Anemia - Amgen

February 22, 2010
Maria Ellis
Executive Secretary for MEDCAC
Centers for Medicare & Medicaid Services
Office of Clinical Standards and Quality,
Coverage and Analysis Group
C1-09-06
7500 Security Boulevard
Baltimore, MD 21244
Re: Medicare Program; Meeting of MEDCAC, March 24, 2010, on Erythropoiesis
Stimulating Agents (ESA) in Anemia Related to Kidney Disease
Dear Ms Ellis:
Joshua Ofman, MD, MSHS
Vice President
Global Coverage and Reimbursement
Global Health Economics
Amgen Inc. (Amgen) is writing to comment on the topics to be addressed at the Medicare
Evidence Development and Coverage Advisory Committee (MEDCAC) March 24, 2010 meeting
on Erythropoiesis Stimulating Agents (ESA) in anemia related to chronic kidney disease (CKD),
which the Centers for Medicare & Medicaid Services (CMS) published on February 19, 2010 on
the website, http://www.cms.hhs.gov/mcd/index_list.asp?list_type=mcac. ESAs are indicated
for the treatment of anemia associated with chronic renal failure (CRF), including patients on
and not on dialysis and Amgen’s comments will be provided separately for these two patient
populations.

As a science-based, patient-care driven company, Amgen is committed to using science and
innovation to dramatically improve people’s lives and is vitally interested in improving access to
innovative drugs and biologicals for Medicare beneficiaries.
Attached you will find our detailed written submission addressing the use of ESAs in anemia
related to CKD. In particular, we call your attention to Appendix 4 which specifically addresses
the voting questions posted on the CMS website February 19, 2010. Amgen appreciates this
opportunity to provide important information and we look forward to the opportunity to discuss
the evidence for ESAs in patients with chronic kidney disease at the upcoming MEDCAC. If
you have any questions or would like to discuss further, then please do not hesitate to contact
me at
Regards,
Joshua J. Ofman, MD, MSHS
Vice President,
Global Coverage and Reimbursement and Global Health Economics

Table of Contents
Page 1
INTRODUCTION............................................................................................................... 4
1. CKD PATIENTS ON DIALYSIS ............................................................................... 7
1.1 Treatment of Anemia in Dialysis Patients Prior to ESAs .............................. 7
1.2 Adverse Sequelae of RBC Transfusion Therapy ......................................... 7
1.3 Development of ESAs and their Impact on Dialysis Patients ..................... 10
1.4 Current ESA Approved Indications............................................................. 12
1.5 Characteristics of the US Dialysis Population ............................................ 13
1.6 Overview of ESA use in General Dialysis Clinical Practice........................ 15
1.7 10 g/dL as the Lower Threshold for ESA Treatment in Dialysis
Patients ...................................................................................................... 17
1.8 Attempts to Improve CV Outcomes in Dialysis Patients by
Targeting Hemoglobin Levels Outside the FDA-Approved Range
(13 g/dL or higher)...................................................................................... 19
1.9 Association of Higher Achieved Hemoglobin and Clinical
Outcome..................................................................................................... 21
1.10 Association of ESA Dose and Adverse Events: Contribution of
Confounding Factors in Dialysis Patients................................................... 21
1.11 New Data from a CV Outcomes Trial Targeting Hemoglobin
Levels above 12 g/dL ................................................................................. 24
1.12 CONCLUSION: ESA THERAPY IS AN ESSENTIAL
TREATMENT FOR DIALYSIS PATIENTS ................................................. 24
2. CKD PATIENTS NOT ON DIALYSIS (CKD-NOD)................................................. 26
2.1 Transfusion Therapy in CKD-NOD Patients............................................... 26
2.2 Adverse Sequelae of Transfusion Therapy ................................................ 27
2.3 Benefits of ESA Therapy in CKD-NOD Patients ........................................ 27
2.4 Anemia and CV Outcomes in CKD-NOD ................................................... 30
2.5 Overview of ESA use in General CKD-NOD Clinical Practice ................... 31
2.6 Trial to Reduce Cardiovascular Events with Aranesp ® Therapy
(TREAT) Trial in Diabetic CKD-NOD Patients............................................ 32
2.7 CONCLUSION: ESA THERAPY IN CKD-NOD PATIENTS IS
AN IMPORTANT TREATMENT OPTION IN THOSE FOR
WHOM TRANSFUSION AVOIDANCE IS A MEANINGFUL
CLINICAL BENEFIT ................................................................................... 34
REFERENCES................................................................................................................ 35
Technical Appendix 1 – Evidence Tables ....................................................................... 40
Technical Appendix 2 - Literature References................................................................ 71
Technical Appendix 3 – PI for Aranesp and EPOGEN ..................................................159
Technical Appendix 4 – Response to Voting Questions ................................................217

List of Tables
Page 2
Table 1. Results from the studies evaluating the association between
EPOGEN ® dose and mortality using methods to address
confounding-by-indication........................................................................ 23
List of Figures
Figure 1. (a) Relationship between the number of transfusions and the risk of
allo-sensitization (measured as panel reactive antibody [PRA]
> 50%; PRA measures anti-human antibodies in the blood) 18 ,
(b) The median time spent on the transplant wait list by the
level of allo-sensitization (measured as PRA < 10% versus
> 10%) and (c) The likelihood of dying while waiting for
transplant 2 .................................................................................................. 8
Figure 2. (a) Long-term (10-year) graft survival of cadaver kidney transplants
according to pre-transplant allo-sensitization (measured as
PRA), and (b) 10-year follow-up of kidney grafts from HLAidentical
sibling donors 20 ............................................................................ 9
Figure 3. Time to transplantation and death among patients 60 years of age
and older on the transplant wait-list 21 ....................................................... 10
Figure 4. Percent of patients receiving a transfusion at baseline and in weeks
1-12 and 13-24 for patients randomized to Epoetin alfa and
placebo treatment (Amgen, data on file).................................................. 11
Figure 5. Improvements in exercise tolerance and physical function observed
when hemoglobin levels were increased to a mean of 10.2
g/dL and 11.8 g/dL with EPOGEN ® compared to placebotreated
patients (Amgen, data on file; EPOGEN ® Prescribing
information [PI]) ....................................................................................... 12
Figure 6. Forest plot of pre-versus post-ESA comparison studies of VO2 23 .................. 13
Figure 7. Transfusion rate in US dialysis patients over time expressed as
percent of patients receiving transfusions in each quarter-year
period 2 . Note: PMMIS and Medicare claims represent all
outpatient transfusion events occurring in the dialysis unit...................... 14
Figure 8. Hemoglobin levels and transfusion rates between 1991 and 2007 2 ............... 14
Figure 9. Mortality rate and hemoglobin levels preceding and following the
introduction of EPOGEN ®27 28 .................................................................. 15
Figure 10. Percentage of hemoglobin measurements in specific ranges for
the dialysis patient population (Amgen data on file; this data is
shared with CMS on an ongoing basis) ................................................... 17
Figure 11. Transfusion risk by the previous month’s hemoglobin level in the
lower arm of the Normal Hematocrit Cardiac Trial (NHCT)
(Amgen, data on file) ............................................................................... 17

Page 3
Figure 12. Transfusion rates by the number of months with an outpatient
hemoglobin below 10 g/dL, and separately below 11 g/dL, in
approximately 160,000 US hemodialysis patients in 2004 34 .................... 18
Figure 13. Percentage of patients with hemoglobin > 13 g/dL for ≥ 3
months 32 .................................................................................................. 20
Figure 14. The risk of transfusion by hemoglobin levels in anemic CKD-NOD
patients treated with ESAs and iron and among those not
receiving treatment (n = 97,636) 66 ........................................................... 27
Figure 15. Transfusion rate among ESA treated CKD-NOD patients with
anemia over time 64 ................................................................................... 28
Figure 16. Hemoglobin levels and associated Kidney Disease Questionnaire
(KDQ) scores for physical symptoms, fatigue, depression,
relationship with others, frustration, and overall KDQ 74
(Clinically meaningful change in KDQ is 0.2-0.5 75 )................................. 29
Figure 17. A meta-analysis of the difference in SF-36 physical function and
vitality scores between higher and lower hemoglobin levels in
CKD-NOD patients; WMD = weighted mean difference 76 ........................ 30
Figure 18. Percentage of hemoglobin measurements in specific ranges for
the CKD-NOD patient population (Amgen data on file; these
data are shared with CMS on an ongoing basis)..................................... 32

Page 4
INTRODUCTION
The approval of the first recombinant erythropoiesis-stimulating agent (ESA), Epoetin
alfa, in 1989 represented an important scientific breakthrough in medicine and
revolutionized the care of patients with anemia of chronic renal failure (CRF). Amgen
Inc. (Amgen), the United States (US) license holder of Epoetin alfa and darbepoetin alfa,
developed erythropoiesis-stimulating agents (ESAs) as supportive therapies to stimulate
red blood cell (RBC) production in order to elevate and maintain hemoglobin
concentrations in patients with CRF and anemia in order to avoid RBC transfusions and
improve patient reported outcomes (PRO). The ability to produce erythropoietin (the
hormone produced by the kidneys to stimulate RBC formation) is impaired in chronic
kidney disease (CKD) patients and this impairment is the primary cause of anemia in this
disease. In the US, Epoetin alfa is marketed under the trade names EPOGEN ® by
Amgen for the treatment of anemia in dialysis, and PROCRIT ® by Centocor Ortho
Biotech Inc. for the treatment of anemia in CKD not on dialysis (NOD). Darbepoetin alfa
is marketed under the trade name Aranesp ® by Amgen for both CKD-NOD and dialysis.
Amgen will focus its comments on the use of EPOGEN ® in dialysis and Aranesp ® in
CKD-NOD, while Centocor Ortho Biotech Inc. will provide comments on PROCRIT ® . In
this document, the term chronic renal failure (CRF) will be used when referring to the US
Food and Drug Administration (FDA)-approved label; otherwise the more commonly
employed term, CKD, will be used. Additionally, the term ESA will be used when
referring to this class of drugs.
ESAs are used to elevate or maintain RBC levels (as manifested by the hemoglobin or
hematocrit determinations) and to decrease the need for RBC transfusions. EPOGEN ®
and Aranesp ® are approved for the treatment of anemia associated with CRF, which
includes patients receiving and not receiving dialysis. Amgen has prepared this
document in response to the Medicare Evidence Development and Coverage Advisory
Committee (MEDCAC) meeting scheduled by the Centers for Medicare and Medicaid
Services (CMS) to review the available evidence on the use of erythropoiesis-stimulating
agents (ESAs) to manage anemia in patients who have chronic kidney disease. Since
the regulatory approval of ESAs for the treatment of anemia in CKD, safety concerns
have been raised based on the results of randomized controlled trials (RCTs) designed
to evaluate the potential for cardiovascular (CV) and survival benefit when normalizing
hemoglobin with ESAs. These trials treated patients to higher hemoglobin targets—

above the current labeled range (hemoglobin 10-12 g/dL); the hemoglobin targets
evaluated in these studies do not reflect how ESAs are used in clinical practice.
Page 5
There are significant differences in the characteristics of dialysis and CKD-NOD patients,
including demographic characteristics, the extent of morbidity and mortality, and the
intensity of clinical management required for their treatment 1, 2 . Most notable among
these differences is that dialysis patients have significantly more co-morbid conditions,
such as: diabetes, hypertension, heart failure, infectious complications, and a more
profound erythropoietin deficiency. These patients are subject to continuous blood loss
from the dialysis procedure, which results in more profound anemia that is nearly
universal in the absence of treatment. Dialysis patients depend on regular dialysis
treatments to sustain life; typically, death ensues within two weeks when dialysis
treatment is withdrawn 3 .
In dialysis patients, ESAs are an important therapy in the management of anemia and
have a positive benefit to risk profile when used according to the FDA-approved label to
reduce the need for RBC transfusions. ESAs are effective at increasing hemoglobin
levels in order to reduce the need for transfusions and to improve physical function and
exercise tolerance in dialysis patients. Transfusion avoidance protects against
cumulative hazards, especially allo-sensitization, which is the generation of antibodies to
foreign antigens that may impair or preclude eligibility for renal transplant, and increase
the likelihood of transplant rejection. Even the recent technique of removing white cells
from blood prior to transfusion (leuko-reduction) does not decrease the hazard of allosensitization
4 . Clinical trials and observational data have demonstrated that the risk of
transfusion increases substantially when hemoglobin levels fall below 10 g/dL; this
supports 10 g/dL as the appropriate lower limit for clinical benefit and aligns with the
labeled range. With increasing hemoglobin, the transfusion benefit continuously
improves through the labeled hemoglobin range of 12 g/dL. Because of the intrinsic
hemoglobin variability seen in dialysis patients, a 2 g/dL range is appropriate, thus
supporting the upper hemoglobin level of 12 g/dL, aligning with the label. The
hemoglobin target range accommodates hemoglobin variability and allows physicians to
treat individual patients to attain the hemoglobin level necessary to avoid transfusion.
Additionally, near-complete surveillance of the dialysis population has not provided
evidence of an increased rate of death when patients are treated to the FDA-approved
ESA labeled hemoglobin range. The evidence supports that the labeled hemoglobin

Page 6
range of 10 to 12 g/dL is necessary and prudent to maximize the benefit of transfusion
avoidance in a way that recognizes and minimizes the risk of cardiovascular (CV) events
that have been observed in trials of high hemoglobin targets (≥ 13 g/dL). This range
allows physicians to manage anemia in dialysis patients effectively.
In CKD-NOD patients, while the prevalence of anemia is lower, anemia can be severe
in some patients and transfusions are more common than has been appreciated.
Among those with severe anemia, the administration of ESAs is appropriate to avoid the
risks of transfusion, and, in particular, the risk of allo-sensitization, which can reduce
both transplant eligibility and graft survival. ESA therapy should be initiated when
hemoglobin levels decline below 10 g/dL and patients are iron replete. The
demonstrated benefit of transfusion reduction with ESA use is observed when
hemoglobin levels are treated to above 10 g/dL. ESA therapy should be individualized
to achieve and maintain hemoglobin between 10 and 12 g/dL. As with dialysis patients,
CKD-NOD patients demonstrate intrinsic hemoglobin variability and the practical
limitations of managing hemoglobin require the use of a hemoglobin range in order to
administer ESAs to maximize the clinical benefit for each patient. A hemoglobin level of
12 g/dL as the upper end of the hemoglobin range will allow physicians the necessary
discretion to manage individual patients so as to reduce transfusions. Avoiding a
hemoglobin target > 12 g/dL provides a safety margin against the higher hemoglobin
target in the Trial to Reduce Cardiovascular Events with Aranesp ® Therapy (TREAT)
(≥ 13.0 g/dL) and the Correction of Hemoglobin and Outcomes in Renal Insufficiency
(CHOIR) (~13.5 g/dL) where risks have been identified. The evidence supports that
ESAs are an important therapy for anemic CKD-NOD patients in whom transfusion
avoidance is a meaningful clinical goal and that ESAs have a positive benefit to risk
profile when used according to the FDA-approved labeling in these patients.
Nonetheless, in response to the results of the TREAT trial, Amgen is working with global
regulatory authorities on potential changes to labeling to provide more detailed guidance
to physicians on optimizing dosing to achieve hemoglobin levels within the target range.
In addition, safety warnings have been strengthened; further labeling changes are likely.
In this evidence review and accompanying supporting documents, the use of ESAs in
CKD patients on dialysis and CKD-NOD patients are presented separately. There is a
technical summary at the end of the review which includes a) relevant citations including
those reporting the results of the registrational trials, and b) evidence tables

summarizing the published literature on transfusion patterns, costs of transfusions,
hemoglobin variability, hospitalization, health resource utilization and patient reported
outcomes in CKD patients. These evidence tables include the most commonly cited
references; a systematic review was not conducted.
Page 7
1. CKD PATIENTS ON DIALYSIS
In 2007 in the US, there were over 527,000 patients with end-stage renal disease
(ESRD) who required either dialysis or a kidney transplant for survival 2 . Typically,
patients receive dialysis treatments three times per week. During each dialysis session,
their blood is circulated through a dialyzer that removes solutes (such as urea and
accumulated fluids that are normally removed by the kidney) and restores electrolyte
balance. Dialysis treatment, while life sustaining, compounds the anemia 5 caused by
deceased erythropoietin levels 6, 7 . Blood is lost during each dialysis procedure 5 , from
frequent blood sampling to monitor laboratory parameters, and from increased bleeding
tendency attributable to anticoagulation (heparin) therapy administered during the
dialysis. There is also an increased risk of gastrointestinal bleeding and re-bleeding
after therapy in dialysis patients 8, 9 . Blood loss is estimated to be between 2.5 and 5.1 L
(5-10 Units) of blood annually 5 , up to double the normal circulating blood volume. Thus,
there are many factors contributing to the anemia in dialysis patients, and, in the
absence of ESA therapy, anemia is often times very severe.
1.1 Treatment of Anemia in Dialysis Patients Prior to ESAs
Prior to the development of ESAs, the treatment options for anemia were limited to RBC
transfusions, and to a smaller extent, androgen and iron therapy 10 . Androgens and iron
therapy had only modest efficacy and substantial side effects, leaving RBC transfusions
as the mainstay of anemia therapy in the pre-ESA era. Data from the pre-ESA era
indicate that patients had a mean hemoglobin of approximately 7 g/dL 11 .
1.2 Adverse Sequelae of RBC Transfusion Therapy
RBC transfusions are by their nature only transiently effective in a population that is
chronically unable to produce sufficient RBCs 12 . They carry a range of hazards including
transmission of blood borne viral diseases, transfusion reactions, acute volume and
potassium overload, and more chronically, iron overload 10, 13-16 . In the pre-ESA era, 55-
60% of dialysis patients received transfusions to avoid severe anemia 11 (Amgen, data on
file).

Page 8
Perhaps the RBC transfusion risk that has the greatest impact on the lives of dialysis
patients is the potential for allo-sensitization to foreign antigens 2, 17, 18 , as development of
such antibodies can delay or preclude kidney transplantation and impair the function of
kidney transplants that do occur 2, 19 (Figure 1a-1c).
Figure 1. (a) Relationship between the number of transfusions and the risk of allosensitization
(measured as panel reactive antibody [PRA] > 50%; PRA measures
anti-human antibodies in the blood) 18 , (b) The median time spent on the transplant
wait list by the level of allo-sensitization (measured as PRA < 10% versus > 10%)
and (c) The likelihood of dying while waiting for transplant 2 .
Early research suggested that a single transfusion event could result in sensitization in
~33% of patients 4 . More recent evidence using enhanced methods for detecting allosensitization
(methods that are now being adopted by the United Network of Organ
Sharing [UNOS]) suggest that as many as 75% of patients are sensitized after a single
transfusion (D Tynan, personal communication; Amgen, data on file). Sensitization by
transfusion is avoidable in these patients; the other two mechanisms that can cause
sensitization, pregnancy and previous transplant 17 are not readily mitigated. The risk of
transfusion and allo-sensitization may be reduced but not eliminated by leuko-reduction 4 .
Allo-sensitization may result in a decline in graft survival 17 among cadaveric
transplanted patients who are sensitized (Figure 2a) and also among HLA-identical
sibling donors (Figure 2b) 20 . This last finding is notable because these transplants are
expected to have lower immunological barriers to transplantation and greatest overall
success.

Page 9
Figure 2. (a) Long-term (10-year) graft survival of cadaver kidney transplants
according to pre-transplant allo-sensitization (measured as PRA), and (b) 10-year
follow-up of kidney grafts from HLA-identical sibling donors 20 .
Kidney transplant is the preferred ESRD treatment modality because successfully
transplanted patients have superior survival and quality of life and significantly reduced
health resource utilization and cost 2 . Receiving a transfusion can limit a dialysis
patient’s likelihood of receiving a transplant by prolonging their wait for a matching
kidney. Patients who wait longer for a transplant are more likely to die rather than
receive a kidney 19 . The effect of prolonged waiting time on the risk of death is most
pronounced among patients 60 years and older. Recent estimates suggest that nearly
50% will die within 5 years if they have not received a transplant 21 (Figure 3).

Page 10
Figure 3. Time to transplantation and death among patients 60 years of age and
older on the transplant wait-list 21 .
In the pre-ESA era, while necessary for management of anemia, the risks related to
transfusion limited their use, and as a result, hemoglobin levels were maintained at a
level of approximately 7 g/dL 11 . Since patients were maintained at these low hemoglobin
levels, they experienced persistent, severe fatigue and had restrictions on physical
functioning—all of which contributed to a poor quality of life.
1.3 Development of ESAs and their Impact on Dialysis Patients
The primary registration trials (five studies) used for the approval of Epoetin alfa
demonstrated correction of anemia and virtual elimination of transfusions (>90%
reduction) in patients treated with ESAs to a mean hemoglobin of 11.7 g/dL (within the
target range of 10.7 to 12.7 g/dL). In placebo treated patients, hemoglobin levels
remained low (< 7 g/dL) and these patients continued to receive multiple transfusions,
while the Epoetin alfa treated group became nearly transfusion independent (Amgen
Clinical Study Report, data on file) (Figure 4).

Page 11
Figure 4. Percent of patients receiving a transfusion at baseline and in weeks 1-12
and 13-24 for patients randomized to Epoetin alfa and placebo treatment (Amgen,
data on file).
The original registration studies also evaluated the impact of raising hemoglobin levels
with ESA therapy on physical functioning and quality of life. Exercise tolerance (meters
walked) and physical function improved compared to placebo in patients receiving
Epoetin alfa with both achieved hemoglobin levels of 10.2 g/dL as well as 11.7 g/dL at 6
months 22 . These improvements were statistically significantly different compared to
placebo, as well as clinically meaningful (Figure 5). Exercise tolerance and physical
function were significantly improved with ESA treatment, rising by approximately 30%
after 6 months of treatment as defined by the meters walked 11, 22 .

Page 12
Figure 5. Improvements in exercise tolerance and physical function observed
when hemoglobin levels were increased to a mean of 10.2 g/dL and 11.8 g/dL with
EPOGEN ® compared to placebo-treated patients (Amgen, data on file; EPOGEN ®
Prescribing information [PI])
1.4 Current ESA Approved Indications
EPOGEN ® and Aranesp ® are approved for the treatment of anemia associated with
CRF, which includes patients receiving and not receiving dialysis. ESAs are used to
elevate or maintain RBC levels (as manifested by the hemoglobin or hematocrit
determinations) and to decrease the need for transfusions in these patients. The
EPOGEN ® label also includes a quality of life benefit (improved exercise tolerance and
physical functioning) as demonstrated in the Canadian Erythropoietin Study Group 22
(Amgen, data on file). The benefit on patient reported outcomes, including improvement
in energy, physical function and exercise tolerance (as measured by VO2 max), resulting
from treatment with ESAs to hemoglobin levels greater than 10 g/dL, has been shown
consistently in subsequent open-label and observational studies and summarized in a
recent meta-analysis 23 (Figure 6).

Figure 6. Forest plot of pre-versus post-ESA comparison studies of VO2 23
Page 13
1.5 Characteristics of the US Dialysis Population
In the US, extensive data on the health care provided to the majority of dialysis patients
(Medicare primary insurer) is captured by the US Renal Data System (USRDS).
Consequently, dialysis patients are subject to near complete surveillance; patient care
information including medications, biochemical parameters and morbidity and mortality
outcomes are systematically collected. This pharmacovigilance system is unique among
disease states. Since USRDS has been in continuous operation for over 20 years, the
evaluation of trends in treatments and outcomes in the dialysis population is possible.
In the pre-ESA era, a substantial fraction of transfusions were administered for chronic
anemia in the outpatient dialysis unit. Almost immediately following the introduction of

EPOGEN ® into clinical practice (June 1989), the transfusion rate among US
hemodialysis patients fell sharply (Figure 7).
Page 14
Figure 7. Transfusion rate in US dialysis patients over time expressed as percent
of patients receiving transfusions in each quarter-year period 2 . Note: PMMIS and
Medicare claims represent all outpatient transfusion events occurring in the
dialysis unit.
By 1992 almost 90% of US dialysis patients received ESA therapy and this treatment
prevalence continues today 24 . In 1994, the EPOGEN ® label was changed to expand the
10 to 11 g/dL hemoglobin range to 10 to 12 g/dL, a change that was supported by
clinical practice guidelines. Between 1992, when the mean population hemoglobin was
~9.8 g/dL, and 2000, when, with increased treatment intensity, the mean population
hemoglobin had risen to ~11.2 g/dL, the total transfusion rate (inpatient plus outpatient)
was again halved 25 (Figure 8).
Figure 8. Hemoglobin levels and transfusion rates between 1991 and 2007 2
Mean Hb (g/dL)
13.00
12.00
11.00
10.00
9.00
1991 1993 1995 1997 1999 2001 2003 2005 2007
Year
Mean Hb
Transfusions
16.00
12.00
8.00
4.00
0.00
Transfusion (% pateints/quarter)

Page 15
These data provided important surveillance information on clinical practice and
population level blood supply utilization supporting the label range (10-12 g/dL) as an
effective strategy for reducing RBC transfusions in outpatient dialysis facilities. The
decline in transfusions was primarily in the outpatient setting; the majority of remaining
transfusions occur in the inpatient setting (75%) 26 .
As part of the routine surveillance of US dialysis patients by USRDS, important clinical
outcomes such as mortality are evaluated annually. These data indicate that since the
introduction of EPOGEN ® into clinical practice and its wide spread adoption as the
primary treatment for anemia over the past 20 years, the overall mortality rate in the
dialysis population has not increased (Figure 9).
Figure 9. Mortality rate and hemoglobin levels preceding and following the
®27 28
introduction of EPOGEN
Thus, there is evidence of an important and tangible clinical benefit of transfusion
avoidance, and no evidence of an overt safety signal when patients are treated
according to the FDA-approved label in general clinical practice.
1.6 Overview of ESA use in General Dialysis Clinical Practice
The goals of the use of medicines in clinical practice are to maximize benefit and
minimize risk consistent with FDA-approved labeling. Reimbursement policies have
been designed, in many respects, to enable physicians to treat individual patients
consistent with the approved label. For ESAs, the current policy allows physicians to
individualize treatment to achieve and maintain hemoglobin levels between 10 and
12 g/dL so as to reduce the need of RBC transfusions and avoid sustained high
hemoglobin levels. Research over the past 10 years has described the considerable

Page 16
variation in hemoglobin levels in dialysis patients 29-31 and has highlighted the difficulty of
maintaining patients within a narrow hemoglobin range. The current CMS payment
policy is clinically-based and takes into account the transient fluctuations in hemoglobin
levels that commonly occur in the dialysis patient population.
Reimbursement policies for ESAs have been dynamic over the last two decades,
reflecting changes in label and practice guidelines for the treatment of anemia in CKD
and the appropriate administration of ESAs with regard to the safety and efficacy profile.
Changes made to the EMP effective January 1, 2008 have had the desired effect in
decreasing the occurrence of excessive hemoglobin levels in dialysis patients; the mean
population hemoglobin has also decreased. Recently published data indicate that mean
hemoglobin levels in the dialysis patient population has decreased from 12.08 g/dL (SD
1.48) in June 2006 to 11.71 g/dL (SD 1.35) in November 2008 32 . Additional surveillance
data on ~87% of dialysis centers in the US indicates that as of December 2009, the
mean hemoglobin continues to decline and is approximately 11.54 g/dL (Amgen data on
file). Analysis of surveillance data also indicates that the majority of physicians (98%)
are acting to appropriately reduce ESA doses when hemoglobin levels exceed the FDAapproved
target range 33 . Examining specific hemoglobin categories, the percentage of
patients with a monthly hemoglobin value greater than 12 g/dL has declined (from 53.1
to 34.1%) and the proportion within the labeled range has increased from 40.7 to 56.8%.
(Amgen data on file) (Figure 10). Amgen routinely shares this surveillance data with
CMS on an ongoing basis.

Page 17
Figure 10. Percentage of hemoglobin measurements in specific ranges for the
dialysis patient population (Amgen data on file; this data is shared with CMS on
an ongoing basis)
1.7 10 g/dL as the Lower Threshold for ESA Treatment in Dialysis
Patients
As demonstrated in both clinical trials in dialysis patients and in observational data, the
risk of transfusion rises significantly as the outpatient hemoglobin in the preceding month
falls below 10 g/dL and this risk increases when hemoglobin levels drop further (Amgen,
data on file) (Figure 11).
Figure 11. Transfusion risk by the previous month’s hemoglobin level in the lower
arm of the Normal Hematocrit Cardiac Trial (NHCT) (Amgen, data on file)
Hazard Ratio (95% CI)
8
7
6
5
4
3
2.5
2
1.5
1.2
1
0.8
0.6
0.4
The Risk of Transfusion by the Previous Month's Hemoglobin Level
< 9 9 - < 10 10 - < 11 11 - < 12 >=12
Data Source: NHCT study, data on file
Hemoglobin level (g/dL)

Page 18
An analysis of nearly 160,000 US hemodialysis patients in Medicare data further indicate
that the risk of transfusion increases substantially the longer hemoglobin levels remain
below 10 g/dL 34 (Figure 12).
Figure 12. Transfusion rates by the number of months with an outpatient
hemoglobin below 10 g/dL, and separately below 11 g/dL, in approximately
160,000 US hemodialysis patients in 2004 34 .
The majority of the evidence from RCTs, as well as observational data, have strongly
supported that maintaining hemoglobin levels above 10 g/dL is essential for avoiding
transfusion, fatigue and decline in physical function. The importance of maintaining
hemoglobin levels above 10 g/dL is recognized by the nephrology community as well as
CMS, and is currently incorporated as a quality metric by which dialysis units are
evaluated 35 . However, because hemoglobin levels are known to vary within individuals
over time (intra-patient variability) 29, 31 , and because of the delayed response between
ESA dosing and hemoglobin changes 36 , it is extremely difficult to maintain hemoglobin
levels within a narrow range in many, if not most, patients 29 .
In the early registration trials, variation in hemoglobin levels was observed in both
placebo-treated and ESA-treated patients (the mean intra-patient standard deviation
[SD] in hemoglobin levels was 0.6 g/dL) 37 ; this degree of variability continues to be
observed in more recent clinical trials. In current clinical practice, the population mean
intra-patient hemoglobin SD is near 0.9 g/dL and accounts for ~80% of the total
variability (1.4 g/dL) that is observed in cross-sectional evaluations of hemoglobin in the
entire US dialysis population 34 .

Page 19
Both the original registration trials and the current FDA-approved ESA labels refer to a
2 g/dL hemoglobin range. The need for this range is based on the following: (i) the
recognized need to maintain hemoglobin levels above 10 g/dL to avoid transfusion; (ii)
the inherent variability in patient hemoglobin levels over time; and (iii) the continued
reduction in transfusions that is observed up to a hemoglobin level of 12 g/dL. Using this
range, physicians are effectively managing hemoglobin levels according to patient needs
in order to avoid unnecessary RBC transfusions.
The influence of hemoglobin variability on the management of anemia in dialysis patients
is well recognized by CMS and has been incorporated into their National Claims
Monitoring Policy for ESAs in hemodialysis patients (EMP). The EMP recognizes that in
administering ESAs to achieve and maintain the 10-12 g/dl hemoglobin range, their
reimbursement policy should account for this inherent variability 38 .
1.8 Attempts to Improve CV Outcomes in Dialysis Patients by Targeting
Hemoglobin Levels Outside the FDA-Approved Range (13 g/dL or
higher)
A number of observational studies had evaluated the association between anemia and
increased risk of CV outcomes in CKD patients 39-41 . Using the available surveillance
data in USRDS, studies have consistently shown that hemoglobin levels below 10 g/dL
are associated with increased rates of hospitalization and death and greater healthcare
resource utilization 42, 43 , and patients who are able to achieve higher hemoglobin levels
(10-12 g/dL) had lower hospitalization and mortality rates. To date, no study has
evaluated CV morbidity and mortality, or healthcare resource utilization benefits when
treating with ESAs to within the labeled range. However, studies were undertaken to
test the hypothesis that normalization or near normalization of hemoglobin levels
(hemoglobin of 13.5 and 14 g/dL) in CKD would result in decreased CV morbidity and
mortality; the results of these RCTs did not confirm this hypothesis and hazard, instead
of benefit, was demonstrated.
The first and only randomized controlled CV outcome trial to address this issue in
dialysis patients, the Normal Hematocrit Cardiac Trial (NHCT), was conducted in
subjects with pre-existing CV disease or heart failure and anemia 44 . The second trial
was in CKD-NOD patients, the Correction of Hemoglobin and Outcomes in Renal
Insufficiency (CHOIR) 45 . NHCT was stopped for futility and CHOIR was stopped for
hazard before the planned study completion. Although CHOIR was conducted in CKD-
NOD patients, it is described here as part of a full discussion of safety with respect to

Page 20
ESA use. Both studies evaluated higher-than-approved hemoglobin targets: (i) NHCT
compared the impact of a hemoglobin target of 14.0 g/dL to a target of 10.0 g/dL on the
time to mortality or nonfatal myocardial infarction and found an excess of events in the
high target arm (hazard ratio [HR] = 1.3, 95% confidence interval [CI]: 0.9-1.8); (ii)
CHOIR compared the impact of a hemoglobin target of 13.5 g/dL to a target of 11.3 g/dL
on the composite endpoint of death, myocardial infarction, hospitalization for congestive
heart failure (without renal replacement therapy) and stroke, and found an excess of
events in the high target arm (HR = 1.34, 95%CI: 1.03 – 1.74).
Based on the results of the NHCT trial, the United States PI 46 for ESAs was revised in
1996 to highlight the observed risks and to caution against normalization of hemoglobin
levels in CKD. Following the publication of the CHOIR results in 2006, the US
prescribing information (USPI) for ESAs were updated in 2007 to include the risks
reported in the study and a Boxed Warning.
It is important to note that the mean achieved hemoglobin level in general clinical
practice (11.7 g/dL in 2008) 32 is much lower than the mean achieved hemoglobin level of
patients in the high target arm of NHCT (13.5 g/dL). Moreover, the proportion of dialysis
patients with a hemoglobin level persistently above 13 g/dL for three consecutive months
decreased to only 2.2% n 2008 32 (Figure 13).
Figure 13. Percentage of patients with hemoglobin > 13 g/dL for ≥ 3 months 32

Page 21
1.9 Association of Higher Achieved Hemoglobin and Clinical Outcome
Patients with lower hemoglobin levels have worse clinical outcomes and quality of life 23,
43
, and this may be due in part to the extent of their comorbidity and other clinical factors
that may cause low hemoglobin. Safety information from RCTs (CHOIR and NHCT)
consistently demonstrates higher hazard associated with randomization to higher
hemoglobin targets. At the same time, analyses of these same trials 28, 44 as well as
surveillance data, suggest that patients who are able to achieve higher hemoglobin
levels experience better clinical outcomes. This paradox of different clinical outcomes
between targeting, as opposed to being able to achieve higher hemoglobin levels, points
out the confounding effect of patients’ health status on the clinical outcomes related to
anemia management.
1.10 Association of ESA Dose and Adverse Events: Contribution of
Confounding Factors in Dialysis Patients
The totality of the evidence informing the benefit and risks of ESAs in CKD patients was
reviewed at the September 2007 joint meeting between the FDA Cardiovascular and
Renal Drugs Advisory Committee (CRDAC) & Drug Safety and Risk Management
Advisory Committee, which included results from CHOIR as well as a review of the
evidence regarding the role of ESA dose on outcomes. In the NHCT, higher ESA doses
were required to achieve the higher target hemoglobin level 44 . This raised the concern
that the risks seen in this high hemoglobin target study were, in part, attributable to ESA
dose. Additional evidence suggesting that higher ESA doses increased mortality risk
was provided by two different analyses of US hemodialysis patients using USRDS data
and found that higher ESA doses were associated with a significantly elevated mortality
risk across all hemoglobin levels 47, 48 . Amgen undertook an extensive evaluation of the
data from RCTs and observational studies to better understand the potential contribution
of ESA dose on mortality and CV and/or thromboembolic risk. These analyses indicate
that there are three important considerations when examining the relationship between
ESA dose and clinical outcomes:
• ESA dose and hemoglobin are strongly correlated because ESA dose is
titrated in response to specific hemoglobin levels.
• Hemoglobin response to ESAs is influenced by the patient's overall health
status; sicker patients require higher doses of ESAs to maintain hemoglobin.

Page 22
• Hemoglobin response to ESAs is not static, but rather, changes within each
individual over time.
In the course of clinical management of dialysis patients, ESA dosing decisions are
made based on a patient’s preceding hemoglobin concentration 33 . Thus, there is a tight
correlation between hemoglobin and ESA dose. This relationship is further complicated
because each patient’s overall health status influences their hemoglobin response to
ESA dosing changes 49 . Since ESA dosing is administered chronically, the hemoglobindose
relationship can change over time 50 , and thus, the dose that is necessary to
maintain a hemoglobin concentration varies over time. An analogous situation exists in
the case of insulin dose and glucose control in the critical care setting. While
prospective randomized trials targeting blood glucose to lower levels using intensive
insulin therapy have demonstrated a significant reduction in mortality 51 , analysis of blood
glucose levels and administered insulin doses has shown a consistent association
between higher insulin doses and greater mortality, regardless of the prevailing blood
glucose level 52 . The authors of this latter investigation recognized the complexity of
targeting blood glucose levels and considered that the control of glucose levels, rather
than insulin doses, was the important determinant of the beneficial effects observed with
lower target blood glucose levels.
This same conceptual view can be applied to ESA therapy and the clinical practice in
which the physician attempts to maintain hemoglobin concentrations within a specified
range over time. It was acknowledged at the 2007 CRDAC meeting that the excess
mortality risk seen in patients receiving high ESA doses was likely related to their
inability to respond to increased ESA doses (referred to as ESA hyporesponsiveness),
rather than the dose itself 53, 54 . Thus, in 2007 following the CRDAC meeting, the concept
of hyporesponsiveness and its management was introduced into the USPI.
Following the 2007 CRDAC meeting, the EPOGEN ® and Aranesp ® labels were further
revised:
• An update was made to the Boxed Warning
• The hemoglobin range of 10 to 12 g/dL was maintained
• Guidance was added for the treatment of patients who did not attain a
hemoglobin level within the range of 10 to 12 g/dL despite the use of
appropriate EPOGEN ® dose titrations over a 12-week period

Page 23
Since 2007, additional analyses have been conducted to further understand the interrelationship
between hyporesponsiveness, EPOGEN ® dosing and adverse clinical
outcomes in dialysis patients. It is now more widely understood that these initial
observations of excess risk observed in patients who require greater EPOGEN ® doses
are largely attributable to patient characteristics, worsening clinical status and poor
hemoglobin response rather than EPOGEN ® dose 53-55 . Patients with the highest
EPOGEN ® doses are those most likely to have low hemoglobin levels, greater CV
disease burden, more inflammation and malnutrition, increased hospitalization
frequency, and are more likely to be dialyzing with a catheter (which promotes infection
and is associated with increased mortality) rather than a permanent vascular access 55-
57 .
The prevalence of higher EPOGEN ® doses in patients with worse overall prognosis
results in confounding-by-indication and can produce biased results attributing risk to
higher doses 50 . Several studies using different analytical techniques to address this
confounding have shown that EPOGEN ® dose is not an independent predictor of
mortality 53, 58-60 60 (Table 1). Importantly, in subsequent analyses with more rigorous
control for confounding, the lead investigators (Zhang, Thamer and Cotter) who originally
suggested that higher EPOGEN ® doses increase mortality risk 48 , no longer demonstrated
this association 59 .
Table 1. Results from the studies evaluating the association between EPOGEN ®
dose and mortality using methods to address confounding-by-indication
Author N Comparison HR (95% CI)
Limited baseline adjustment
Zhang 2004 48 94,569 >22K vs. 40K vs. 20-30K U/wk 0.91 (0.67, 1.22)
Bradbury 2008 55 22,955 Log EPO dose 1.01 (0.99, 1.03)
Wang 2010 60 27,791 >49K vs. 37.5% vs. 0-12.5%^ 0.86 (0.69, 1.08)
^ Average monthly dose change over 3 months among patients with low hemoglobin levels

Page 24
1.11 New Data from a CV Outcomes Trial Targeting Hemoglobin Levels
above 12 g/dL
Since the review of the totality of the evidence at FDA CRDAC 2007, the results of
TREAT, a large Amgen-sponsored phase III prospective randomized clinical trial, have
become available 61 . TREAT was a study designed in 2002, and conducted between
2004 and 2009, to address the hypothesis that correction of anemia in CKD-NOD
patients with mild anemia (hemoglobin < 11 g/dL at enrollment) and type 2 diabetes
would improve CV morbidity and mortality. The detailed results for TREAT are
described in the CKD-NOD section. In brief, the study did not meet the primary objective
of showing an improvement in CV morbidity and mortality (HR = 1.05, 95% CI: 0.94 –
1.17). Patients in the Aranesp ® treated arm experienced an almost two-fold increased
risk of stroke compared with the placebo group (HR = 1.92, 95% CI: 1.38-2.68 with an
annualized incidence rate of 1.1% in the placebo arm and 2.1% in the Aranesp ® treated
arm. Consistent with Amgen’s previous response to safety signals identified in clinical
trials, Amgen has proactively revised the EPOGEN ® and Aranesp ® labels to include this
risk.
1.12 CONCLUSION: ESA THERAPY IS AN ESSENTIAL TREATMENT FOR
DIALYSIS PATIENTS
ESAs are an essential therapy in the management of anemia in dialysis patients and
have a positive benefit to risk profile when used according to the FDA-approved label.
ESAs are effective at increasing hemoglobin levels in order to avoid transfusions and to
improve physical function and exercise tolerance in dialysis patients. Transfusion
avoidance protects against cumulative hazards—especially allo-sensitization to foreign
antigens—and improves the eligibility for and outcome of renal transplant. This is a
particularly relevant issue for many potential transplant candidates who often times
spend many years on dialysis, thereby increasing both their exposure to the risks of
transfusions as well as their mortality rates from having been on dialysis. A lower
hemoglobin limit of 10.7 g/dL was studied in the original registrational clinical trials.
Subsequently, additional clinical trials and observational data have demonstrated that
the risk of transfusion increases substantially when hemoglobin levels fall below 10 g/dL;
this supports 10 g/dL as the appropriate lower limit of the labeled range. The transfusion
benefit continuously improves through the labeled hemoglobin range of 12 g/dL, and
because of the intrinsic hemoglobin variability seen in dialysis patients; a 2 g/dL range is
appropriate with an upper hemoglobin level of 12 g/dL. Additionally, the near-complete

Page 25
surveillance of the US dialysis population by USRDS has not provided evidence of an
increased rate of death when patients are treated to the FDA-approved ESA labeled
hemoglobin range. Thus, the labeled hemoglobin range of 10 to 12 g/dL is necessary
and prudent to maximize the benefit of transfusion avoidance and minimize the risk of
CV events associated with high hemoglobin targets (≥ 13 g/dL), and allows physicians to
effectively manage anemia in dialysis patients.

Page 26
2. CKD PATIENTS NOT ON DIALYSIS (CKD-NOD)
There are an estimated 26 million people in the US with CKD-NOD, of which,
approximately 22 million have moderate to severe loss of kidney function (estimated
glomerular filtration rate [eGFR] 15 to 60 mL/min/1.73 m 2 ) 62 . As renal function
decreases and endogenous erythropoietin production declines, anemia develops 7 .
While anemia is ubiquitous and severe in dialysis patients (in the absence of treatment),
the prevalence of anemia is lower in CKD-NOD. However, it is estimated that > 50% of
Medicare CKD-NOD patients have some degree of anemia 2 ; the prevalence of anemia
increases with the stage of CKD 63 . The difference in the prevalence of anemia in CKD-
NOD relative to dialysis is a key distinguishing clinical characteristic between these two
populations.
2.1 Transfusion Therapy in CKD-NOD Patients
While the prevalence of anemia in CKD-NOD is lower than in patients who are on
dialysis, the anemia in CKD-NOD can be severe and the transfusion burden in these
patients is surprisingly high. In Medicare patients with CKD-NOD and chronic anemia,
the annual transfusion rate ranged from 17% to 25% between 1992 and 2004, and the
risk was three-fold higher than in CKD-NOD patients without anemia, and 10 fold higher
than those without CKD 64 . In 2004 there were approximately 400,000 anemic CKD-NOD
patients covered by Medicare; this roughly translates to 60,000 to 100,000 transfusion
events annually in the Medicare population alone. In the Veteran’s Administration (VA)
healthcare system, transfusion rates are similarly elevated in the anemic CKD-NOD
population 65 . Transfusions are significantly more common in non-treated patients when
compared to patients receiving ESAs and iron, and increase markedly when hemoglobin
levels fall below 10 g/dL 66 (Figure 14).

Page 27
Figure 14. The risk of transfusion by hemoglobin levels in anemic CKD-NOD
patients treated with ESAs and iron and among those not receiving treatment (n =
97,636) 66
2.2 Adverse Sequelae of Transfusion Therapy
RBC transfusions carry a range of hazards including transmission of viral disease and
transfusion reactions. As with dialysis patients, transfusion related complications such
as acute volume and potassium overload, and more chronically, iron overload. can be
problematic for CKD-NOD patients 10, 13-16 . Finally, and of unique importance to CKD
patients, RBC transfusions can result in allo-sensitization to foreign antigens 2, 19 that can
delay or preclude kidney transplantation 2 and impact overall graft survival in transplanted
patients 20 . Currently, 15% of transplants occur in CKD-NOD patients before they initiate
dialysis 2 and the numbers are growing. USRDS projects that ~700,000 patients will be
receiving dialysis or be transplanted by the year 2020 2 . The largest group progressing to
dialysis is projected to be 45-64 year-olds, all of whom will become Medicare
beneficiaries upon developing ESRD; these patients will likely be prime candidates for
renal transplantation. Avoidance of transfusion to maintain transplant eligibility is of
critical importance because kidney transplant is the preferred treatment modality for
ESRD. Successfully transplanted patients have superior survival and quality of life and
incur lower health care costs 2 .
2.3 Benefits of ESA Therapy in CKD-NOD Patients
The clinical benefit of transfusion avoidance in anemic dialysis patients was
demonstrated by the registrational clinical trials with EPOGEN ® and established
hemoglobin as the key outcome for approval of ESAs in the non-dialysis setting. The

Page 28
Aranesp ® clinical trial program evaluated hemoglobin response in CKD-NOD subjects
and demonstrated that hemoglobin levels could be raised and maintained within targeted
hemoglobin range; approval for this indication was granted by the FDA in 2001 67 . In the
original registration studies with EPOGEN ® in dialysis patients, transfusion avoidance
occurred when hemoglobin levels were raised above 10 g/dL 68 , and the same efficacy
was demonstrated in an open-label single-arm study of anemic CKD-NOD patients 69 . In
this study, transfusion events were reduced by ~70% when hemoglobin levels were
raised above 10 g/dL. This benefit has also been seen in general clinical practice. In
the Medicare population with CKD-NOD and anemia, transfusion rates have declined
from 30 to 15% among patients treated with ESAs between 1995 and 2004 64 (Figure
15).
Figure 15. Transfusion rate among ESA treated CKD-NOD patients with anemia
over time 64
There has been, and continues to be, significant interest in the nephrology community
toward better understanding the effects of treatment with ESAs to hemoglobin levels
above 10 g/dL on CKD-NOD patients’ symptoms and functioning from the patient’s
perspective; treatment of these patient reported domains is not an approved indication
for Aranesp ® . A number of trials, including double-blind placebo-controlled, open-label
randomized controlled trials as well as single-arm and observational studies have been
conducted to study this question. In summary, all of these studies suggest
improvements in patient-reported outcomes, although the degree of benefit varies

Page 29
across studies. Studies which compare achieved hemoglobin levels below 10 g/dL to
those above 10g/dL demonstrate more improvement than those which compare
improvements in patient reported outcomes between high and low hemoglobin targets.
For example, in an open-label RCT, SF-36 scores decreased in the untreated arm
(achieved hemoglobin of 8.9 g/dL) for both physical function and energy while there was
a significant increase in these subscale scores in the treated arm (achieved hemoglobin
of 10.5 g/dL) 70 . Similar improvements have been demonstrated in fatigue, and activity
levels 71-73 . A variety of studies have shown that there is a clinically meaningful
improvement in physical function and vitality scores, as measured by various patientreported
outcome instruments, as hemoglobin levels rise above 10 g/dL 74 (Figure 16).
Figure 16. Hemoglobin levels and associated Kidney Disease Questionnaire
(KDQ) scores for physical symptoms, fatigue, depression, relationship with
others, frustration, and overall KDQ 74 (Clinically meaningful change in KDQ is 0.2-
0.5 75 )
A recent meta-analysis showed that there was a modest difference in physical function
and vitality scores in the SF-36 when scores for patients treated to higher hemoglobin

Page 30
targets (approximately 12.0 – 15 g/dL across studies) were compared to those treated to
lower hemoglobin targets (approximately 9.5 – 11.5 g/dL across studies) 76 (Figure 17).
Figure 17. A meta-analysis of the difference in SF-36 physical function and vitality
scores between higher and lower hemoglobin levels in CKD-NOD patients; WMD =
weighted mean difference 76
2.4 Anemia and CV Outcomes in CKD-NOD
As previously discussed, a number of observational studies demonstrated an
association between anemia and an increased risk of adverse CV outcomes in CKD-
NOD patients 77, 78 . Based on these studies, RCTs were undertaken to test the
hypothesis that normalization or near normalization of hemoglobin levels in CKD-NOD
patients would result in decreased CV morbidity and mortality. The NHCT in dialysis
subjects with pre-existing CV disease or heart failure and anemia 44 and the CHOIR
study in CKD-NOD subjects with anemia 45 , addressed this question. Although NHCT
was conducted in dialysis patients, it is described here as part of a full discussion of
safety with respect to ESA use. The detailed results of these studies and the
subsequent revisions to the ESA labels were discussed in the previous section. In brief,
both studies evaluated higher-than-approved hemoglobin targets and found an excess of
events in the high target arm: (1) NHCT hazard ratio [HR] = 1.3, 95% confidence interval
[CI]: 0.9-1.8; (2) CHOIR HR = 1.34, 95%CI: 1.03 – 1.74). Based on the results of the
NHCT trial, the USPI for ESAs were revised in 1996 to highlight the observed risks and
to caution against normalization of hemoglobin levels in CKD. Following the publication

of the CHOIR results in 2006, the USPI for ESAs was updated in 2007 to include the
risks reported in the study including a Boxed Warning.
Page 31
The totality of the evidence informing the benefit and risks of ESA in CKD patients was
reviewed at the September 2007 joint meeting between the FDA CRDAC & Drug Safety
and Risk Management Advisory Committee, which included results from CHOIR.
Following the CRDAC meeting, the ESA labels were further revised to highlight the
following safety information:
• The Boxed Warning was updated
• The hemoglobin range of 10 to 12 g/dL was maintained
• Dosing guidance language was added on how to treat patients who did not
attain a hemoglobin level within the range of 10 to 12 g/dL despite the use of
appropriate Aranesp ® dose titrations over a 12-week period
2.5 Overview of ESA use in General CKD-NOD Clinical Practice
Nephrologists have responded to the findings in these well-publicized clinical trials by
treating to lower achieved hemoglobin levels; doses have also declined. The proportion
of ESA-treated CKD-NOD patients with hemoglobin levels > 12 g/dL has dropped
significantly (30% to 9%) and the proportion within the 10 to 12 g/dL hemoglobin range is
nearly 70% (Amgen data on file) (Figure 18). As of June 2009, the average hemoglobin
level among ESA-treated CKD-NOD patients was 10.7 g/dL 79 . Additionally, the average
monthly Aranesp ® dose in clinical practice in 2009 was 134 mcg.

Page 32
Figure 18. Percentage of hemoglobin measurements in specific ranges for the
CKD-NOD patient population (Amgen data on file; these data are shared with CMS
on an ongoing basis).
2.6 Trial to Reduce Cardiovascular Events with Aranesp ® Therapy
(TREAT) Trial in Diabetic CKD-NOD Patients
In 2009, the results of TREAT, a large Amgen-sponsored phase III prospective
randomized clinical trial, became available 61 . TREAT was designed in 2002, and
conducted between 2004 and 2009, to explore whether there were benefits with ESA
treatment beyond the supportive care indication for which ESAs were approved. The
rationale for this study was based on the observation that patients with higher achieved
hemoglobin levels experienced lower CV morbidity and mortality 77, 78 . The specific
objective of the study was to show that, when compared to placebo, treatment with
Aranesp ® could reduce time to first occurrence of death, CV morbidity or end-stage renal
disease; although it also assessed other endpoints such as quality of life. TREAT
enrolled diabetic, early-stage CKD-NOD patients (eGFR < 60 mL/min/1.73 m 2 ) with mild
anemia (hemoglobin < 11 g/dL). TREAT was a 4,038 patient, randomized, double-blind,
placebo-controlled study that examined treatment with ESA to a hemoglobin target of
13 g/dL compared to placebo (with Aranesp ® as rescue therapy when hemoglobin fell
below 9 g/dL). The median achieved hemoglobin in the Aranesp ® treated arm was 12.5
g/dL and the median achieved hemoglobin in the placebo arm was 10.6 g/dL.
TREAT did not demonstrate a reduction in the risk of the primary composite endpoint
(time to first event of death, myocardial infarction, hospitalization for congestive heart
failure, stroke and myocardial ischemia) (HR = 1.05, 95% CI: 0.94-1.17). Patients in the
Aranesp ® treated arm experienced an almost two-fold increased risk of stroke compared

Page 33
with the placebo group (HR = 1.92, 95% CI: 1.38-2.68 with an annualized incidence rate
of 1.1% in the placebo arm and 2.1% in the Aranesp ® treated arm). In December 2009,
consistent with Amgen’s previous response to safety signals identified in clinical trials,
Amgen proactively revised the ESA labels to include the risk of stroke observed in
TREAT, including enhancement of the Boxed Warning language and the Warnings and
Precautions section.
Amgen has undertaken a number of analyses to further understand the increased risk of
stroke (Amgen data on file). To date, analyses did not identify an association between
achieved hemoglobin or baseline characteristics (with the exception of a nominally
significant interaction with baseline aspirin use), and the risk of stroke. Within either the
Aranesp ® or placebo group, there was no difference in mean hemoglobin between
subjects who developed stroke and subjects who did not develop stroke. For subjects in
both treatment groups, systolic and diastolic blood pressure was higher in subjects who
developed stroke compared with subjects who did not develop stroke. Consistent with
the post-hoc analyses of other RCTs (NHCT, CHOIR) and observational studies, higher
achieved hemoglobin levels were associated with a lower risk of CV and mortality
events.
TREAT also assessed the association of treatment with ESAs with patient reported
outcomes in this mildly anemic CKD-NOD population. A modest improvement was seen
in the mean change in the FACT-Fatigue subscale score for subjects in the Aranesp ® -
treated group from baseline to week 25 (4.2 points for the Aranesp ® -treated group and
2.8 points for the placebo group), with a difference between the two groups of 1.33 (95%
CI: 0.64, 2.02; p < 0.0001). No statistically significant changes were seen between the
two groups in the SF-36 physical function and energy subscale scores.
While not an endpoint in the study, TREAT provides valuable information on the impact
of ESAs on transfusions in CKD-NOD. Subjects randomized to the Aranesp ® arm had
approximately half the rate of RBC transfusions compared to placebo-treated patients
(22.8 per 100 patient-years for placebo-treated versus 13.1 per 100 patient-years in the
Aranesp ® -treated arm; annualized rates). Additionally, analyses of the transfusion data
in TREAT demonstrated a continuous increase in the risk of transfusion as hemoglobin
levels decline (HR = 0.56, 95% CI 0.49-0.65) 61 . In general clinical practice, there is a
strong relationship between declining hemoglobin levels and increased transfusion risk
and this risk is particularly elevated when hemoglobin levels drop below 10 g/dL 65, 66 .

Page 34
TREAT was not a study designed to assess the optimal hemoglobin target range for
treatment in CKD-NOD and does not provide evidence to support the use of ESAs
beyond transfusion reduction in the population studied. During the treatment phase, the
median hemoglobin in the arm randomized to receive Aranesp ® was 12.5 g/dL and the
median hemoglobin in the arm randomized to receive placebo with rescue therapy (at
9 g/dL) was 10.6 g/dL. TREAT only enrolled patients with hemoglobin < 11 g/dL; the
median baseline hemoglobin in TREAT was 10.4 g/dL. In current clinical practice, most
of the patients enrolled in TREAT would not receive ESA therapy; CKD-NOD patients
are currently initiated on therapy when hemoglobin declines to below 10 g/dL (Amgen,
data on file). Furthermore, the mean hemoglobin in anemic CKD-NOD patients who are
currently receiving ESA therapy in the US is approximately 10.7 g/dL 79 , which is similar
to the placebo arm of TREAT and considerably less than the target hemoglobin of
13 g/dL.
2.7 CONCLUSION: ESA THERAPY IN CKD-NOD PATIENTS IS AN
IMPORTANT TREATMENT OPTION IN THOSE FOR WHOM
TRANSFUSION AVOIDANCE IS A MEANINGFUL CLINICAL BENEFIT
Although the prevalence of anemia is lower in CKD-NOD, anemia can be severe and
transfusions are frequent. Anemic CKD-NOD patients are vulnerable to the risks of
transfusions, particularly the risk of allo-sensitization, and its potential impact on
transplant eligibility and graft survival. ESA therapy should be initiated when hemoglobin
declines below 10 g/dL because the demonstrated benefit of transfusion reduction with
ESA use is observed when hemoglobin levels are treated to above 10 g/dL. ESA
therapy should be individualized to achieve and maintain hemoglobin between 10 and
12 g/dL. As with dialysis patients, the intrinsic hemoglobin variability seen in CKD-NOD
and the practical limitations of managing hemoglobin, require the use of a hemoglobin
range to administer ESAs to maximize the clinical benefit for each patient. A
hemoglobin level of 12.0 g/dL as the upper end of the hemoglobin range will allow
physicians to manage patients so as to reduce transfusions. Avoiding a hemoglobin
target > 12 g/dL provides a safety margin against the higher hemoglobin target
(≥ 13.0 g/dL) in TREAT where risks have been identified. Evidence supports that ESAs
are an important therapy for anemic CKD-NOD patients in whom transfusion avoidance
is a meaningful clinical goal and believes that ESAs have a positive benefit to risk profile
when used according to the FDA-approved label in these patients.

REFERENCES
Page 35
1. Go AS, Chertow GM, Fan D, McCulloch CE and Hsu CY: Chronic kidney disease
and the risks of death, cardiovascular events, and hospitalization. The New England
journal of medicine. 2004;351: 1296-1305.
2. USRDS: USRDS 2009 Annual data report: atlas of end-stage renal disease in the
United States. Bethesda, MD, National Institutes of Health, National Institute of
Diabetes and Digestive Kidney Diseases, 2009
3. Birmele B, Francois M, Pengloan J, et al.: Death after withdrawal from dialysis: the
most common cause of death in a French dialysis population. Nephrol Dial
Transplant. 2004;19: 686-691.
4. Karpinski M, Pochinco D, Dembinski I, Laidlaw W, Zacharias J and Nickerson P:
Leukocyte reduction of red blood cell transfusions does not decrease
allosensitization rates in potential kidney transplant candidates. J Am Soc Nephrol.
2004;15: 818-824.
5. Sargent JA and Acchiardo SR: Iron requirements in hemodialysis. Blood purification.
2004;22: 112-123.
6. Caro J, Brown S, Miller O, Murray T and Erslev AJ: Erythropoietin levels in uremic
nephric and anephric patients. The Journal of laboratory and clinical medicine.
1979;93: 449-458.
7. Erslev AJ and Besarab A: Erythropoietin in the pathogenesis and treatment of the
anemia of chronic renal failure. Kidney international. 1997;51: 622-630.
8. Cheung J, Yu A, LaBossiere J, Zhu Q and Fedorak RN: Peptic ulcer bleeding
outcomes adversely affected by end-stage renal disease. Gastrointestinal
endoscopy. 2010;71: 44-49.
9. Wasse H, Gillen DL, Ball AM, et al.: Risk factors for upper gastrointestinal bleeding
among end-stage renal disease patients. Kidney international. 2003;64: 1455-1461.
10. Eschbach JW and Adamson JW: Anemia of end-stage renal disease (ESRD). Kidney
international. 1985;28: 1-5.
11. Churchill DN, Taylor DW, Cook RJ, et al.: Canadian Hemodialysis Morbidity Study.
Am J Kidney Dis. 1992;19: 214-234.
12. Eschbach JW, Egrie JC, Downing MR, Browne JK and Adamson JW: Correction of
the anemia of end-stage renal disease with recombinant human erythropoietin.
Results of a combined phase I and II clinical trial. The New England journal of
medicine. 1987;316: 73-78.
13. Beauregard P and Blajchman MA: Hemolytic and pseudo-hemolytic transfusion
reactions: an overview of the hemolytic transfusion reactions and the clinical
conditions that mimic them. Transfusion medicine reviews. 1994;8: 184-199.
14. Despotis GJ, Zhang L and Lublin DM: Transfusion risks and transfusion-related proinflammatory
responses. Hematol Oncol Clin North Am. 2007;21: 147-161.
15. Dodd RY, Notari EPt and Stramer SL: Current prevalence and incidence of infectious
disease markers and estimated window-period risk in the American Red Cross blood
donor population. Transfusion. 2002;42: 975-979.
16. Simon GE and Bove JR: The potassium load from blood transfusion. Postgraduate
medicine. 1971;49: 61-64.
17. Hardy S, Lee S-H and Terasaki PI: Sensitization 2001, in PI, CJaT (ed): Clinical
Transplants 2001, UCLA, 2001
18. USRDS: USRDS 2004 Annual data report: atlas of end-stage renal disease in the
United States. Bethesda, MD, National Institutes of Health, National Institute of
Diabetes and Digestive Kidney Diseases, 2004

Page 36
19. UNOS: United Network for Organ Sharing. Accessed at: 4http://www.unos.org/ .
Accessed on Feb 12, 2010.,
20. Opelz G: Non-HLA transplantation immunity revealed by lymphocytotoxic antibodies.
Lancet. 2005;365: 1570-1576.
21. Schold J, Srinivas TR, Sehgal AR and Meier-Kriesche HU: Half of kidney transplant
candidates who are older than 60 years now placed on the waiting list will die before
receiving a deceased-donor transplant. Clin J Am Soc Nephrol. 2009;4: 1239-1245.
22. CESG: Association between recombinant human erythropoietin and quality of life
and exercise capacity of patients receiving haemodialysis. Canadian Erythropoietin
Study Group. BMJ. 1990;300: 573-578.
23. Johansen KL, Finkelstein FO, Revicki DA, Gitlin M, Evans C and Mayne TJ:
Systematic Review and Meta-analysis of Exercise Tolerance and Physical
Functioning in Dialysis Patients Treated With Erythropoiesis-Stimulating Agents. Am
J Kidney Dis. 2010;In press.
24. Pisoni RL, Bragg-Gresham JL, Young EW, et al.: Anemia management and
outcomes from 12 countries in the Dialysis Outcomes and Practice Patterns Study
(DOPPS). Am J Kidney Dis. 2004;44: 94-111.
25. USRDS: USRDS 2007 Annual data report: atlas of end-stage renal disease in the
United States. Bethesda, MD, National Institutes of Health, National Institute of
Diabetes and Digestive Kidney Diseases, 2007
26. Ibrahim HN, Ishani A, Foley RN, Guo H, Liu J and Collins AJ: Temporal trends in red
blood transfusion among US dialysis patients, 1992-2005. Am J Kidney Dis. 2008;52:
1115-1121. .
27. USRDS: USRDS 2003 Annual data report: atlas of end-stage renal disease in the
United States. Bethesda, MD, National Institutes of Health, National Institute of
Diabetes and Digestive Kidney Diseases, 2003
28. Amgen: Background information for the joint meeting between the Cardiovascular
and Renal Drugs Advisory Committee & Drug Safety and Risk Management Advisory
Committee. Accessed at: 4http://www.fda.gov/ohrms/dockets/ac/07/briefing/2007-
4315b1-04-AMGEN.pdf, 2007
29. Ebben JP, Gilbertson DT, Foley RN and Collins AJ: Hemoglobin level variability:
associations with comorbidity, intercurrent events, and hospitalizations. Clin J Am
Soc Nephrol. 2006;1: 1205-1210. .
30. Fishbane S and Berns JS: Hemoglobin cycling in hemodialysis patients treated with
recombinant human erythropoietin. Kidney Int. 2005;68: 1337-1343.
31. Lacson E, Jr., Ofsthun N and Lazarus JM: Effect of variability in anemia
management on hemoglobin outcomes in ESRD. Am J Kidney Dis. 2003;41: 111-
124.
32. Spiegel DM, Khan I, Krishnan M and Mayne TJ: Changes in hemoglobin level
distribution in US dialysis patients from June 2006 to November 2008. Am J Kidney
Dis. 2009;55: 113-120.
33. Collins AJ, Brenner RM, Ofman JJ, et al.: Epoetin alfa use in patients with ESRD: an
analysis of recent US prescribing patterns and hemoglobin outcomes. Am J Kidney
Dis. 2005;46: 481-488.
34. Medicare: Analysis conducted by the Chronic Disease Research Group of the
Minneapolis Medical Research Foundation under DUA #16234 and #16235 with the
Centers For Medicare and Medicaid Services,
35. Medicare: Dialysis facility compare. Accessed at:
4http://www.medicare.gov/Dialysis/Include/DataSection/Questions/SearchCriteria.asp
?version=default&browser=IE%7C6%7CWin2000&language=English&defaultstatus=
0&pagelist=Home,

Page 37
36. Uehlinger DE, Gotch FA and Sheiner LB: A pharmacodynamic model of
erythropoietin therapy for uremic anemia. Clinical pharmacology and therapeutics.
1992;51: 76-89.
37. Feng A and Comstock T: Within-subject hemoglobin variability in placebo-controlled
Epoetin alfa studies of patients on hemodialysis with anemia. J Am Soc Nephrol
[abstract]. 2009;20: 168A.
38. CMS: CMS Manuap System. Pub 100-04 Medicare Claims Processing. Transmittal
751, Change Request 4135, DHHS CMS 11-10-2005. National monitoring policy for
EPO and Aranesp for end stage renal disease (ESRD) patients treated in renal
dialysis facilities 2005
39. Cannella G, La Canna G, Sandrini M, et al.: Reversal of left ventricular hypertrophy
following recombinant human erythropoietin treatment of anaemic dialysed uraemic
patients. Nephrol Dial Transplant. 1991;6: 31-37.
40. Wizemann V, Kaufmann J and Kramer W: Effect of erythropoietin on ischemia
tolerance in anemic hemodialysis patients with confirmed coronary artery disease.
Nephron. 1992;62: 161-165.
41. Pascual J, Teruel JL, Moya JL, Liano F, Jimenez-Mena M and Ortuno J: Regression
of left ventricular hypertrophy after partial correction of anemia with erythropoietin in
patients on hemodialysis: a prospective study. Clinical nephrology. 1991;35: 280-
287.
42. Collins AJ, Li S, St Peter W, et al.: Death, hospitalization, and economic associations
among incident hemodialysis patients with hematocrit values of 36 to 39%. J Am Soc
Nephrol. 2001;12: 2465-2473.
43. Volkova N and Arab L: Evidence-based systematic literature review of
hemoglobin/hematocrit and all-cause mortality in dialysis patients. Am J Kidney Dis.
2006;47: 24-36.
44. Besarab A, Bolton WK, Browne JK, et al.: The effects of normal as compared with
low hematocrit values in patients with cardiac disease who are receiving
hemodialysis and epoetin. N Engl J Med. 1998;339: 584-590.
45. Singh AK, Szczech L, Tang KL, et al.: Correction of anemia with epoetin alfa in
chronic kidney disease. N Engl J Med. 2006;355: 2085-2098.
46. Amgen: EPOGEN (R) (Epoetin alfa) Prescribing Information. Amgen Inc., Thousand
Oaks, CA.
47. Cotter DJ, Stefanik K, Zhang Y, Thamer M, Scharfstein D and Kaufman J:
Hematocrit was not validated as a surrogate end point for survival among epoetintreated
hemodialysis patients. Journal of clinical epidemiology. 2004;57: 1086-1095.
48. Zhang Y, Thamer M, Stefanik K, Kaufman J and Cotter DJ: Epoetin requirements
predict mortality in hemodialysis patients. Am J Kidney Dis. 2004;44: 866-876.
49. Bradbury BD, Critchlow CW, Weir MR, Stewart R, Krishnan M and Hakim RH:
Impact of elevated C-reactive protein levels on erythropoiesis- stimulating agent
(ESA) dose and responsiveness in hemodialysis patients. Nephrol Dial Transplant.
2009;24: 919-925.
50. Bradbury BD, Brookhart MA, Winkelmayer WC, et al.: Evolving statistical methods to
facilitate evaluation of the causal association between erythropoiesis-stimulating
agent dose and mortality in nonexperimental research: strengths and limitations. Am
J Kidney Dis. 2009;54: 554-560.
51. van den Berghe G, Wouters P, Weekers F, et al.: Intensive insulin therapy in the
critically ill patients. The New England journal of medicine. 2001;345: 1359-1367.
52. Finney SJ, Zekveld C, Elia A and Evans TW: Glucose control and mortality in
critically ill patients. JAMA. 2003;290: 2041-2047.

Page 38
53. Bradbury BD, Danese MD, Gleeson M and Critchlow CW: Effect of epoetin alfa dose
changes on hemoglobin and mortality in hemodialysis patients with hemoglobin
levels persistently below 11 g/dL. Clin J Am Soc Nephrol. 2009;4: 630-637.
54. Kilpatrick RD, Critchlow CW, Fishbane S, et al.: Greater epoetin alfa responsiveness
is associated with improved survival in hemodialysis patients. Clin J Am Soc
Nephrol. 2008;3: 1077-1083.
55. Bradbury BD, Wang O, Critchlow CW, et al.: Exploring relative mortality and epoetin
alfa dose among hemodialysis patients. Am J Kidney Dis. 2008;51: 62-70.
56. Kalantar-Zadeh K and Aronoff GR: Hemoglobin variability in anemia of chronic
kidney disease. J Am Soc Nephrol. 2009;20: 479-487.
57. Solid CA, Foley RN, Gilbertson DT and Collins AJ: Perihospitalization hemoglobinepoetin
associations in U.S. hemodialysis patients, 1998 to 2003. Hemodial Int.
2007;11: 442-447.
58. Bradbury BD, Do TP, Winkelmayer WC, Critchlow CW and Brookhart MA: Greater
Epoetin alfa (EPO) doses and short-term mortality risk among hemodialysis patients
with hemoglobin levels less than 11 g/dL. Pharmacoepidemiol Drug Saf. 2009;18:
932-940.
59. Zhang Y, Thamer M, Cotter D, Kaufman J and Hernan MA: Estimated effect of
epoetin dosage on survival among elderly hemodialysis patients in the United States.
Clin J Am Soc Nephrol. 2009;4: 638-644.
60. Wang O, Kilpatrick RD, Critchlow CW, et al.: Relationship between Epoetin alfa dose
and mortality: findings from a marginal structural model. Clin J Am Soc Nephrol.
2010;In press.
61. Pfeffer MA, Burdmann EA, Chen CY, et al.: A trial of darbepoetin alfa in type 2
diabetes and chronic kidney disease. The New England journal of medicine.
2009;361: 2019-2032.
62. Coresh J, Selvin E, Stevens LA, et al.: Prevalence of chronic kidney disease in the
United States. JAMA. 2007;298: 2038-2047.
63. Astor BC, Muntner P, Levin A, Eustace JA and Coresh J: Association of kidney
function with anemia: the Third National Health and Nutrition Examination Survey
(1988-1994). Arch Intern Med. 2002;162: 1401-1408.
64. Ibrahim HN, Ishani A, Guo H and Gilbertson DT: Blood transfusion use in nondialysis-dependent
chronic kidney disease patients aged 65 years and older. Nephrol
Dial Transplant. 2009;24: 3138-3143.
65. Lawler EV, Gagnon DR, Fink J, et al.: Initiation of anaemia management in patients
with chronic kidney disease not on dialysis in the Veterans Health Administration.
Nephrol Dial Transplant. 2010;In press.
66. Lawler EV, Bradbury BD, Fonda JR, Gaziano JM and Gagnon DR: Transfusion
burden among patients with chronic kidney disease and anemia. Clin J Am Soc
Nephrol. 2010;In press.
67. Amgen: Aranesp (R) (Epoetin alfa) Prescribing Information. Amgen Inc., Thousand
Oaks, CA.
68. Eschbach JW, Abdulhadi MH, Browne JK, et al.: Recombinant human erythropoietin
in anemic patients with end-stage renal disease. Results of a phase III multicenter
clinical trial. Annals of internal medicine. 1989;111: 992-1000.
69. Provenzano R, Garcia-Mayol L, Suchinda P, et al.: Once-weekly epoetin alfa for
treating the anemia of chronic kidney disease. Clin Nephrol. 2004;61: 392-405.
70. Revicki DA, Brown RE, Feeny DH, et al.: Health-related quality of life associated with
recombinant human erythropoietin therapy for predialysis chronic renal disease
patients. Am J Kidney Dis. 1995;25: 548-554.

Page 39
71. Benz R, Schmidt R, Kelly K and Wolfson M: Epoetin alfa once every 2 weeks is
effective for initiation of treatment of anemia of chronic kidney disease. Clin J Am
Soc Nephrol. 2007;2: 215-221. .
72. Kleinman KS, Schweitzer SU, Perdue ST, Bleifer KH and Abels RI: The use of
recombinant human erythropoietin in the correction of anemia in predialysis patients
and its effect on renal function: a double-blind, placebo-controlled trial. Am J Kidney
Dis. 1989;14: 486-495.
73. Alexander M, Kewalramani R, Agodoa I and Globe D: Association of anemia
correction with health related quality of life in patients not on dialysis. Curr Med Res
Opin. 2007;23: 2997-3008.
74. Lefebvre P, Vekeman F, Sarokhan B, Enny C, Provenzano R and Cremieux PY:
Relationship between hemoglobin level and quality of life in anemic patients with
chronic kidney disease receiving epoetin alfa. Curr Med Res Opin. 2006;22: 1929-
1937.
75. Jaeschke R, Singer J and Guyatt GH: Measurement of health status. Ascertaining
the minimal clinically important difference. Controlled clinical trials. 1989;10: 407-
415.
76. Clement FM, Klarenbach S, Tonelli M, Johnson JA and Manns BJ: The impact of
selecting a high hemoglobin target level on health-related quality of life for patients
with chronic kidney disease: a systematic review and meta-analysis. Archives of
internal medicine. 2009;169: 1104-1112.
77. Al-Ahmad A, Rand WM, Manjunath G, et al.: Reduced kidney function and anemia
as risk factors for mortality in patients with left ventricular dysfunction. Journal of the
American College of Cardiology. 2001;38: 955-962.
78. Sarnak MJ, Tighiouart H, Manjunath G, et al.: Anemia as a risk factor for
cardiovascular disease in The Atherosclerosis Risk in Communities (ARIC) study.
Journal of the American College of Cardiology. 2002;40: 27-33.
79. Regidor D, McClellan WM, Kewalramani R, Sharma A and Bradbury BD: Changes in
erythropoiesis stimulating agent (ESA) dosing and hemoglobin levels in U.S. nondialysis
chronic kidney disease patients between 2005 and 2009. submitted (under
review). 2010.

Technical Appendix 1 – Evidence Tables Page 40
Technical Appendix 1 – Evidence Tables
Table 1a: Transfusion Burden in CKD Patients
STUDY OBJECTIVE STUDY POPULATION TRANSFUSION OUTCOME
Retrospective observational
Gilbertson
(2009) 1
Describe the influence of anemia
management interventions (ESA, Iron,
HD patients receiving ESAs
Hb variability groups:
• Increased time below 11 g/dL is associated with increased
risk of transfusions
and transfusions) on Hb variability • Patients with Hb < 11 for six • Transfusion rate was 21.6 per 100 for patients with Hb
months (L-L), N=2,165
fluctuating between < 11 g/dL and 11-12.5 g/dL over 6
• Patients with Hb 11-12.5 for six months
consecutive months (I-I),
• Transfusion rate was 121.7 per 100 for patients with Hb
N=9,646
persistently below 11 g/dL for six consecutive months
• Patients with Hb > 12.5 for six
consecutive months (H-H),
N=3,750
• Patients with Hb fluctuating
between < 11 and 11-12.5 over
six months (L-I), N=29,222
• Patients with Hb > 11 for six
consecutive months (I-H),
N=50,680
• Patients with Hb levels across
Hb distribution (L-H), N=64,257
Ibrahim (2008) 2 Retrospective cohort study to address
patterns and trends in red blood cell
(RBC) transfusions in US hemodialysis
patients (1992-2005)
Ibrahim (2009) 3 Retrospective cohort study to address
patterns and trends in red blood cell
(RBC) transfusions in the CKD and non-
CKD population in Medicare between
1992 to 2004
All Medicare HD patients between
1992 and 2005
N=77,347 (1992)
N=164,933 (2005)
> 90% of hemodialysis received ESA
treatment during this period
All Medicare beneficiaries with CKD
between 1992 and 2004
Total CKD N=301,000
N=14,057 (1992)
N=42,225 (2004)
Total CKD with anemia N=118,895
N=4371 (1992)
• Mean population hemoglobin increased from 9.7 g/dL in 1992
to 11.8 g/dL in 2004
• The total transfusion rate declined from 535.33 per 1000
patient-years in 1992 to 263.65 per 1000 patient-years in
2005
• For transfusions occurring in the outpatient setting only,
transfusions declined from 225.63 per 1000 patient-years in
1992 to 66.66 per 1000 patient-years in 2005
• The proportion of CKD patients receiving ESA therapy
increased from 0.8% in 1992 to 7.5% in 2004
• The total transfusion among anemic CKD patients not
receiving ESA therapy was 249.2 per 1000 patient-years in
1992 and 174.7 per 1000 patient-years in 2004
• The total transfusion among anemic CKD patients who
received ESA therapy was 572.9 per 1000 patient-years in

Technical Appendix 1 – Evidence Tables Page 41
STUDY OBJECTIVE STUDY POPULATION TRANSFUSION OUTCOME
N=18,579 1992 and 200.1 per 1000 patient-years in 2004
Lawler (2010) 4 Estimate the burden of transfusion by
anemia treatment and Hb level in CKD-
NOD patients
Lawler (2010) 5
Examine initiation of anemia management
among anemic CKD-NOD patients
CKD-NOD (CKD ≥ Stage 3)
Hb< 11 g/dL
No anemia treatment, N=35,364
Iron treatment only, N=35,420
ESA treatment only, N=16,439
ESA + Iron treatment, N=10,413
CKD-NOD patients (eGFR 15-

Technical Appendix 1 – Evidence Tables Page 42
Table 1b: Cost of Red Blood Cell Transfusion
STUDY OBJECTIVE STUDY POPULATION* OUTCOME
Prospective Observational
Agrawal et al, (2006) 1
Assess direct costs of blood Hospital
Average blood transfusion of 2 units to £546.12. (2004)
(UK)
transfusion through a time
and motion study
Blood cost, staff time, blood-bank cost,
disposables, overnight stay, transfusion
related complications
Glenngard (2005) 12
Assess direct costs of blood Societal
Swedish kronor (SEK) 6330 for filtered allogeneic RBCs and
(Sweden)
transfusion through interviews
with hospital staff and
published data
Acquisition, Typing, Testing, Preparation,
Compatibility testing, Administration at the
ward, Materials at the ward, Cell saver costs.
SEK 5394 for autologous
RBCs for surgery patients per 2-unit transfusion (2003)
Ueno et al (2006) 23
(US)
Assess direct and indirect
costs of blood transfusion
through interviews,
prospective time and motion
and retrospective chart review
Minuk et al (2008) 3 Assess indirect costs of blood
transfusion through a
prospective, observational
study
Retrospective Observational
Amin (2004) 4
(Canada)
Assess direct costs of blood
transfusion using a coststructure
analysis,
observational
Hospital
RBC acquisition/preparation, testing, and
administration costs
N=20
Patients receiving RBC transfusion in a
tertiary-care cancer center.
N=1,279
Societal
Cost of blood collection, production, and
distribution
Cost of hospital transfusion service
processing RBC transfusion. Transfusion
reaction management
/RBC unit
Direct Costs: One unit costs: US$582/RBC unit
Two unit costs: US$1,139/2-unit transfusion
(2005)
Indirect Costs: The mean (SD) time duration was 98.5 (12.7)
and 92.5 (21.2) minutes for the first and second units, of blood
transfusion, respectively.
Transfusions are also related to increased number of physician
visits
A patient receiving 1 RBC unit spent a total of 109 + 19
minutes (mean + SD) , independent of the time consumed by
activities pre- and post- transfusion.
US$264.81
(95% CI, $256.29-
$275.65)
(2002)

Technical Appendix 1 – Evidence Tables Page 43
STUDY OBJECTIVE STUDY POPULATION* OUTCOME
Cremieux (2000) 5
Assess direct costs of blood Provider
Total: US$491.00/RBC unit (1998)
(US)
transfusion using
retrospective chart review,
observational
Total direct costs
Norum (2008) 6
(Norway)
Assess direct costs of blood
transfusion using
retrospective, observational
Societal
Health care, infections, patient, family, other
sectors
€1069
(2005)
These evidence tables include the most commonly cited references; a systematic review was not conducted.
1. Agrawal S, Davidson N, Walker Mc, Gibson S, et al. Assessing the total costs of blood delivery to hospital oncology and haematology patients. Current Medical
Research and Opinion. 2006. Vol. 22(10):1903–1909.
2. Glenngard AH, Persson U and Soderman C: Costs associated with blood transfusions in Sweden--the societal cost of autologous, allogeneic and perioperative RBC
transfusion. Transfus Med. 2005;15: 295-306.
3. Ueno W, Beveridge RA and Kales AN: Costs of outpatient red blood cell transfusions. J Support Oncol. 2006;4: 494-495.
4. Minuk L, Chin-Yee I, Hibbert A, Chan D and Xenocostasn A: Red blood cell transfusion and chemotherapy administration: a study of resource utilization. Community
Oncology;5: 598-603.
5. Cremieux PY, Barrett B, Anderson K and Slavin MB: Cost of outpatient blood transfusion in cancer patients. J Clin Oncol. 2000;18: 2755-2761.
6. Norum J and Moen MA: Practice and costs of red blood cell (RBC) transfusion in an oncological unit. Anticancer Res. 2008;28: 459-464.

Technical Appendix 1 – Evidence Tables Page 44
Table 2. Impact of Transfusion on Allo-Sensitization (Panel Reactive Antibodies [PRA]) and Transplantation
STUDY OBJECTIVE STUDY POPULATION STUDY RESULTS
Retrospective observational
Bucin (1988) 1 To determine the effect of pre-transplant
transfusions on long-term transplant
survival
Buscaroli
(1992) 2
Cardarelli
(2005) 3
To determine the association of
intermediate PRA levels (30-60%) on
transplant outcomes
Examine the prevalence and significance
of anti-HLA and donor-specific antibody
(DSA) long term after renal transplantation
Cecka (1988) 4 To determine effect of peak PRA on time
to transplant
Deierhoi (1989) 5 Determine incidence of renal transplant by
PRA pattern
Primary renal transplants, treated
with azathioprine or prednisolone
Subgroups:
pre-transplant transfusions, N=79
no pre-transplant transfusions, N=37
241 renal transplant patients with
peak PRA 12 months prior to
transplant
Low PRA 60%
1) Persistently high: sustaining high
PRA (> 60%), N=46
2) Persistently low: 1 peak and
decline ≤ 30%, N=22
3) Variable PRA: ≥ 2 PRA peaks and
• Decreased 2-year graft survival in patients with pre-transplant
transfusions.
• 2-year cumulative graft survival:
• Transfused: 81%
• No transfusion: 97% (p < 0.05)
• Better graft survival 1 year post transplant for patients in the
lower PRA group compared to the higher PRA group (90%
vs. 79%, p

Technical Appendix 1 – Evidence Tables Page 45
STUDY OBJECTIVE STUDY POPULATION STUDY RESULTS
decline ≤ 30%), N=24
Study follow-up period was not
Hardy (2001) 6 Describe graft survival by transfusion
history and PRA level
Katznelson
(1997) 7
Describe association between highest
pre-transplant PRA and transplant
survival
Lietz (2003) 8 Describe association between highest
pre-transplant PRA and transplant
survival
Nicol (1993) 9 Determine effect of PRA, transfusions and
other donor/recipient factors on transplant
survival
Opelz (2005) 10 Assess the association between the
presence of lymphocytotoxic antibodies
before transplantation and the outcome of
kidney grafts for a) HLA identical sibling
and b) cadaveric kidney transplants
Soosay (2003) 11 Characterize the impact of transfusion on
sensitization and time to transplant
provided
Transplanted UNOS registrants
PRA ≤10%, N=27,787
PRA>10%, N=7,447
Cadaveric transplants in UNOS
registrants
Patients with PRA ≤20%, N=20,924
Patients with PRA >20%, N=5,570
Transplanted kidney survival ≥ 6
months
Groups:
ESA, N=235
Transfusion, N=207
No treatment, N=60
Renal transplant patients
Graft failure attributed to rejection,
N=48
Control (functioning graft at 3 years),
N=300
PRA levels from the last pretransplant
serum level
HLA Living relative Subgroup:
PRA 0 N=3001
PRA 1-50 N=803
PRA > 50 N=244
PRA levels from the last pretransplant
serum level
Cadaveric Kidney Transplant
Subgroup:
No PRA N=116562
PRA 1-50 N=36314
PRA > 50 N=7610
Patients on renal transplant wait list
stratified by PRA
N=244
PRA categorized as:
PRA 0−9%
• Lower 3-year kidney transplant survival as transfusion
burden increased
• At two years graft survival among patients with
PRA>20% was 76%
• At two years graft survival among patients with PRA

Technical Appendix 1 – Evidence Tables Page 46
STUDY OBJECTIVE STUDY POPULATION STUDY RESULTS
UNOS 12 Transplant registry
Scientific Registry of Transplant
Recipients (SRTR), annual report for
transplant patients.
USRDS
(2009) 13
Surveillance data assessing treatment
patterns among ESRD patients
Vella (1998) 14 Determine transfusion burden, PRA
levels, and reason for sensitization in
patients on the renal transplant waitlist in
the pre compared to post ESA era (1989
compared 1994)
Zhou (1991) 15 Describe effects of sensitization on
waiting time and graft survival
PRA 10−59%
PRA 60−79%
PRA 80−100%
CKD patients receiving a living donor
kidney transplant by PRA level.
PRA levels by graft survival 3
months, 1 year, 5 years, and 10
years.
ESRD patients
Patients 18 and older listed for a firsttime
kidney-only transplant in the
given year.
Patients on renal transplant waitlist
Year 1989, patients N=205 (pre ESA)
Year 1994 patients N=126 (post ESA)
Nonsensitized=PRA 50%, N=2,615
These evidence tables include the most commonly cited references; a systematic review was not conducted.
• Higher PRA levels increase time to transplant
• Patients with PRA 0 – 9% waited 7 months on average
• Patients with PRA 80 – 100% waited >35 months on
average
• Despite graft survival being highest in recipients of living
donor kidneys, 2006 survival rates at 3 months, 1 year, 5
years, and 10 years show marked reduction in survival as
PRA level increases
• PRA = 0-9% graft survival was 97.9% at 3 mos, 96% at 1
yr, 81% at 5 years and 57.5% at 10 years
• PRA = 10-79% graft survival was 97.4% at 3 mos, 95.2%
at 1 yr, 76.3% at 5 years and 51.6% at 10 years
• PRA > 80% graft survival was 93.8% at 3 mos, 90.9% at
1 yr, 67.4% at 5 years and 45.4% at 10 years
• Wait time for transplant continues to increase nationwide
• Higher sensitized patients (PRA ≥ 10%) wait twice as
long (>4 yrs vs. >2 yrs) as patients with PRA < 10%
• Patients on the transplant waitlist longer have a higher
likelihood of death
• PRA levels were lower the in 1994 cohort vs. 1989 cohort in
each of the three different PRA categories (p=0.008)
• 49.8% of 1989 cohort had PRA = 0-9% compared to
63.5% in the 1994 cohort
• 25.9% of the 1989 cohort had PRA = 10-79% compared
to 12.7% in the 1994 cohort
• 24.4% of the 1989 cohort had PRA > 80% compared to
23.8% of the 1994 cohort
• At one year graft survival among patients with PRA 0-10%
was 79%
• At one year graft survival among patients with PRA 10-50%
was 78%
• At one year graft survival among patients with PRA > 50%
was 72%

Technical Appendix 1 – Evidence Tables Page 47
1. Besarab A, Bolton WK, Browne JK, et al.: The effects of normal as compared with low hematocrit values in patients with cardiac disease who are receiving
hemodialysis and epoetin. N Engl J Med. 1998;339: 584-590.
2. Pfeffer MA, Burdmann EA, Chen CY, et al.: A trial of darbepoetin alfa in type 2 diabetes and chronic kidney disease. The New England journal of medicine.
2009;361: 2019-2032.
3. Provenzano R, Garcia-Mayol L, Suchinda P, et al.: Once-weekly epoetin alfa for treating the anemia of chronic kidney disease. Clin Nephrol. 2004;61: 392-405.
4. Boelaert JR, Daneels RF, Schurgers ML, Matthys EG, Gordts BZ and Van Landuyt HW: Iron overload in haemodialysis patients increases the risk of bacteraemia:
a prospective study. Nephrol Dial Transplant. 1990;5: 130-134.
5. Churchill DN, Taylor DW, Cook RJ, et al.: Canadian Hemodialysis Morbidity Study. Am J Kidney Dis. 1992;19: 214-234.
6. Ibrahim HN, Ishani A, Foley RN, Guo H, Liu J and Collins AJ: Temporal trends in red blood transfusion among US dialysis patients, 1992-2005. Am J Kidney Dis.
2008;52: 1115-1121. .
7. Ibrahim HN, Ishani A, Guo H and Gilbertson DT: Blood transfusion use in non-dialysis-dependent chronic kidney disease patients aged 65 years and older.
Nephrol Dial Transplant. 2009;24: 3138-3143.
8. Seliger: Timing of ESA initiation and adverse outcomes in non-dialysis dependent CKD: a propensity matched observational cohort study. Clin J Am Soc Nephrol.
2010;In press.

Technical Appendix 1 – Evidence Tables Page 48
Table 3: Transfusion Reduction with ESA Treatment
STUDY
Randomized trial
OBJECTIVE STUDY POPULATION OUTCOMES
Amgen study Double-blind, partial cross over
8701 (data on
file)
Correction of anemia with intravenous
EPOGEN ® HD patients, Hb ≤ 10 g/dL
in HD patients and estimate
Target Hb 10.7-12.7 g/dL
N=68
reduction in transfusions.
Treatment duration 24 weeks
Correction phase, 12 weeks
Placebo, N=32
EPOGEN
Maintenance phase: 12 weeks
® • For patients randomized to receive EPOGEN
treated, N=36
® , transfusions
were reduced from 58% at baseline to 17% at 12 weeks and
to 0% at 24 weeks.
• 72% of patients randomized to placebo received transfusion
at baseline and this persisted at 12 weeks (63%). At this
point, patients were switched to EPOGEN ® and by 24 weeks
only 17% were receiving transfusion.
Amgen 8904
(data on file)
Double-blind, partial cross over
Estimate correction of anemia with
subcutaneous EPOGEN ® and reduction in
transfusions, PRO
Target Hb 10.7-12.7
Treatment duration 24 weeks
Correction phase, 12 weeks
Maintenance phase: 12 weeks
Besarab (1998) 1 Open label
To examine the risks and benefits of
normalizing the hematocrit in patients with
cardiac disease who were undergoing
hemodialysis.
Pfeffer (2009) 2
and Amgen data
on file
Single Arm Trial
Provenzano
(2004) 3
Randomized, double-blind, placebocontrolled
In patients with Type 2 Diabetes and
Chronic Kidney Disease that does not
require dialysis and concomitant anemia,
increasing Hb levels with the use of
Darbepoetin alfa to lower the rates of
death, cardiovascular events and endstage
renal disease
Prospective, multi-center, open-label
study
To study the safety and efficacy with
changes in health-related quality of life.
PD patients, Hb ≤ 10 g/dL
Target Hb 10 10.7-12.7 g/dL
N=152
Placebo, N=74
EPOGEN® treated, N=78
HD Patients with congestive heart failure or
ischemic heart disease; serum transferring
saturation of ≥ 20%. Patients treated to:
High target Hb arm (14 g/dL), N=618
Low target Hb arm (10 g/dL), N=615
Median treatment duration: 14 months
Diabetic CKD-NOD patients (20 to
60 mL/min/1.73 m 2 ) with anemia
(Hb ≤ 11 g/dL).
Treated arm, N=2012
Placebo arm, N=2025
Anemic CKD-NOD patients
N=1557 (enrolled)
N=1338 (efficacy)
16-week study
• Patients receiving SC EPOGEN ® were able to achieve target
Hb with almost complete elimination of need of transfusion by
12 weeks.
• Over the course of the study, 21% of the patients in the high
Hb target arm received a transfusions compared to 31% in
the low Hb target arm received transfusion (p

Technical Appendix 1 – Evidence Tables Page 49
STUDY OBJECTIVE STUDY POPULATION OUTCOMES
Mean baseline Hb=9.1 g/dL
Mean follow-up Hb=11.6 g/dL (at 16 weeks)
Prospective observational
Boelaert (1990) 4 Examine the increased risk of bacteremia
with transfusional iron overload in patients
on hemodialysis
Churchill
(1992) 5
Determine the transfusion practices prior
to the availability of EPOGEN ®
Retrospective observational
Ibrahim (2008) 6 Retrospective cohort study to address
patterns and trends in red blood cell
(RBC) transfusions in US hemodialysis
patients (1992-2005)
Ibrahim (2009) 7 Retrospective cohort study to address
patterns and trends in red blood cell
(RBC) transfusions in the CKD and non-
CKD population in Medicare between
1992 to 2004
HD adult patients who have received
transfusion in the past, and stratified by iron
level:
Ferretin < 500
Ferritin 500-1000
Ferritin ≥1000
N=158
Incident HD patients (after one month on
dialysis) and prevalent patients who remained
on dialysis for 6 months
N=1158
All Medicare HD patients between 1992 and
2005
N=77,347 (1992)
N=164,933 (2005)
> 90% of hemodialysis received ESA
treatment during this period
All Medicare beneficiaries with CKD between
1992 and 2004
Total CKD N=301,000
N=14,057 (1992)
N=42,225 (2004)
Total CKD with anemia N=118,895
N=4371 (1992)
N=18,579
• Acquired transfusional iron overload in hemodialysis patients
is associated with a greater risk of bacteremia
• The bacterimic incidence was 2.92 times higher in the Ferritin
> 1000 than in the Ferritin < 1000 and was independent of
age or dialysis vintage
• Mean pre-transfusion Hb 6.87 g/dL
• 639 patients (55.2%) received at least 1 unit RBC transfusion
(8.1 units/pt yr)
o 7% had at least one transfusion reaction
o The probability of surviving 12 months without receiving
a blood transfusion was 47.2% for males and 27.5% for
females
• Mean population Hb increased from 9.7 g/dL in 1992 to
11.8 g/dL in 2004
• The total transfusion rate declined from 535.33 per 1000
patient-years in 1992 to 263.65 per 1000 patient-years in
2005
• For transfusions occurring in the outpatient setting only,
transfusions declined from 225.63 per 1000 patient-years in
1992 to 66.66 per 1000 patient-years in 2005
• The proportion of CKD patients receiving ESA therapy
increased from 0.8% in 1992 to 7.5% in 2004
• The total transfusion among anemic CKD patients not
receiving ESA therapy was 249.2 per 1000 patient-years in
1992 and 174.7 per 1000 patient-years in 2004
• The total transfusion among anemic CKD patients who
received ESA therapy was 572.9 per 1000 patient-years in
1992 and 200.1 per 1000 patient-years in 2004

Technical Appendix 1 – Evidence Tables Page 50
STUDY OBJECTIVE STUDY POPULATION OUTCOMES
Seliger (2010) 8
Determine the association of timing of
ESA initiation with clinical outcomes
Patient with CKD-NOD (CKD
< 60 mL/min/1.73 m
among CKD-NOD population
2 ) and anemia (excluded
• Transfusion incidence rate was 19.5 per 100 person-years
among the delayed ESA treatment initiation patients and 13.4
patients with ESRD, hematological
per 100 person-years among the early ESA treatment
malignancy, on chemotherapy, hospitalized for
have gastrointestinal bleeding)
initiation patients (HR=0.71, 95% CI: 0.59-0.97)
Early ESA treatment initiation (Hb 10-11 g/dL)
N=705
Delayed ESA treatment initiation
(Hb < 10 g/dL)
N=705
Early and delayed treatment patients were
propensity score matched
These evidence tables include the most commonly cited references; a systematic review was not conducted
1. Besarab A, Bolton WK, Browne JK, et al.: The effects of normal as compared with low hematocrit values in patients with cardiac disease who are receiving
hemodialysis and epoetin. N Engl J Med. 1998;339: 584-590.
2. Pfeffer MA, Burdmann EA, Chen CY, et al.: A trial of darbepoetin alfa in type 2 diabetes and chronic kidney disease. The New England journal of medicine.
2009;361: 2019-2032.
3. Provenzano R, Garcia-Mayol L, Suchinda P, et al.: Once-weekly epoetin alfa for treating the anemia of chronic kidney disease. Clin Nephrol. 2004;61: 392-405.
4. Boelaert JR, Daneels RF, Schurgers ML, Matthys EG, Gordts BZ and Van Landuyt HW: Iron overload in haemodialysis patients increases the risk of bacteraemia:
a prospective study. Nephrol Dial Transplant. 1990;5: 130-134.
5. Churchill DN, Taylor DW, Cook RJ, et al.: Canadian Hemodialysis Morbidity Study. Am J Kidney Dis. 1992;19: 214-234.
6. Ibrahim HN, Ishani A, Foley RN, Guo H, Liu J and Collins AJ: Temporal trends in red blood transfusion among US dialysis patients, 1992-2005. Am J Kidney Dis.
2008;52: 1115-1121. .
7. Ibrahim HN, Ishani A, Guo H and Gilbertson DT: Blood transfusion use in non-dialysis-dependent chronic kidney disease patients aged 65 years and older.
Nephrol Dial Transplant. 2009;24: 3138-3143.
8. Seliger: Timing of ESA initiation and adverse outcomes in non-dialysis dependent CKD: a propensity matched observational cohort study. Clin J Am Soc Nephrol.
2010;In press.

Technical Appendix 1 – Evidence Tables Page 51
Table 4: Patient Reported Outcomes (PROs) in Dialysis Patients
STUDY PRO INSTRUMENT TREATMENT N BASELINE Hb
MEAN (SD) or
Mean (95% CI)
(g/dL)
RCT
Furuland
(2003) 1
48-76 weeks
N=416
Foley
(2001) 2
48 weeks
N=104
(13.5 g/dL
arm, n=60;
10 g/dL arm,
n=54)
KDQ-physical
symptoms
KDQ- Fatigue
KDQ- Depression
KDQ- Frustration
KDQ- Relations
KDQ-fatigue
Normal Hb with EA 129 10.9±1.1
Sub normal Hb with EA 124 11.0±0.9
Normal Hb with EA 129
10.9±1.1
Sub normal Hb with EA 124 11.0±0.9
Normal Hb with EA 129 10.9±1.1
Sub normal Hb with EA 124 11.0±0.9
Normal Hb with EA 129 10.9±1.1
Sub normal Hb with EA 124 11.0±0.9
Normal Hb with EA 129 10.9±1.1
Sub normal Hb with EA 124 11.0±0.9
EA target 13.5 g/dL
(Concentric LV
hypertrophy)
EA target 13.5 g/dL
(LV dilation)
EA target 10 g/dL
(Concentric LV
hypertrophy)
EA target 10 g/dL
(LV dilation)
34 12.2 (11.9,12.5)
39 12.3 (12.0,12.5)
36
36
10.4 (10.2,10.6)
10.4 (10.2,10.6)
BASELINE
PRO
3.83
(1.67–7.00)
4.00
(1.00–7.00)
4.67
(1.17–7.00)
5.00
(1.00–7.00)
5.40
(1.40–7.00)
5.60
(1.60–7.00)
6.00
(3.00–7.00)
6.00
(1.00–7.00)
5.17
(1.83–6.67)
5.00
(1.17–6.83)
4.41
(4.10,4.73)
4.53
(4.15,4.91)
Hb (g/dL)/Hct(%)
LEVELS or
CHANGE IN Hb
(g/dL)/Hct (%)
LEVELS
2.2-3.7
0.3-0.8
2.2-3.7
0.3-0.8
2.2-3.7
0.3-0.8
2.2-3.7
0.3-0.8
2.2-3.7
0.3-0.8
NR
NR
CHANGE IN PRO STAT
SIGNIF
Median (Min-Max)
0.66
(–1.17–3.00)
0.25
(–1.17–2.66)
0.16
(–2.50–2.34)
-0.33
(-3.17–2.00)
0.00
(-4.00–2.60)
-0.40
(-2.40–2.20)
0.00
(-1.67–2.34)
0.00
(-2.00–2.00)
0.00
(-2.66–2.67)
0.00
(-2.67–2.33)
Mean (95%CI)
to week 24:
0.46 (0.17,0.75)
to week 48:
0.25 (-0.05,0.55)
To week 24:
-0.04 (-0.22,0.14)
To week 48:
-0.01 (-0.30,0.29)
0.02
0.05
0.01
0.05
0.19
To week
24:
P=0.004
To week
48:
P=0.22

Technical Appendix 1 – Evidence Tables Page 52
STUDY PRO INSTRUMENT TREATMENT N BASELINE Hb
MEAN (SD) or
Mean (95% CI)
(g/dL)
KDQ-depression
KDQ-relationships
KDQ-physical
symptoms
KDQ—Frustration
EA target 13.5 g/dL
(Concentric LV
hypertrophy)
EA target 13.5 g/dL
(LV dilation)
EA target 10 g/dL
(Concentric LV
hypertrophy)
EA target 10 g/dL
(LV dilation)
EA target 13.5 g/dL
(Concentric LV
hypertrophy)
EA target 13.5 g/dL
(LV dilation)
EA target 10 g/dL
(Concentric LV
hypertrophy)
EA target 10 g/dL(LV
dilation)
EA target 13.5 g/dL
(Concentric LV
hypertrophy)
EA target 13.5 g/dL
(LV dilation)
EA target 10 g/dL
(Concentric LV
hypertrophy)
EA target 10 g/dL(LV
dilation)
EA target 13.5 g/dL
(Concentric LV
hypertrophy)
34
39
12.2 (11.9,12.5)
12.3 (12.0,12.5)
36 10.4 (10.2,10.6)
36
10.4 (10.2,10.6)
34 12.2 (11.9,12.5)
39 12.3 (12.0,12.5)
36
36
10.4 (10.2,10.6)
10.4 (10.2,10.6)
34 12.2 (11.9,12.5)
39 12.3 (12.0,12.5)
36 10.4 (10.2,10.6)
36 10.4 (10.2,10.6)
34 12.2 (11.9,12.5)
EA target 13.5 g/dL 39 12.3 (12.0,12.5)
BASELINE
PRO
5.02
(4.69,5.35)
5.20
(4.81,5.59)
5.04
(4.78,5.29)
4.96
(4.49,5.32)
3.49
(3.17,3.81)
3.68
(3.28,4.08)
4.93
(4.57,5.30)
Hb (g/dL)/Hct(%)
LEVELS or
CHANGE IN Hb
(g/dL)/Hct (%)
LEVELS
NR
NR
NR
NR
NR
NR
NR
CHANGE IN PRO STAT
SIGNIF
To week 24:
0.30 (0.06,0.53)
To week 48:
0.05 (-0.28,0.37)
To week 24:
0.00 (-0.23,0.24)
To week 48:
0.20 (-0.11,0.51)
To week 24: 0.31
(0.15,0.48)
To week 48:
-0.07 (-0.38,0.24)
To week 24:
-0.04 (-0.31,0.23)
To week 48:
0.20 (-0.15,0.46)
To week 24:
0.79 (0.39,1.20)
To week 48:
1.10 (0.55,1.64)
To week 24:
0.71 (0.38,1.04)
To week 48:
1.17 (0.76,1.57)
To week 24:
0.16 (-0.08,0.39)
To week 48:
-0.07 (-0.38,0.24)
Week
24:
P=0.083
Week
48:
P=0.49
Week
24:
P=0.025
Week
48:
P=0.31
Week
24:
P=0.76
WEEK
48:
P=0.84
Week
24:
P=0.36
Week

Technical Appendix 1 – Evidence Tables Page 53
STUDY PRO INSTRUMENT TREATMENT N BASELINE Hb
MEAN (SD) or
Mean (95% CI)
(g/dL)
Muirhead
(1992) 3
(SC Epoetin
alfa vs IV
Epoetin alfa)
24 weeks
Physical
Fatigue
Relationship
Depression
(LV dilation)
EA target 10 g/dL
(Concentric LV
hypertrophy)
EA target 10 g/dL
(LV dilation)
SC 45
IV 36
SC 45
IV 36
SC 45
IV 36
36 10.4 (10.2,10.6)
36 10.4 (10.2,10.6)
After run-in:
8.01±1.37
After run-in:7.74
±1.53
After run-in:
8.01±1.37
After run-in:7.74
±1.53
After run-in:
8.01±1.37
After run-in:7.74
±1.53
BASELINE
PRO
5.07
(4.66,5.08)
4.3±1.2
4.3±1.1
4.5±1.4
4.3±1.4
5.2±1.0
4.9±1.2
Hb (g/dL)/Hct(%)
LEVELS or
CHANGE IN Hb
(g/dL)/Hct (%)
LEVELS
NR
Mean (SD)
After stabilization:
11.8±0.81;
24 week:
10.93±1.05
After stabilization
11.5±0.8
24 week:
11.2±1.1
After stabilization:
11.7±0.81;
24 week:
10.93±1.05
After stabilization
11.5±0.8
24 week:
11.2±1.1
After stabilization:
11.7±0.81;
24 week:
10.93±1.05
After stabilization
11.5±0.8
24 week:
11.2±1.1
SC 45 After run-in: 5.1±1.4 After stabilization:
CHANGE IN PRO STAT
SIGNIF
To week 24:
-0.01 (-0.31,0.27)
To week 48:
0.15 (-0.15,0.46)
48:
P=0.31
After stabilization: 0.53±0.890*;
week 8
0.70±1.0*;
week 16
0.55±1.1*;
week 24
0.69±1.2*
(*p

Technical Appendix 1 – Evidence Tables Page 54
STUDY PRO INSTRUMENT TREATMENT N BASELINE Hb
MEAN (SD) or
Mean (95% CI)
(g/dL)
Frustration
Global physical
Global emotional
Non-randomized controlled studies
IV 36
SC 45
IV 36
SC 45
IV 36
SC 45
IV 36
BASELINE
PRO
Hb (g/dL)/Hct(%)
LEVELS or
CHANGE IN Hb
(g/dL)/Hct (%)
LEVELS
8.01±1.37 11.7±0.81;
24 week:
10.93±1.05
After run-in:7.74
±1.53
After run-in:
8.01±1.37
After run-in:7.74
±1.53
After run-in:
8.01±1.37
After run-in:7.74
±1.53
After run-in:
8.01±1.37
After run-in:7.74
±1.53
5.0±1.4
5.3±1.3
4.9±1.5
4.5±1.2
4.4±1.2
5.2±1.1
4.9±1.2
After stabilization
11.5±0.8
24 week:
11.2±1.1
After stabilization:
11.7±0.81;
24 week:
10.93±1.05
After stabilization
11.5±0.8
24 week:
11.2±1.1
After stabilization:
11.7±0.81;
24 week:
10.93±1.05
After stabilization
11.5±0.8
24 week:
11.2±1.1
After stabilization:
11.7±0.81;
24 week:
10.93±1.05
After stabilization
11.5±0.8
24 week:
11.2±1.1
CHANGE IN PRO STAT
SIGNIF
0.15±0.94
0.00±0.97
0.07±0.98
0.07±0.86
0.03±0.94
-0.06±0.98
-0.02±0.98
0.46±0.75*
0.54±0.92*
0.35±1.1*
0.46±1.1*
0.12±0.69
0.11±0.76
0.01±0.83
0.08±0.82

Technical Appendix 1 – Evidence Tables Page 55
STUDY PRO INSTRUMENT TREATMENT N BASELINE Hb
MEAN (SD) or
Mean (95% CI)
(g/dL)
Barany
(1993) 4
Physical activities,
physical symptoms,
sleep, sexual
adjustment, emotional
wellbeing
Open-label single arm studies
Evans
(1990) 5
NR
24(EPO)
8 (control)
BASELINE
PRO
Kamotsky Index EA 333 NR 3.30 (1.33)
Hb (g/dL)/Hct(%)
LEVELS or
CHANGE IN Hb
(g/dL)/Hct (%)
LEVELS
CHANGE IN PRO STAT
SIGNIF
NR NR NR NR P

Technical Appendix 1 – Evidence Tables Page 56
STUDY PRO INSTRUMENT TREATMENT N BASELINE Hb
MEAN (SD) or
Mean (95% CI)
(g/dL)
BASELINE
PRO
Hb (g/dL)/Hct(%)
LEVELS or
CHANGE IN Hb
(g/dL)/Hct (%)
LEVELS
CHANGE IN PRO STAT
SIGNIF
Social isolation 19.3% 13.6%; 14.2% 0.032
mobility 21.1% 16.3%;19.0% 0.263
Job or work 37.1% 24.4%; 28.9% 0.080
home 55.0% 41.6%; 42.4% 0.010
Social life 46.9% 36%; 40.8% 0.160
Personal relationship 28.4% 21%; 21.3% 0.160
Sex life 49.2% 40.9%; 35.7% 0.004
hobbies 51.1% 33.1%; 32.9% 0.001
Holidays 63.6% 57.9%; 64.6% 0.780
Well-being EA 333 NR 11.24 (3.13)
Psychological affect 5.05 (1.35)
Life satisfaction 5.63 (1.77)
Average hematocrit
1 st follow-up (mean
4.4 months): 0.35;
2 nd follow-up (mean
10.3 months): 0.34
These evidence tables include the most commonly cited references; a systematic review was not conducted
11.81 (2.66);
11.87 (2.70)
5.30 (1.07); 5.25
(1.20)
5.92 (1.61); 6.02
(1.59)
1. Furuland H, Linde T, Ahlmen J, Christensson A, Strombom U and Danielson BG: A randomized controlled trial of haemoglobin normalization with epoetin alfa in
pre-dialysis and dialysis patients. Nephrol Dial Transplant. 2003;18: 353-361.
2. Foley RN, Parfrey PS, Morgan J, et al.: Effect of hemoglobin levels in hemodialysis patients with asymptomatic cardiomyopathy. Kidney international. 2000;58:
1325-1335.
3. Muirhead N, Laupacis A and Wong C: Erythropoietin for anaemia in haemodialysis patients: results of a maintenance study (the Canadian Erythropoietin Study
Group). Nephrol Dial Transplant. 1992;7: 811-816.
4. Barany P, Pettersson E and Konarski-Svensson JK: Long-term effects on quality of life in haemodialysis patients of correction of anaemia with erythropoietin.
Nephrol Dial Transplant. 1993;8: 426-432.
5. Evans RW, Rader B and Manninen DL: The quality of life of hemodialysis recipients treated with recombinant human erythropoietin. Cooperative Multicenter EPO
Clinical Trial Group. JAMA. 1990;263: 825-830.
0.004
0.003
0.017

Technical Appendix 1 – Evidence Tables Page 57
Table 5: Cardiovascular and All-cause Morbidity/Mortality Outcomes Related to Target and Achieved Hemoglobin Levels
STUDY OBJECTIVE STUDY POPULATION OUTCOMES
Randomized trials
Besarab (1998) 1
Open label
To examine the risks and benefits of
normalizing the hematocrit in patients with
cardiac disease who were undergoing
hemodialysis.
Drueke (2006) 2 Open label
Determine effect of treating with ESA to a
high Hb target of 13-15 g/dL compared to
a low Hb target of 10.5 – 11.5 g/dL on CV
events on patients with Stage IV CKD
patients.
Singh (2006) 3 Open label
Determine effect of treating with ESA to a
high Hb target of 13.5 g/dL compared to a
low Hb target of less than 11.3 g/dL on CV
events and mortality
Parfrey (2005) 9 Randomized double blind
Determine effect of treating with ESA to a
high Hb target of 13.5 – 14.5 g/dL
compared to a low Hb target of less than
9.5 - 11.5 g/dL on left ventricular volume
index
Pfeffer (2009) 4 Randomized, double-blind, placebocontrolled
Determine effect of treating with ESA to a
target of 13 g/dL on composite CV
morbidity and mortality compared to
placebo (with rescue at Hb

Technical Appendix 1 – Evidence Tables Page 58
STUDY OBJECTIVE STUDY POPULATION OUTCOMES
Retrospective observational studies
Collins (2001) 5 To assess mortality, hospitalization, and
Medicare allowable expenditures in
incident hemodialysis patients
Ofsthun (2003) 6
Li (2004) 7
To determine the association between
hemoglobin level and risk of mortality and
hospitalization.
To determine the association between Hct
and cardiac risk
Regidor (2006) 8 Determine the association between
achieved hemoglobin level and CV/all
cause mortality
Incident HD patients
Time frame: January 1, 1996 and June 30,
1998
N: 66,761
HD patients in Fresenius Medical Care-North
America facilities
N:44,550
Time frame: July 1, 1998~June 30, 2000
Incident HD patients
At least 9 months f/u
N: 50,579
Time frame: Jan 1, 1998 ~Dec 31, 1999.
HD patients in DaVita database
N=58,058
Twelve categories of blood hemoglobin (< 9
g/dL, ≥ 14 g/dl, and 10 g/dL, increments of
0.5 g/dL in between)
These evidence tables include the most commonly cited references; a systematic review was not conducted.
• Relative risk of death and all-cause, cardiovascular, and
infectious hospitalizations increased significantly with
declining Hb (HCT).
• There is an increased risk of death with decreasing Hb,
comparing Hb 39% to 33-36%
• Compared to a time-varying hemoglobin of 11.5-12 g/dL, a
time-varying hemoglobin of 9.5-

Technical Appendix 1 – Evidence Tables Page 59
Table 6: Relationship Between Achieved Hemoglobin (Hematocrit) and the Risk of Hospitalization and Healthcare Resource
Utilization (HRU)
STUDY OBJECTIVE STUDY POPULATION HOSPITALIZATION AND HRU OUTCOME SUMMARY
Retrospective data analysis
Xia (1999) 1
To evaluate the association
between hematocrit level and
risk of hospitalization
Collins (2001) 2 To evaluate the association
between baseline hematocrit
level and the risk of
hospitalization and death
Ofsthun (2003) 3
Li (2004) 4
To evaluate the association
between hemoglobin level and
hospitalization burden and
length of stay
To evaluate the association
between hematocrit and risk of
hospitalization
Seliger (2010) 5 To evaluate the association
between the timing of ESA
treatment initiation and the risk
of hospitalization
US HD patients
N=71,717
Time frame: July 1993-December 1994
Incident US HD patients
N=66,761
Time frame: 1996 - 1998
Follow-up period: 1 year
US HD patients
N=44,550
Time frame: July 1998-June 2000.
Incident US HD patients with at least 9 months of
follow-up.
N=50,579
Time frame: January, 1998- December, 1999.
CKD-NOD patients
(eGFR 15-

Technical Appendix 1 – Evidence Tables Page 60
STUDY OBJECTIVE STUDY POPULATION HOSPITALIZATION AND HRU OUTCOME SUMMARY
Prospective Data Analysis
Churchill
(1995) 6
To evaluate the effect of ESA
use on utilization of hospital
Multicenter Canadian prospective cohort study in
HD patients
• There were 58 hospitalizations in the EPO group vs. 97 in the nontreated
group.
resources in hemodialysis N=67 EPO treated hemodialysis patients
• Average days hospitalized per admission: 15.3 days (EPO treated) vs.
patients
Baseline Hb 78.6 g/L
N=67 non-treated matched patient
Baseline Hb 74.0 g/L
23.2 (non-treated)
These evidence tables include the most commonly cited references; a systematic review was not conducted
1. Xia H, Ebben J, Ma JZ and Collins AJ: Hematocrit levels and hospitalization risks in hemodialysis patients. J Am Soc Nephrol. 1999;10: 1309-1316.
2. Collins AJ, Li S, St Peter W, et al.: Death, hospitalization, and economic associations among incident hemodialysis patients with hematocrit values of 36 to 39%. J
Am Soc Nephrol. 2001;12: 2465-2473.
3. Ofsthun N, Labrecque J, Lacson E, Keen M and Lazarus JM: The effects of higher hemoglobin levels on mortality and hospitalization in hemodialysis patients.
Kidney Int. 2003;63: 1908-1914.
4. Li S, Foley RN and Collins AJ: Anemia, hospitalization, and mortality in patients receiving peritoneal dialysis in the United States. Kidney Int. 2004;65: 1864-1869.
5. Seliger: Timing of ESA initiation and adverse outcomes in non-dialysis dependent CKD: a propensity matched observational cohort study. Clin J Am Soc Nephrol.
2010;In press.
6. Churchill DN, Muirhead N, Goldstein M, et al.: Effect of recombinant human erythropoietin on hospitalization of hemodialysis patients. Clinical nephrology.
1995;43: 184-188.

Technical Appendix 1 – Evidence Tables Page 61
Table 7: Hemoglobin Variability in CKD Patients
STUDY OBJECTIVE STUDY POPULATION OUTCOMES
Randomized prospective trial
West (2007) 1 Open-label, single-center
To characterize Hb variability
using functional data analysis
Retrospective data analysis
Fishbane
(2005) 2
Lacson (2003) 3
To describe hemoglobin cycling in
ESA-treated HD patients; to analyze
etiologic factors; assess Hb
fluctuations on iron storage; and to
study factors associated with a
greater degree of cycling.
To quantify Hb level variability and
the likelihood of falling within the Hb
level goal range of 11 to 12 g/dL.
Ebben (2006) 4 To characterize hemoglobin level
variability in US dialysis patients
Ad-hoc analysis of data from prevalent
HD patients, iron-replete
N=151
HD Patients on chronic maintenance
outpatient hemodialysis in a single center
N=281
Time frame: 1998-2003
US HD patients in Fresenius Medical Care
N=65,009
Time frame: Jan 1-Dec 31, 2000
US HD patients
N=159,720
Assess variability of 6-month period based
on patterns of hemoglobin change using the
categories
• < 11 g/dL
• 11-12.5 g/dL
• For majority of patients, Hb instability was small, typified
by the overall mean value of 0.63 g/dL per mo
• The data show trajectories that vary widely and form a
continuum, rather than obvious subsets
• The quantitative analysis of individual responses within
the system revealed a continuous spectrum of measurable
Hb instability, overlapping for each treatment group
• Greater than 90% of patients experienced Hb cycling defined as
one or more annual, full, extended cycle of amplitude >1.5 Hb
g/dL of a duration more than 8 weeks
• A subpopulation of patients was identified with more frequent
cycling (>2 cycles per year). No clinical factors were identified in
this group other than they were more responsive to ESA than
other patients (ERI 1036 ± 659 U/week/g Hb compared to 1992 ±
701 U/week/g Hb respectively) (p=0.02)
• 38.4% patients had Hb 11~12 g/dL. In only 8% did Hb levels
consistently remain less than 11 g/dL, and in 18%, greater than 12
g/dL all year. 29% (18,633 patients) moved from below to above
target range or vice versa. The average individual patient is
expected to have a±1.4-g/dL fluctuation in 3-month rolling average
Hgb levels per year.
• Greater mean facility Hb level correlated with a greater
percentage of patients with Hb levels greater than 10 g/dL
(R 2 =0.49) and greater than 12.5 g/dL (R 2 =0.61).
• Despite increased mean Hb levels and EPO and iron use, the
spread of the Hb distribution curve remained unchanged in the
last 6 years.
• 1.8% of patients had Hb < 11 for six consecutive months
• 6.5% of patients had Hb 11-12.5 for six consecutive months
• 2.0% of patients had Hb > 12.5 for six consecutive months
• 21.3% of patients had Hb fluctuations between < 11 and 11-12.5
over six months
• 28.9% of patients had Hb fluctuations between 11-12.5 and >
12.5 g/dL over six months

Technical Appendix 1 – Evidence Tables Page 62
STUDY OBJECTIVE STUDY POPULATION OUTCOMES
• > 12.5 g/dL
• 39.5% of patients had Hb < 11 and > 12.5 g/dL over six months
• Hospitalization events, comorbidity and intercurrent events were
association with persistently low (Hb < 11 for six months) and
transiently low (Hb < 11 and between 11 and 12.5) hemoglobin
levels
Gilbertson
(2008) 5
To evaluate association between
hemoglobin variability groups and
US HD patients receiving ESAs
N=159,720
• Patients with Hb < 11 for six consecutive months had
significantly elevated risk of mortality compared to those with Hb
mortality
Time frame: 2004
11-12.5 for six months (HR 2.18; 95%CI:1.93-2.45)
Hemoglobin variability groups:
• Patients with Hb < 11 and 11-12.5 had an elevated risk of death
• Patients with Hb < 11 for six months
(HR=1.44, 95% CI: 1.33-1.56)
• Patients with Hb 11-12.5 for six
consecutive months
• Patients with Hb > 12.5 for six
consecutive months
• Patients with Hb fluctuating between <
11 and 11-12.5 over six months
• Patients with Hb > 11 for six
consecutive months
• Patients with Hb levels across Hb
distribution
Spiegel (2009) 6
To evaluate the impact of recent
events on ESA dosing patterns and
consequently on hemoglobin (Hb)
distribution characteristics in dialysis
patients.
Yang (2007) 7 To evaluate the association between
hemoglobin variability and mortality
Minutolo (2009) 8 To evaluate whether intensity of
epoetin therapy affects hemoglobin
variability and renal survival in CKD-
HD patients from ~87% of dialysis centers in
the US.
N: 743,000
Time frame: June 2006-November 2008
HD patients have ≥3 Hb measurements, no
drop-out in exposure window
N=19,150
Hb variability measured by:
• absolute level of Hb (intercept),
• temporal trend in Hb (slope)
• Hb variability (residual standard
deviation).
Adult CKD-NOD patients with anemia
patients receiving epoetin
N: 137 patients (1,198 visits)
• Mean Hb level decreased by 0.37 g/dL during the indicated period
• standard deviation (SD) of the population Hb level distribution
decreased by 0.14 g/dL and skewness increased by -0.10
• Hb measurements in specific ranges changed as follows: >12
g/dL, decreased by 11.3 percentage points; 10-12 g/dL, increased
by 9.4 percentage points; and 13
g/dL for ≥3 months decreased by 2.9 percentage points
• Survival analyses indicated that each 1g/dL increase in the
residual standard deviation was associated with a 33% increase
in rate of death, even after adjusting for multiple baseline
covariates
• Patient characteristics accounted for very little of the variation in
hemoglobin variability metric
• Limitation: hospitalizations and other acute events were not
included in the model used
• Lower hemoglobin response to first epoetin dosage was an
independent predictor of higher therapeutic index(TI) (difference
between rates of visits that required EPO dosage change and

Technical Appendix 1 – Evidence Tables Page 63
STUDY OBJECTIVE STUDY POPULATION OUTCOMES
NOD patients. Time frame: April 30, 2002~October 31,
2005
These evidence tables include the most commonly cited references; a systematic review was not conducted.
those with effective EPO change) (P = 0.002)
• From lower to higher tertile of TI, Hb variability increased, as
shown by the reduction of time with Hb at target (time in target,
from 9.2 ±2.0 to 3.0±2.2 mo; p < 0.0001) and the wider values of
within-patient Hb standard deviation (from 0.70 to 0.96; P = 0.005)
and Hb fluctuations across target (p < 0.0001)
• Lack of adjustment of EPO worsens Hb variability in CKD. Hb
variability may be associated with renal survival, but further
studies are needed to explore the association versus causal
relationship
1. West RM, Harris K, Gilthorpe MS, Tolman C and Will EJ: Functional data analysis applied to a randomized controlled clinical trial in hemodialysis patients
describes the variability of patient responses in the control of renal anemia. J Am Soc Nephrol. 2007;18: 2371-2376.
2. Fishbane S and Berns JS: Hemoglobin cycling in hemodialysis patients treated with recombinant human erythropoietin. Kidney Int. 2005;68: 1337-1343.
3. Lacson E, Jr., Ofsthun N and Lazarus JM: Effect of variability in anemia management on hemoglobin outcomes in ESRD. Am J Kidney Dis. 2003;41: 111-124.
4. Ebben JP, Gilbertson DT, Foley RN and Collins AJ: Hemoglobin level variability: associations with comorbidity, intercurrent events, and hospitalizations. Clin J Am
Soc Nephrol. 2006;1: 1205-1210. .
5. Gilbertson DT, Ebben JP, Foley RN, Weinhandl ED, Bradbury BD and Collins AJ: Hemoglobin level variability: associations with mortality. Clin J Am Soc Nephrol.
2008;3: 133-138.
6. Spiegel DM, Khan I, Krishnan M and Mayne TJ: Changes in hemoglobin level distribution in US dialysis patients from June 2006 to November 2008. Am J Kidney
Dis. 2009;55: 113-120.
7. Yang W, Israni RK, Brunelli SM, Joffe MM, Fishbane S and Feldman HI: Hemoglobin variability and mortality in ESRD. J Am Soc Nephrol. 2007;18: 3164-3170.
8. Minutolo R, Chiodini P, Cianciaruso B, et al.: Epoetin therapy and hemoglobin level variability in nondialysis patients with chronic kidney disease. Clin J Am Soc
Nephrol. 2009;4: 552-559.

Technical Appendix 1 – Evidence Tables Page 64
Table 8: Patient Reported Outcomes (PROs) in CKD-NOD Patients
STUDY DESIGN CHANGE
IN PRO
Mean
(SD)
Double-blind RCT – Energy/Vitality/Activity
Kleinman (1989) 1 *****
Energy (Single item VAS)
N = 14
PRO FU=12 weeks
Pfeffer (2009) 2
Energy/Vitality subscale of
the SF-36
N= 4047
PRO FU= 25 weeks
Singh (2006) 3
LASA Energy score
N= 1432
PRO FU= NR
The US Recombinant
Human Erythropoietin
Predialysis Study Group
(1991) 4 ***
Energy (5-point Likert
Scale)
N= 117
PRO FU=8 weeks
Treated (target
Hct: 38 – 40 %
Placebo NR
STAT
SIG §
EFFECT SIZE MCID BASELINE Hb
OR HCT
Hb OR
Hct AT
STUDY
END
CHANGE IN
Hb OR Hct
PER UNIT
CHANGE IN
ENERGY /
VITALITY /
ACTIVITY**
NR NE NE NR NR 7.7% (Hct) NE
NR*
NE NE NR NR -0.1% (Hct) NE
DA (High Hb
Target: 13 g/dL)
+2.6(9.9)
p = 0.20
0.234 No 10.4 (9.8-10.8) 12.5 g/dL 2.1 g/dL 1.2
Placebo +2.1(9.7)
0.190 No 10.4 (9.8-10.8) 10.6 g/dL 0.2 g/dL 10.5
High Hb Target
(13.5 g/dL)
Low Hb Target
(11.3 g/dL)
Anemia corrected
[Target Hct: 40%
(men) and 37%
(women)]
Open-label RCTs - Energy/Vitality/Activity
+16.6
(23.7) *
+15.5
(23.1) *
+1.45
(NR)
P = 0.7
0.70
(Moderate)
0.67
(Moderate)
p

Technical Appendix 1 – Evidence Tables Page 65
Drueke (2006) 5
Energy/Vitality subscale of
the SF-36
N =603
PRO FU= 12 months
MacDougall (2008) 6
Energy/Vitality subscale of
the SF-36
N = 297
PRO FU=29 weeks
High Hb Target
(13.0 – 15.0 g/dL)
Low Hb Target
(10.5 – 11.5 g/dL)
Darbepoetin alfa
CERA
+3.8
(NR)
-0.5
(NR)
+7.5
(NR)
+11.2
(NR)
STUDY DESIGN CHANGE
IN PRO
Mean
(SD)
Revicki (1995) 7
Energy/Vitality subscale of
the SF-36
N= 83
Untreated N=40
ESA treated N=43
PRO FU= 48 weeks
Treated
(target Hct: 36 %
Untreated
+5.8
(3.6) *
-3.1
(2.0)
p< 0.001
NR ***
STAT
SIG §
p = 0.04
NE No 11.6 (0.6) g/dL 13.5 g/dL 1.9 g/dL 2.0
NE No 11.6 (0.6) g/dL 11.6 g/dL 0.0 g/dL NE
NE Yes 10.2 (0.7) g/dL 12.2 g/dL 2.02 g/dL 3.7
NE Yes 10.2 (0.6) g/dL 12.3 g/dL 2.12 g/dL 5.3
EFFECT SIZE MCID BASELINE Hb
OR HCT
Hb OR
Hct AT
STUDY
END
CHANGE IN
Hb OR HCT
PER UNIT
CHANGE IN
ENERGY /
VITALITY /
ACTIVITY**
1.90 (Large) Yes 26.8%(4.5) 31.5% 4.7 % (Hct) 1.2
-0.84 (Large) No 26.8% (3.6) 25.8% -1.0 % (Hct) 3.1

Technical Appendix 1 – Evidence Tables Page 66
Rossert (2006) 8
Energy/Vitality subscale of
the SF-36
N= 224
PRO FU=16 weeks
Singh (2006) 3
Energy/Vitality subscale of
the SF-36
N= 1432
PRO FU= NR
Singh (2006) 3
LASA Activity Score
N=1432
PRO FU= NR
High Hb Target
(13.0 – 15.0 g/dL)
Low Hb Target
(11.0 – 12.0 g/dL)
High Hb Target
(13.5 g/dL)
Low Hb Target
(11.3 g/dL)
High Hb Target
(13.5 g/dL)
Low Hb Target
(11.3 g/dL)
NR p = 0.04
+10.0
(22.6) *
+8.2
(22.4) *
+15.0(39.
9) *
+13.3(29
.8) *
Open-label single-arm studies – Energy/Vitality/Activity
Alexander (2007) 9
Energy/Vitality subscale of
the SF-36
N= 48
PRO FU=16 weeks
Benz (2007) 10
Energy/Vitality subscale of
the SF-36
N= 67
PRO FU=16 weeks
+14.9
(3.2)
+14.1
(23.9)
p = 0.6
p

Technical Appendix 1 – Evidence Tables Page 67
Benz (2007) 10
LASA Activity Score
N=67
PRO FU=16 weeks
+17.0(21.
9)
STUDY DESIGN CHANGE
IN PRO
Mean
(SD)
Benz (2007) 10
LASA Energy Score
N= 67
PRO FU=16 weeks
Provenzano (2004) 11
LASA Activity Score
N=1184
PRO FU=16 weeks
Provenzano (2004) 11
LASA Energy Score
N= 1184
PRO FU=16 weeks
Provenzano (2005) 12
LASA Activity Score
Parallel groups
N= 519
PRO FU=16 weeks
Double-blind RCT – Physical Function
DA (High Hb
Pfeffer (2009) Target: 13 g/dL)
2
Subscale of the SF-36
N= 4047
PRO FU= 25 weeks
Placebo
+20.6
(21.8)
+24.5(24.
9)
+27.9
(23.4)
+5.4
(22.3)
p

Technical Appendix 1 – Evidence Tables Page 68
Drueke (2006) 5
Subscale of the SF-36
N=603
PRO FU=12 months
MacDougall (2008) 6
Subscale of the SF-36
N=297
PRO FU=29 weeks
Revicki (1995) 7
Subscale of the SF-36
N=83; PRO FU 48 weeks
Untreated N=40
ESA treated N=43
High Hb Target
(13.0 – 15.0 g/dL)
Low Hb Target
(10.5 – 11.5 g/dL)
DA
CERA
Treated
(target Hct: 36 %
+3.3 (NR) p

Technical Appendix 1 – Evidence Tables Page 69
Abu-Alfa (2008) 14
SF-36 Physical Composite
Score
N=277
PRO FU=52 weeks
Alexander (2007) 9
Subscale of the SF-36
N=48
PRO FU=16 weeks
Benz (2007) 10
Subscale of the SF-36
N=67
PRO FU=16 weeks
Islam (2005) 15
Nottingham Scale
N=45
PRO FU=6 months
Kleinman (1989) 16
Single item VAS
N=14
PRO FU=12 weeks
Single arm +3.3 (8.7) P

Technical Appendix 1 – Evidence Tables Page 70
** Change in PRO divided by change in Hb (or change in Hematocrit)
***The US Recombinant Human Erythropoietin Predialysis Study Group (1991) randomized patients to one of four treatment groups, including placebo. Quality of life results were reported
by two groups, anemia corrected and anemia not corrected.
****The US Recombinant Human Erythropoietin Predialysis Study Group (1991) reports the percent of patients in each arm showing at least 6% change in hematocrit.
***** Kleinman (1989) reports “quality of life improvement” in treated patients, but not in patients in placebo group.
These evidence tables include the most commonly cited references; a systematic review was not conducted.
1. Kleinman KS, Schweitzer SU, Perdue ST, Bleifer KH and Abels RI: The use of recombinant human erythropoietin in the correction of anemia in predialysis patients
and its effect on renal function: a double-blind, placebo-controlled trial. Am J Kidney Dis. 1989;14: 486-495.
2. Pfeffer MA, Burdmann EA, Chen CY, et al.: A trial of darbepoetin alfa in type 2 diabetes and chronic kidney disease. The New England journal of medicine.
2009;361: 2019-2032.
3. Singh AK, Szczech L, Tang KL, et al.: Correction of anemia with epoetin alfa in chronic kidney disease. N Engl J Med. 2006;355: 2085-2098.
4. The US Recombinant Human Erythropoietin Predialysis Study Group. Double-blind, placebo-controlled study of the therapeutic use of recombinant human
erythropoietin for anemia associated with chronic renal failure in predialysis patients. Am J Kidney Dis. 1991;18: 50-59.
5. Drueke TB, Locatelli F, Clyne N, et al.: Normalization of hemoglobin level in patients with chronic kidney disease and anemia. N Engl J Med. 2006;355: 2071-2084.
6. Macdougall IC, Walker R, Provenzano R, et al.: C.E.R.A. corrects anemia in patients with chronic kidney disease not on dialysis: results of a randomized clinical
trial. Clin J Am Soc Nephrol. 2008;3: 337-347.
7. Revicki DA, Brown RE, Feeny DH, et al.: Health-related quality of life associated with recombinant human erythropoietin therapy for predialysis chronic renal
disease patients. Am J Kidney Dis. 1995;25: 548-554.
8. Rossert J, Levin A, Roger SD, et al.: Effect of early correction of anemia on the progression of CKD. Am J Kidney Dis. 2006;47: 738-750.
9. Alexander M, Kewalramani R, Agodoa I and Globe D: Association of anemia correction with health related quality of life in patients not on dialysis. Curr Med Res
Opin. 2007;23: 2997-3008.
10. Benz R, Schmidt R, Kelly K and Wolfson M: Epoetin alfa once every 2 weeks is effective for initiation of treatment of anemia of chronic kidney disease. Clin J Am
Soc Nephrol. 2007;2: 215-221. .
11. Provenzano R, Garcia-Mayol L, Suchinda P, et al.: Once-weekly epoetin alfa for treating the anemia of chronic kidney disease. Clin Nephrol. 2004;61: 392-405.
12. Provenzano R, Bhaduri S and Singh AK: Extended epoetin alfa dosing as maintenance treatment for the anemia of chronic kidney disease: the PROMPT study.
Clinical nephrology. 2005;64: 113-123.
13. Roger SD, McMahon LP, Clarkson A, et al.: Effects of early and late intervention with epoetin alpha on left ventricular mass among patients with chronic kidney
disease (stage 3 or 4): results of a randomized clinical trial. J Am Soc Nephrol. 2004;15: 148-156.
14. Abu-Alfa AK, Sloan L, Charytan C, et al.: The association of darbepoetin alfa with hemoglobin and health-related quality of life in patients with chronic kidney
disease not receiving dialysis. Curr Med Res Opin. 2008;24: 1091-1100.
15. Islam S, Rahman H and Rashid HU: Effect rHuEpo on predialysis CRF patients: study of 45 cases. Bangladesh Medical Research Council bulletin. 2005;31: 83-
87.
16. Kleinman KS, Schweitzer SU, Perdue ST, Bleifer KH and Abels RI: The use of recombinant human erythropoietin in the correction of anemia in predialysis patients
and its effect on renal function: a double-blind, placebo-controlled trial. Am J Kidney Dis. 1989;14: 486-495.

Technical Appendix 2 – Literature References Page 71
Technical Appendix 2 – Literature References

584
·
August 27, 1998
The New England Journal of Medicine
THE EFFECTS OF NORMAL AS COMPARED WITH LOW HEMATOCRIT VALUES
IN PATIENTS WITH CARDIAC DISEASE WHO ARE RECEIVING HEMODIALYSIS
AND EPOETIN
ALLEN
ANATOLE
BESARAB,
M.D., W. KLINE
BOLTON,
M.D., J
R. NISSENSON,
M.D., DOUGLAS
M. OKAMOTO,
PH.D.,
S
ABSTRACT
Background In patients with end-stage renal disease,
anemia develops as a result of erythropoietin
deficiency, and recombinant human erythropoietin
(epoetin) is prescribed to correct the anemia partially.
We examined the risks and benefits of normalizing
the hematocrit in patients with cardiac disease
who were undergoing hemodialysis.
Methods We studied 1233 patients with clinical evidence
of congestive heart failure or ischemic heart
disease who were undergoing hemodialysis: 618 patients
were assigned to receive increasing doses of
epoetin to achieve and maintain a hematocrit of 42
percent, and 615 were assigned to receive doses of
epoetin sufficient to maintain a hematocrit of 30 percent
throughout the study. The median duration of
treatment was 14 months. The primary end point was
the length of time to death or a first nonfatal myocardial
infarction.
Results After 29 months, there were 183 deaths
and 19 first nonfatal myocardial infarctions among
the patients in the normal-hematocrit group and 150
deaths and 14 nonfatal myocardial infarctions among
those in the low-hematocrit group (risk ratio for the
normal-hematocrit group as compared with the lowhematocrit
group, 1.3; 95 percent confidence interval,
0.9 to 1.9). Although the difference in event-free
survival between the two groups did not reach the
prespecified statistical stopping boundary, the study
was halted. The causes of death in the two groups
were similar. The mortality rates decreased with increasing
hematocrit values in both groups. The patients
in the normal-hematocrit group had a decline
in the adequacy of dialysis and received intravenous
iron dextran more often than those in the low-hematocrit
group.
Conclusions In patients with clinically evident
congestive heart failure or ischemic heart disease
who are receiving hemodialysis, administration of
epoetin to raise their hematocrit to 42 percent is not
recommended. (N Engl J Med 1998;339:584-90.)
©1998, Massachusetts Medical Society.
ADVANCED kidney failure usually leads to
anemia, primarily as a result of deficient
renal erythropoietin production. Most patients
undergoing hemodialysis are treated
with recombinant human erythropoietin (epoetin)
to stimulate erythropoiesis and correct the anemia
partially. In a random sample of 940 patients at 188
U.S. hemodialysis centers obtained before the initi-
EFFREY
TEVE
K. BROWNE,
PH.D.,
J
J. SCHWAB,
M.D.,
AND
OAN
C. EGRIE
DAVID
, PH.D.,
A. GOODKIN,
M.D.
ation of this study, we found that 69 percent of the
patients had hematocrits of 27 to 33 percent, 15
percent had values below 27 percent, and 16 percent
had values above 33 percent (unpublished data). Yet
the normal ranges for hematocrit values are 37 to 48
percent for women and 42 to 52 percent for men,
prompting the question of whether increasing the
doses of epoetin would benefit patients who are
undergoing hemodialysis. Cerebral oxygen delivery
among patients with ischemic cerebrovascular disease,
for example, is maximal when the hematocrit is
40 to 45 percent. 2
Cardiac disease is the most common cause of death
among patients who are regularly receiving dialysis. 3
Among these patients, partial correction of anemia
reduces exercise-induced cardiac ischemia4,5
and ameliorates
the left ventricular hypertrophy4,6-9
that predisposes
patients to death and cardiac-related morbidity.
10-12 In two pilot studies of patients who were
receiving hemodialysis who maintained a normal hem-
atocrit value
13
(and unpublished data), there were no
increases in blood pressure, thrombosis of the vascular
access site, seizures, or cardiovascular events, and
normal hematocrit values were associated with an
improved quality of life, shorter hospitalizations, and
increased exercise performance. The present trial was
designed to examine the benefits and risks of a normal
hematocrit in a large group of patients with cardiac
disease who were undergoing hemodialysis.
METHODS
Study Subjects
In this randomized, prospective, open-label trial, we studied
1233 patients with congestive heart failure or ischemic heart disease
who were undergoing hemodialysis at 51 centers (see the Appendix).
The institutional review boards at all centers approved the
protocol, and all the patients gave written, informed consent. All
the patients had end-stage renal disease and were undergoing longterm
hemodialysis, and they had hematocrit values of 27 to 33
percent while receiving epoetin during the four weeks before enrollment.
Ninety percent of the patients received epoetin intrave-
From the Division of Nephrology and Hypertension, Henry Ford Hospital,
Detroit (A.B.); the Department of Internal Medicine, Division of Nephrology,
University of Virginia Health Sciences Center, Charlottesville
(W.K.B.); Amgen, Thousand Oaks, Calif. (J.K.B., J.C.E., D.M.O., D.A.G.);
the Department of Medicine, Division of Nephrology, University of California
at Los Angeles, Los Angeles (A.R.N.); and the Department of Medicine,
Nephrology Division, Duke University Medical Center, Durham,
N.C. (S.J.S.). Address reprint requests to Dr. Goodkin at Clinical Research,
Amgen, 1 Amgen Center Dr., Thousand Oaks, CA 91320.
Downloaded from www.nejm.org at Amgen, Inc. on September 2, 2008 .
Copyright © 1998 Massachusetts Medical Society. All rights reserved.
Page 72
1

NORMAL VS. LOW HEMATOCRIT IN PATIENTS WITH CARDIAC DISEASE RECEIVING HEMODIALYSIS AND EPOETIN
nously, and 88 percent received it three times per week. According
to the protocol, all patients had to have documented congestive
heart failure (defined as the need for hospitalization or nonroutine
ultrafiltration for congestive heart failure in the preceding two
years) or ischemic heart disease (defined as angina pectoris requiring
medication in the preceding two years, coronary artery disease
documented by cardiac catheterization, or prior myocardial infarction)
and a serum transferrin saturation of 20 percent or higher.
Exclusion criteria included a diastolic blood pressure of 100 mm
Hg or more; a life expectancy of less than six months; severe cardiac
disability (New York Heart Association class IV); myocardial
infarction, percutaneous transluminal coronary angioplasty, or coronary-artery
bypass grafting in the three months before the study
began; pericardial disease; cardiac valvular disease likely to require
surgery; cardiac amyloidosis; and androgen therapy.
Study Protocol
The patients were randomly assigned to one of two groups.
One was a normal-hematocrit group, in which the patients received
increasing doses of epoetin alfa (Epogen, Amgen, Thousand
Oaks, Calif.) to achieve and maintain hematocrit values of
42 percent (±3 percentage points). The other was a low-hematocrit
group, in which the patients received doses of epoetin sufficient
to maintain the hematocrit at 30 percent (±3 percentage
points). The planned duration of the study was three years after
the enrollment of the last patient. The epoetin was administered
intravenously or subcutaneously, according to the route of administration
at base line, and at the same frequency per week as
before the study. In the normal-hematocrit group, the dose was
increased by a factor of 1.5 on study entry. Subsequently, doses
were increased by 25 percent of the base-line dose if the hematocrit
had not increased by at least 2 percentage points during the
preceding two weeks. If the hematocrit increased by more than
4 percentage points in a two-week period, the dose was reduced
by 25 U per kilogram of body weight. Iron status (as determined
by the serum transferrin saturation and serum ferritin concentration)
was reevaluated in patients who had no responses to increases
in the dose of epoetin. In the low-hematocrit group, the dose
was adjusted by 10 to 25 U per kilogram at two-week intervals,
when needed, to maintain a hematocrit of 30 percent. To avoid
spuriously elevated hematocrit values that result from swelling of
red cells during transport to the central laboratory, we calculated
the hematocrit values (in percentage points) by multiplying the
hemoglobin concentrations (in grams per deciliter) by 3, and all
hematocrit values after randomization are expressed in this way.
Evaluations
All adverse events, hospitalizations, coronary-artery bypass graft
or percutaneous transluminal coronary angioplasty procedures,
thromboses at vascular access sites or any change or procedure involving
a vascular access site, transfusions, and deaths were recorded.
Serial records were kept of the patients’ hemodialysis regimens;
blood pressure; hematologic, chemical, and coagulation
profiles; serum transferrin saturation; and concomitant drug therapy.
We also recorded Kt/V, a unitless measure of the adequacy
of hemodialysis therapy, where K is the rate of urea clearance by
the artificial kidney, t is the duration of each hemodialysis session,
V is the volume of distribution of urea within the patient, and
1.20 is considered to be the minimal recommended value, although
the measurement technique varied among the centers and
was estimated from the urea reduction ratio for the small number
of patients treated at centers that did not calculate Kt/V. Blood
specimens for laboratory analyses were drawn just before hemodialysis.
Quality of life was assessed at base line and every six
months thereafter with the 36-item Medical Outcomes Study
Short-Form Health Survey, 14 which evaluates eight health-related
aspects: physical function, social function, physical role, emotional
role, mental health, energy, pain, and general health perceptions.
Each portion of the test is scored on a scale that ranges
from 0 (severe limitation) to 100 (no limitation).
The primary end point was the length of time to death or a
first nonfatal myocardial infarction. A myocardial infarction was
considered to be fatal if death occurred within 24 hours after the
infarction; the event was then counted as a death in tabulating the
primary end point. Three criteria were required for the diagnosis
of myocardial infarction: clinical suspicion of acute myocardial infarction,
a peak serum creatine kinase concentration that was
more than 1.5 times the upper limit of the normal range, and a
high fraction of creatine kinase MB. Secondary end points were
congestive heart failure requiring hospitalization, angina pectoris
requiring hospitalization, coronary-artery bypass grafting, percutaneous
transluminal coronary angioplasty, hospitalization for all
causes, change in cardiovascular drugs, red-cell transfusion, and
changes in the quality-of-life scores.
Statistical Analysis
We estimated that 1000 patients were required to provide the
study with a power of 90 percent to detect a 20 percent difference
in primary event-free survival after three years at an overall
a level of 0.05 for two-sided tests, for a risk ratio of 1.3 with the
log-rank test used to compare Kaplan–Meier curves. We used a
Lan–DeMets procedure for group-sequential testing with an
O’Brien–Fleming boundary. 15 The trial was stopped at the third
interim analysis. We report the risk ratio, with 95 percent confidence
intervals adjusted for the previous interim analyses, using
the method of repeated confidence intervals. 16 Analyses were conducted
on an intention-to-treat basis (data on patients who discontinued
the study regimen, switched to peritoneal dialysis, underwent
kidney transplantation, or died before receiving study
medication were included). We compared the adjusted event-free
times to death or myocardial infarction using Cox proportionalhazards
regression analysis with 11 prespecified covariates. 17 We
also assessed mortality using one prospectively defined subgroup
analysis that excluded patients in the normal-hematocrit group
who did not attain the target hematocrit value. Cumulative incidences
were calculated for other secondary end points.
RESULTS
A total of 1233 patients were enrolled between
October 27, 1993, and March 31, 1996; 618 were
assigned to the normal-hematocrit group, and 615
to the low-hematocrit group. The study period ranged
from 4 days to 30 months (median, 14 months).
Concern about safety at the third interim analysis
prompted the independent data monitoring committee
to recommend that the study be stopped.
The interim results of the study did not reach the
prespecified stopping boundary corresponding to an
overall 5 percent level of significance, complicating
the reporting and interpretation of P values.
The base-line characteristics of the two groups,
including the mean hematocrit values and mean epoetin
doses, were similar (Table 1). By six months the
mean hematocrit in the normal-hematocrit group
had increased to the target range. To maintain their
hematocrit values in the target range these patients
required approximately three times as much epoetin
as before the study (Fig. 1). Although the mean
hematocrit values were stable in the low-hematocrit
group throughout the study and in the normalhematocrit
group after six months, the values varied
considerably in individual patients, often in association
with intercurrent illnesses.
The probability of the primary end point (death
Downloaded from www.nejm.org at Amgen, Inc. on September 2, 2008 .
Copyright © 1998 Massachusetts Medical Society. All rights reserved.
Volume 339 Number 9
Page 73
·
585

or a first nonfatal myocardial infarction) is shown in
Figure 2. After 29 months, there were 183 deaths
and 19 first nonfatal myocardial infarctions among
the patients in the normal-hematocrit group and
150 deaths and 14 nonfatal myocardial infarctions
among the patients in the low-hematocrit group
(risk ratio for the normal-hematocrit group as compared
with the low-hematocrit group, 1.3; 95 percent
confidence interval, 0.9 to 1.9). Increasing age,
the presence of peripheral vascular disease, New
York Heart Association class III cardiac disability,
and the absence of hypertension at base line were
significant risk factors for death or a first nonfatal
myocardial infarction for both groups combined,
whereas sex, race, the type of vascular access, Kt/V,
and the presence of congestive heart failure, ischemic
heart disease, and diabetes mellitus were not. These
11 prespecified base-line characteristics do not explain
the differences in outcomes in the two groups, because
adjustment for these factors did not change
the risk ratio for the normal-hematocrit group as
586
TABLE
1. BASE-LINE
CHARACTERISTICS
OF THE PATIENTS.*
CHARACTERISTIC
·
August 27, 1998
The New England Journal of Medicine
NORMAL-
HEMATOCRIT
GROUP
(N=618)
LOW-
HEMATOCRIT
GROUP
(N=615)
Age (yr)
65±12 64±12
35
Female sex (%)
Race or ethnic group (%)
50
52
White
Black
45
41
42
44
30
Low-hematocrit group
Hispanic
8
9
Other
6
5
Duration of dialysis (yr)
Cause of renal failure (%)
3.2±3.6 3.1±3.3
25
Diabetes mellitus
Hypertension
42
28
46
27
0 10
20
30
Glomerulonephritis
Other
7
23
8
19
700
Type of vascular access (%)
Graft
66
67
600
Natural fistula
23
23
Catheter
10
10
500
Not specified
2
0
Hypertension (%)
Diabetes mellitus (%)
71
54
69
58
400
Normal-hematocrit group
Peripheral vascular disease (%)
Cardiac-related hospitalization (%)
39 38
300
Angina pectoris
32
28†
Congestive heart failure
Myocardial infarction
44
25
47
23
200
Coronary-artery bypass graft
Percutaneous transluminal coronary
angioplasty
20
10
19
9
100
Low-hematocrit group
New York Heart Association class (%)
0
I
II
29
51
31
52
0 10 20
30
III
Hematocrit (%)
19
30.5±3.0
15
30.5±2.9
Months
Epoetin dose (U/kg/wk) 146±103 153±119 Figure 1. Mean Monthly Hematocrit Values and Epoetin Doses
*Plus–minus values are means ±SD. Because of rounding, not all percentages
total 100.
†P=0.04.
during the Study in the Normal-Hematocrit and Low-Hematocrit
Groups.
Both the mean hematocrit values and the mean doses were
significantly different between the two groups from one month
onward (P

NORMAL VS. LOW HEMATOCRIT IN PATIENTS WITH CARDIAC DISEASE RECEIVING HEMODIALYSIS AND EPOETIN
Probability of Death or
Myocardial Infarction (%)
NO. AT RISK
Normal hematocrit
Low hematocrit
Figure 2. Kaplan–Meier Estimates of the Probability of Death or a First Nonfatal Myocardial Infarction
in the Normal-Hematocrit and Low-Hematocrit Groups.
ing hospitalization, congestive heart failure requiring
hospitalization, coronary-artery bypass grafting,
or percutaneous transluminal coronary angioplasty
(Table 3). During the study, 129 patients in the
normal-hematocrit group (21 percent) received redcell
transfusions, as compared with 192 patients in
the low-hematocrit group (31 percent) (P

CAUSE
OF DEATH
*Because of rounding, percentages do not total 100.
March 31, 1996, decreased at higher hematocrit values,
but the mortality rate in the normal-hematocrit
group was higher than that in the low-hematocrit
group for any given range of hematocrit values (Fig.
3). When the average hematocrit replaced group assignment
in the Cox regression analysis of the pre-
588
·
TABLE
2. CAUSES
OF DEATH.*
August 27, 1998
The New England Journal of Medicine
NORMAL-
HEMATOCRIT
GROUP
(N=195)
no. (%)
LOW-
HEMATOCRIT
GROUP
(N=160)
Cardiovascular causes
125 (64) 112 (70)
Cardiac arrest 30 (15) 24 (15)
Acute myocardial infarction 22 (11) 28 (18)
Arrhythmia 19 (10) 14 (9)
Unwitnessed death 18 (9) 9 (6)
Cerebrovascular accident 14 (7) 9 (6)
Coronary artery disease 6 (3) 7 (4)
Cardiomyopathy 5 (3) 6 (4)
Ischemic bowel 3 (2) 2 (1)
Congestive heart failure 2 (1) 2 (1)
Valvular heart disease 1 (1) 3 (2)
Cardiogenic shock 1 (1) 0
Other 4 (2) 8 (5)
Noncardiovascular causes
70 (36) 48 (30)
Sepsis or infection 32 (16) 22 (14)
Voluntary withdrawal from dialysis 11 (6) 7 (4)
Cancer 4 (2) 3 (2)
Hemorrhage 4 (2) 2 (1)
Encephalopathy 3 (2) 2 (1)
Hyperkalemia 3 (2) 2 (1)
Hepatic failure 0 3 (2)
Other 13 (7) 7 (4)
TABLE
3. INCIDENCE
OF SEVEN
SECONDARY
END
POINTS.
END
POINT
NORMAL-
HEMATOCRIT
GROUP
(N=618)
no. (%)
LOW-
HEMATOCRIT
GROUP
(N=615)
*The P values were calculated with Fisher’s exact test.
P
VALUE*
Red-cell transfusion 129 (21) 192 (31)

NORMAL VS. LOW HEMATOCRIT IN PATIENTS WITH CARDIAC DISEASE RECEIVING HEMODIALYSIS AND EPOETIN
who survived received an average of 152±150 mg of
iron dextran per four-week period and those who
died received an average of 214±190 mg per fourweek
period (P

590 · August 27, 1998
APPENDIX
The following principal investigators participated in the study: M. Allon,
Birmingham, Ala.; G. Appel, New York; R. Benz, Wynnewood, Pa.; J.
Berns, Philadelphia; K. Bleifer, Van Nuys, Calif.; J. Bower, Jackson, Miss.;
W. Cleveland, Atlanta; I. Cruz, Washington, D.C.; D. Domoto, St. Louis;
S. Fishbane, Mineola, N.Y.; D. Gentile, Orange, Calif.; L. Glowacki, Dallas;
J. Goldman, Philadelphia; J. Hertel, Augusta, Ga.; D. Hoffman, Miami; H.
Karp, Somers Point, N.J.; C. Kaupke, Orange, Calif.; K. Kleinman, Tarzana,
Calif.; S. Korbet, Chicago; A. Lauer, Brockton, Mass.; V. Lim, Iowa
City, Iowa; J. Lindberg, New Orleans; M. Linsey, Pasadena, Calif.; J. Mac-
Laurin, Columbus, Ohio; T. Marbury, Orlando, Fla.; S. Mischel, Hammond,
Ind.; D. Molony, Houston; J. Navarro, Tampa, Fla.; E. Paganini,
Cleveland; D. Price, Boston; R. Raja, Philadelphia; J. Reiter, Missoula,
Mont.; J. Rimmer, Burlington, Vt.; P. Schoenfeld, San Francisco; W. Smith,
New Orleans; D. Spiegal, Denver; M. Stegman, Memphis, Tenn.; K. Stenzel,
New York; J. Sugihara, Honolulu; G. Ting, Mountain View, Calif.; N.
Vaziri, Orange, Calif.; M. Walczyk, Portland, Oreg.; D. Van Wyck, Tucson,
Ariz.; S. Wen, Madison, Wis.; B. Wilkes, Manhasset, N.Y.; M. Weiner, Lancaster,
Pa.; and B. Wood, Kansas City, Mo. Data monitoring committee:
P. Parfrey (chairman), St. John’s, Newf., Canada; W. Bennett, Portland,
Oreg.; T. Fleming, Seattle; D. Levy, Framingham, Mass.; N. Muirhead,
London, Ont., Canada; M. Pfeffer, Boston; C. Winearls, Oxford, England.
REFERENCES
1. Case Records of the Massachusetts General Hospital: normal reference
laboratory values. N Engl J Med 1992;327:718-24.
2. Kusunoki M, Kimura K, Nakamura M, Isaka Y, Yoneda S, Abe H. Effects
of hematocrit variations on cerebral blood flow and oxygen transport
in ischemic cerebrovascular disease. J Cereb Blood Flow Metab 1981;1:
413-7.
3. Rostand SG, Rutsky EA. Cardiac disease in dialysis patients. In: Nissenson
AR, Fine RN, Gentile DE, eds. Clinical dialysis. 2nd ed. Norwalk,
Conn.: Appleton & Lange, 1990:409-46.
4. Macdougall IC, Lewis NP, Saunders MJ, et al. Long-term cardiorespiratory
effects of amelioration of renal anaemia by erythropoietin. Lancet
1990;335:489-93. [Erratum, Lancet 1990;335:614.]
5. Wizemann V, Kaufmann J, Kramer W. Effect of erythropoietin on ischemia
tolerance in anemic hemodialysis patients with confirmed coronary
artery disease. Nephron 1992;62:161-5.
6. Cannella G, La Canna G, Sandrini M, et al. Reversal of left ventricular
hypertrophy following recombinant human erythropoietin treatment of
anaemic dialysed uraemic patients. Nephrol Dial Transplant 1991;6:31-7.
7. Pascual J, Teruel JL, Moya JL, Liano F, Jimenez-Mena M, Ortuno J.
Regression of left ventricular hypertrophy after partial correction of anemia
with erythropoietin in patients on hemodialysis: a prospective study. Clin
Nephrol 1991;35:280-7.
8. Goldberg N, Lundin AP, Delano B, Friedman EA, Stein RA. Changes
in left ventricular size, wall thickness, and function in anemic patients treated
with recombinant human erythropoietin. Am Heart J 1992;124:424-
7.
9. Zehnder C, Zuber M, Sulzer M, et al. Influence of long-term amelioration
of anemia and blood pressure control on left ventricular hypertrophy
in hemodialyzed patients. Nephron 1992;61:21-5.
10. Silberberg JS, Barre PE, Prichard SS, Sniderman AD. Impact of left
ventricular hypertrophy on survival in end-stage renal disease. Kidney Int
1989;36:286-90.
11. Parfrey PS, Griffiths SM, Harnett JD, et al. Outcome of congestive
heart failure, dilated cardiomyopathy, hypertrophic hyperkinetic disease,
and ischemic heart disease in dialysis patients. Am J Nephrol 1990;10:213-
21.
12. Parfrey PS, Harnett JD, Barre PE. The natural history of myocardial
disease in dialysis patients. J Am Soc Nephrol 1991;2:2-12.
13. Eschbach JW, Glenny R, Robertson T, et al. Normalizing the hematocrit
(HCT) in hemodialysis patients (HDP) with EPO improves quality
of life (Q/L) and is safe. J Am Soc Nephrol 1993;4:425. abstract.
14. McHorney CA, Ware JE Jr, Raczek AE. The MOS 36-Item Short-
Form Health Survey (SF-36). II. Psychometric and clinical tests of validity
in measuring physical and mental health constructs. Med Care 1993;31:
247-63.
The New England Journal of Medicine
15. Lan KKG, DeMets DL. Discrete sequential boundaries for clinical trials.
Biometrika 1983;70:659-63.
16. Jennison C, Turnbull BW. Statistical approaches in interim monitoring
of medical trials: a review and commentary. Stat Sci 1990;5:299-317.
17. Cox DR. Regression models and life-tables. J R Stat Soc [B] 1972;34:
187-220.
18. Eschbach JW, Abdulhadi MH, Browne JK, et al. Recombinant human
erythropoietin in anemic patients with end-stage renal disease: results of a
phase III multicenter clinical trial. Ann Intern Med 1989;111:992-1000.
19. Evans RW, Rader B, Manninen DL, Cooperative Multicenter EPO
Clinical Trial Group. The quality of life of hemodialysis recipients treated
with recombinant human erythropoietin. JAMA 1990;263:825-30.
20. Beusterien KM, Nissenson AR, Port FK, Kelly M, Steinwald B, Ware
JE Jr. The effects of recombinant human erythropoietin on functional
health and well-being in chronic dialysis patients. J Am Soc Nephrol 1996;
7:763-73.
21. Marsh JT, Brown WS, Wolcott D, et al. rHuEPO treatment improves
brain and cognitive function of anemic dialysis patients. Kidney Int 1991;
39:155-63.
22. Lundin AP, Akerman MJ, Chesler RM, et al. Exercise in hemodialysis
patients after treatment with recombinant human erythropoietin. Nephron
1991;58:315-9.
23. Braumann KM, Nonnast-Daniel B, Böning D, Böcker A, Frei U. Improved
physical performance after treatment of renal anemia with recombinant
human erythropoietin. Nephron 1991;58:129-34.
24. Veys N, Vanholder R, Ringoir S. Correction of deficient phagocytosis
during erythropoietin treatment in maintenance hemodialysis patients. Am
J Kidney Dis 1992;19:358-63.
25. Sennesael JJ, Van der Niepen P, Verbeelen DL. Treatment with recombinant
human erythropoietin increases antibody titers after hepatitis B vaccination
in dialysis patients. Kidney Int 1991;40:121-8.
26. Collins A, Ma J, Umen A. Hematocrit as a time-dependent risk factor.
J Am Soc Nephrol 1994;5:439. abstract.
27. Lowrie EG, Ling J, Lew NL. The anemia of ESRD and related
thoughts about iron and EPO therapy. Waltham, Mass.: National Medical
Care, August 10, 1995 (memorandum).
28. Canadian Erythropoietin Study Group. Association between recombinant
human erythropoietin and quality of life and exercise capacity of patients
receiving haemodialysis. BMJ 1990;300:573-8.
29. Eschbach JW. Erythropoietin is not a cause of access thrombosis. Semin
Dial 1993;6:180-4.
30. Muirhead N. Erythropoietin is a cause of access thrombosis. Semin
Dial 1993;6:184-8.
31. Nissenson AR, Pickett JL, Theberge DC, Brown WS, Schweitzer SV.
Brain function is better in hemodialysis (HD) patients (PTS) when hematocrit
is normalized with erythropoietin (rHuEPO). J Am Soc Nephrol
1996;7:1459. abstract.
32. Bárány P, Svendenhag J, Katzarski K, et al. Physiological effects of correcting
anemia in hemodialysis patients (HD pts) to a normal HB. J Am
Soc Nephrol 1996;7:1472. abstract.
33. Benz RL, Pressman MR, Hovick ET, Peterson DD. Relationship between
anemia of chronic renal failure (ACRF) and sleep, sleep disorders
and daytime alertness: benefits of normalizing hematocrit (the Sleepo Trial).
J Am Soc Nephrol 1996;7:1473. abstract.
34. Renal Data System. USRDS 1996 annual data report. Bethesda, Md.:
National Institute of Diabetes and Digestive and Kidney Diseases, April
1996:23.
35. Salonen JT, Nyyssonen K, Korpela H, Tuomilehto J, Seppanen R, Salonen
R. High stored iron levels are associated with excess risk of myocardial
infarction in eastern Finnish men. Circulation 1992;86:803-11.
36. Sullivan JL. Iron and the sex difference in heart disease risk. Lancet
1981;1:1293-4.
37. Tielemans CL, Lenclud CM, Wens R, Collart FE, Dratwa M. Critical
role of iron overload in the increased susceptibility of haemodialysis patients
to bacterial infections: beneficial effects of desferrioxamine. Nephrol
Dial Transplant 1989;4:883-7.
38. Hoen B, Kessler M, Hestin D, Mayeux D. Risk factors for bacterial
infections in chronic haemodialysis adult patients: a multicentre prospective
survey. Nephrol Dial Transplant 1995;10:377-81.
39. Collins A, Ebben J, Ma J. Frequent IV iron dosing is associated with
higher infectious deaths. J Am Soc Nephrol 1997;8:190A. abstract.
Downloaded from www.nejm.org at Amgen, Inc. on September 2, 2008 .
Copyright © 1998 Massachusetts Medical Society. All rights reserved.
Page 78

Members of the Canadian
Erythropoietin Study Group
are given at the end of this
paper.
Correspondence and
requests for reprints to:
Dr Andreas Laupacis,
Robarts Research Institute,
PO Box 5015, London,
Ontario N6A 5KB, Canada.
BrMed3 1990;300:573-8
BMJ VOLUME 300 3 MARCH 1990
Association between recombinant human erythropoietin and quality
of life and exercise capacity of patients receiving haemodialysis
Canadian Erythropoietin Study Group
Abstract
Objective-To determine whether recombinant
human erythropoietin improves the quality of life
and exercise capacity of anaemic patients receiving
haemodialysis.
Design-A double blind, randomised, placebo
controlled study.
Setting-Eight Canadian university haemodialysis
centres.
Patients-118 Patients receiving haemodialysis
aged 18-75 with haemoglobin concentrations

and a data clerk and was responsible for adjusting the
dose of erythropoietin, prescribing iron supplements
or transfusions, and sending haematological data to the
coordinating centre. The blinded team consisted of
nurses in the dialysis unit and our study group and all
doctors in the dialysis unit other than those in the
unblinded team; this team carried out routine clinical
care and recorded adverse reactions and other clinical
events but did not have access to the results of
haematological tests or know the dose of erythropoietin
or placebo that each patient was receiving. The nurses
in the study group administered tests to assess quality
of life and exercise capacity.
All patients were followed up for six months, and
outcome measures were assessed before randomisation
and at two, four, and six months. The two main
outcomes assessed were quality of life and exercise
capacity. We thought that quality of life would be
most effectively assessed by a questionnaire aimed
specifically at patients receiving haemodialysis. To put
the results into perspective, however, we included two
global measures of quality of life.
We thus developed a kidney disease questionnaire
using the methods suggested by Guyatt et al.' We also
used the sickness impact profile, which assesses health
state based on behaviour and has been used in the
national kidney dialysis and kidney transplant study,9
and the time trade off technique described by Churchill
et al," from which a "utility" is derived for each
patient. These three measures of quality of life are
described in more detail in the appendix.
To assess exercise capacity we used the six minute
walk test and an exercise stress test. The six minute
walk was conducted in enclosed corridors that were
between 25 and 40 m long. Patients were instructed to
walk from end to end, covering as much ground as
possible in the allotted time. We gave patients an
exercise stress test using a modified Naughton
protocol2 conducted in 11 two minute stages, with the
speed of the treadmill increasing from 1 -6 to 5 5 km/h
and the incline increasing from 0 to 16%. If the patient
was able to exercise for 22 minutes the test was
continued for a further eight minutes at a speed of
5 5 km/h and a gradual increase in the incline from 16%
to 25%. Standard encouragement was given during
both exercise tests.
We calculated the size of sample that would ensure
that a clinically important difference among the three
groups at six months would be detected with the time
trade off technique, sickness impact profile, and six
minute walk test (the only outcome measures for which
TABLE i-Characteristics ofpatients at beginning ofstudy into effects oflow and high doses of erythropoietin.
Values are means (SD) unless stated otherwise
Placebo Low ervthropoietin High ervthropoietin
group group group
(n =40) (n =40) n-38)
Age (years) 48 (16) 44 (16) 43 (15)
No of men 25 19 26
No who were anephric 6 13 10
Nowho had had a transplant 14 16 12
Duration of haemodialysis (years)* 25 (3 1) 46(47) 44 51)
No of transfusions during previous year 7-3 (8 3) 6 6 (6 8) 5 6 (7 4)
No dependent on transfusiont 19 19 11
Systolic blood pressure (mm Hg)t 143 (19) 137 (22) 143 (21)
Diastolic blood pressure (mm Hg)# 80 (12) 78 (14) 78 12))
No receiving antihypertensive drugs 10 13 12
Haemoglobin (g/l) 71 (9) 69 (10) 71 (12)
Serum iron ([tmol/l) 19 (14) 18 (9) 22 (15)
Serum ferritin (ug/l) 703 (866) 1044 (1690) 1273 (2303)
Median serum ferritin ([g/l) 318 406 345
Serum creatinine (4imolI1) 1057 (246) 1090 (260) 1076 290)
Serum urea (mmol/l) 28-0 (6 0) 27-3 (7 0) 27 8 (7 2)
Serum potassium (mmol/l) 5-0 (0 6) 5-1 (0 6) 51 (0-7)
*p=0 07 Among groups.
tPatients who had received dialysis for one or more years were considered to be dependent on transfusion if they
received six or more units of blood in the year before randomisation. Patients who had received dialyzsis for less than
one year were considered to be dependent on transfusion if they had received more than two units of blood in the
three months before randomisation.
tPlacebo group n= 37; low erythropoietin group n= 39; high ervthropoietin group n= 37.
574
Page 80
data on variance were available before the studv). " We
required 30 patients in each group (two tailed (C of 0 05
and a power of 90o). Assuming a withdrawal rate of
25% due to renal transplantation, adverse effects, and
concomitant events, this meant that we needed a
sample of 120 patients.
The outcome variables were analysed only for
patients who had completed all four evaluations
(before randomisation and at two, four, and six
months). Analysis of variance for repeated measures
was used. The main hypothesis tested was that the
mean response profiles of v . rariables evaluated at the
four time points were the same among the tl-iree
groups.4 In addition to the overall comparison among
the three treatment groups, two orthogonal contrasts
were specified to compare the patients receiving
placebo with the patients treated with erythropoietin
(both groups) and the patients in the low erythropoietin
group with those in the high erythropoietin group.
Pearson's correlation coefficient was computed
to examine the relation between the change in
haemoglobin concentration and the change in qualitv
of life and exercise capacity. Kendall's T statistic
and the corresponding p values were computed to
detect any significant associations between the
incidence of side effects and treatment with
erythropoietin.
Results
PATIENT CHARACTERISTICS
Forty patients were randomised to receive placebo,
40 low dose erythropoietin, and 38 high dose erythropoietin.
Renal failure was due to glomerulonephritis
(50 patients), pyelonephritis or interstitial nephritis
(24), polycystic kidney disease (14), renal vascular
disease (seven), chronic renal failure of unknown cause
(three), nephritis induced by drugs (two), and other
causes ( 18). Table I shows the characteristics of
patients in the three groups at the beginning of the
study; the three groups were comparable. Patients
receiving placebo had been treated by dialysis for a
shorter time than the two other groups (p=0 07).
When age, duration of dialysis, whether patients had
received a transplant or were anephric, dependence on
transfusion, and antihypertensive treatment at the
time of randomisation were examined with effect of
treatment in the analysis the results of the unadjusted
analysis were hardly affected.
WITHDRAWALS FROM STUI)Y
Nineteen patients were withdrawn during the study:
eight in the placebo group (because of transplarntation
(five), non-compliance (one), reaction to transfusion
(one), seizure and death (one)); six in the low erythropoietin
group (transplantation (two), hypertension
(one), hypertension and seizure (one), subarachnoid
haemorrhage and seizure (one), pregnancy (one)); and
five in the high erythropoietin group (transplantation
(three), hypertension (two)). Six patients were
withdrawn before the follow up at two months, and the
13 others were withdrawn before the follow up at four
months. The patient who became pregnant continued
to receive erythropoietin but had a spontaneous
miscarriage at 11- 1 2 weeks' gestation.
CHANGES IN HAEMOGLOBIN CONCENTRATION
Figure 1 shows the changes in haemoglobin
concentration in the three groups. The mean (SD)
haemoglobin concentration at six months was 74
(12) g/l in the placebo group, 102 (10) g/l in the low
erythropoietirl group, and 117 (14) g/l in the high
erythropoietin group. The mean haemoglobin
concentration reached the target range by six weeks in
patients in the low erythropoietin group and by
BMJ VOLUME 300 3 MARCH 1990

IABLE I- Scores for qualitv of liJe and exercise capacity obtained for patienits wzith end stage renal disease before and two and six months aflter start of treatment with placebo or low or
high doses of ervthropoietin
Kidney disease questionnaire:
Phvsical*
Fatigue
Relationships
Depression
Frustration
Sickness impact profilet:
Global
Physical
Psychosocial
Time trade off technique
Exercise stress test: (minutes walkedl
Six minute walk test' distance walked m
*Placebo group n= 31.
tLowest score represents best quality of life.
Placebo group
(n=32
Low ervthropoietin group
n =34)
High erythropoietin group
(n = 33)
At At At At At At
Before 2 months 6 months Before 2 months 6 months Before 2 months 6 months
tPlacebo group n= 29; low ervthropoietin group n= 31; high ers thropoietin group n= 27.
,Placebo group n= 31; low erythropoietin group n 32; high erv thropoietin group n= 30.
Significance of
difference
Among the Ervthropoietin High dose v
three z} low dose
groups placebo ervthropoletin
TABLE III-Symptoms defined by patients as being a problem associated with end stage renal disease before and two and six months after start of treatment with placebo or low or high
doses of erythropozetin
Fatigue
Decreased strength
Sleeping abnormalities
Aching legs and bones
Shortness of breath
SexualitV
100
No
of
patients
67
45
45
35
20
18
Placebo group
At At
Before 2 months 6 months Before
4 1
4 1
4 0
3 6
3-6
4-6
4 6
4.7
4.3
4.4
4.7
4-6
4 1
4-2
4.5
4.3
4.4
5 .2
Low ervthropoietin group High erythropoietin group
3-1
2-8
3 8
3 0
4.3
3 0
At At At At
2 months 6 months Before 2 months 6 months
4 3
4-0
4-0
4 2
5.4
4 2
5 4
5 3
4 0
4.4
5.9
4 2
High
erythropoietin
group
. Low
erythropoietin
group
37 4.9 50
4-0 5 3 5 3
4.3 5 3 56
4-0 5-0 5.7
4-2 5 8 5-8
3-0 3.9 5.0
Significance of diffeience
Among the Erythropoietin High dose z'
three Z) low dose
groups placebo ervthropoietin

TABLE Iv-Mean blood pressure (mm Hg) in the three treatment groups before (time 0) and two, four, and six months after start of treatment
Placebo group (n = 29) Low ervthropoietin group n =31) High erythropoietin group (n =32) Significance of difference
Among the Erythropoietin High dose V lOw dose
0 2 4 6 0 2 4 6 0 2 4 6 three groups v placebo erythropoietin
Sy'stolic 147 144 142 143 137 138 140 137 144 146 141 144 NS NS NS
Diastolic 80 78 77 79 76 79 80 78 78 84 83 85 0-023

phosphorus, calcium, urea, and creatinine or white
cell count among the groups throughout the study. At
six months the platelet count had fallen by a mean (SD)
of 5 (39) x 109/1 in the placebo group and increased by
25 (64)x 109/l in the low erythropoietin group and
23 (47) x 109/1 in the high erythropoietin group
(p=0005, placebo v both erythropoietin groups
throughout study).
BLINDING
At the end of the study all patients were asked
whether they thought they had received erythropoietin
or placebo. Nineteen patients in the placebo group
and 52 patients receiving erythropoietin correctly
identified their treatment. Six of the placebo group
thought that they had received erythropoietin while
two of the patients receiving erythropoietin believed
that they had been given placebo. The other patients
could not say which they had received.
Discussion
Treatment with erythropoietin corrected the
anaemia associated with end stage renal disease in
patients being treated by haemodialysis. Generally,
patients receiving erythropoietin were much less
fatigued, complained of less severe physical symptoms,
and had increased exercise tolerance compared with
patients given placebo. There was no significant
difference in quality of life or exercise tolerance
between patients receiving low and high doses of
erythropoietin. A significant correlation was found
between changes in haemoglobin concentration and
changes in quality of life and exercise capacity, but
the correlation coefficient was modest. Although
erythropoietin caused a significant improvement in
quality of life and exercise capacity, many other
factors, such as uraemia, bone disease, and the
psychological burdens of haemodialysis, also adversely
affect the wellbeing of such patients, and these factors
may not be improved by erythropoietin.
The effect of erythropoietin on psychosocial
function was less impressive than that on fatigue,
physical symptoms, and exercise tolerance. Some
significant (but moderate) improvements were seen in
responses to the kidney disease questionnaire but not
in the sickness impact profile. As the kidney disease
questionnaire was developed for patients with renal
failure it is not surprising that it was more responsive in
our patients.
The two measures of overall quality of life yielded
different results, an improved quality of life being
detected in patients treated with erythropoietin by the
sickness impact profile but not by the time trade
off technique. The time trade off technique allows
patients to give their own weighting to the physical,
social, and emotional factors that affect their quality of
life, while the sickness impact profile applies fixed
weights that may not precisely have reflected the
concerns of our patients. The time trade off technique,
which asks patients how many years of their current
health state they would be willing to give up to receive
erythropoietin is a rigorous test; the fact that the time
trade off score did not change in patients taking
erythropoietin probably indicates that despite the drug
many of the patients receiving dialysis still had several
adverse effects from their disease. The time trade off
technique has been successfully used to detect changes
in quality of life,'6 and complete insensitivity of the test
is therefore unlikely to be the reason for the fairly
constant scores in our study.
We examined several outcome measures, and thus
multiple testing may possibly have caused some results
to appear significant by chance. We expected the effect
of erythropoietin on fatigue, however, to be the largest,
Page 83
and the high levels of significance we obtained make it
extremely unlikely that the results occurred by chance
alone.
It is difficult to recommend confidently a target
haemoglobin concentration for patients treated with
erythropoietin. In general, the patients' quality of life
and exercise capacity rose with increasing haemoglobin
concentration, but patients in the high erythropoietin
group had a greater incidence of hypertension (and
perhaps clotting of the vascular access) than patients in
the other groups. Decisions on each patient's target
haemoglobin concentration should be made clinically
taking into account not only increased quality of life
and exercise capacity but also the patient's likelihood
of developing hypertension or thrombosis.
The ultimate clinical role of erythropoietin in the
treatment of anaemic patients receiving haemodialysis
will depend on the patient's need for transfusion,
the ability of erythropoietin to prevent anti-HLA
sensitisation, the cost of the drug, and the drug's
efficacy and side effects in other populations receiving
haemodialysis.
Appendix
KIDNEY DISEASE QUESTIONNAIRE
The kidney disease questionnaire is specific for
patients with end stage renal disease and was developed
from the methods described by Guyatt et al.8 We
interviewed patients receiving haemodialysis and
health care staff and reviewed existing quality of life
variables. This generated a list of 130 items that could
potentially have an adverse affect on the quality of life
of the patients. Fifty patients receiving haemodialysis
were then asked to rank these items in terms of their
importance.
The items that were ranked most frequently and
judged to be most important by the patients were
included in the questionnaire. The final questionnaire
consisted of 26 questions divided into five sections:
physical symptoms, fatigue, depression, relationships,
and frustration. In the section on physical symptoms
each patient was asked to identify those physical
symtoms or problems that most affected his or her life,
and these were used for that patient throughout the
study.
All questions were scored on a seven point Likert
scale (7 = no problem, 1= a severe problem). For
example, a question from the fatigue section, "How
often in the past two weeks have you felt low in
energy?" was accompanied by a choice of answers: (1)
All the time; (2) most of the time; (3) a good bit of the
time; (4) some of the time; (5) a little of the time;
(6) hardly any of the time; and (7) none of the time.
A change in mean score of 0 5 in each section
represented a minimal, clinically important difference,
and a mean change of 1 0 represented a large clinical
change. 17
SICKNESS IMPACT PROFILE
The sickness impact profile is a questionnaire based
on behaviour that contains 136 statements about
dysfunction related to health in 12 types of activity:
sleep and rest, eating, work, home management,
recreation and pastimes, ambulation, mobility, body
care and movement, social interaction, alertness,
emotional behaviour, and communication.9 Some of
these can be aggregated to give a physical score (body
care and movement, mobility, and ambulation) and
a psychosocial score (emotional behaviour, social
interaction, ambulation, and communication). An
overall score combining all questions can also be
derived.
When answering the questionnaire the patients were
asked to think about themselves that day and to
BMJ VOLUME 300 3 MARCH 1990 577

Department of Community
Medicine and General
Practice, Radcliffe
Infirmary, Oxford
OX2 6HE
E M I Williams, MB, senior
registrar
L Jones, BA, computer scientist
M P Vessey, FFCM, professor
K McPherson, PHD, lecturer
Correspondence to: Dr
E M I Williams, Oxford
Regional Health Authority,
Old Road, Headington,
Oxford OX3 7LF.
Br,1edj 1990;300:578-9
determine which statements described them and were
related to their health. The patient could answer only
yes or no to each question. An example of a question
about sleep and rest is, "I spend much of the day lying
down in order to rest."
TIME TRADE OFF TECHNIQUE
The time trade off technique is used to derive
a "utility," a score between 0 and 1, in which
1 represents perfect health and 0 a state in which the
patient is indifferent between life and death. " With the
help of visual aids patients were asked how many years
of their current health they would be willing to forgo to
achieve perfect health. For example, if a 36 year old
patient stated that she was unable to choose between
four years of perfect health and 40 years of her current
health the utility of her current health state was
4/40=0 10 (she would be willing to give up 36 years of
her current health to achieve four years of perfect
health).
The Canadian Erythropoietin Study Group comprises the
following groups. Executive committee: David Churchill,
Paul Keown (chairman), Andreas Laupacis, Norman
Muirhead, Dalice Sim, David Slaughter. Management
committee: Paul Keown, Andreas Laupacis (chairman),
Norman Muirhead, Beth Sarazin, Denise Short, Dalice Sim,
David Slaughter, Cindy Wong. Study centres: Victoria
General Hospital, Halifax, Nova Scotia (Allan Cohen,* David
Hirsch, Kailash Jindal); Royal Victoria Hospital, Montreal,
Quebec (Paul Barre,* Andrew Gonda, Tom Hutchinson,
Sarah Prichard, Denis Roy); Ottawa Civic Hospital and
Ottawa General Hospital, Ottawa, Ontario (Andrew
Lazarovits,* Steven Nadler,* Gerald Posen, Eli Rabin); St
Joseph's Hospital, Hamilton, Ontario (Colin Barnes, David
Churchill,* Sandra Donnelly, David Ludwin, Kinsey
Smith); University Hospital and Victoria Hospital, London,
Ontario, and Laurentian Hospital, Sudbury, Ontario
(William Fay, David Hollomby, Tony Jevnikar, Paul
Keown, Norman Muirhead*); Health Sciences Centre and St
Boniface Hospital, Winnipeg, Manitoba (Keevin Bernstein,
Adrian Fine, John McKenzie,* Brian Penner, Brent
Schacler, Ashley Thomson); Foothills Hospital, Calgary,
Alberta (Ellen Burgess, Allan Jones, Henry Mandin*); St
Paul's and Vancouver General Hospitals, Vancouver,
Short term increase in risk of
breast cancer associated with
full term pregnancy
E M I Williams, L Jones, M P Vessey,
K McPherson
Women who have their first full term pregnancy after
age 35 are at higher risk of breast cancer than
nulliparous women,' and the proportion of young
women with breast cancer who have had children is
higher than expected.7 These and other findings
suggest that women who have full term pregnancies
have a transiently increased risk of breast cancer that is
followed by long term protection.) We applied the
analysis described by Bruzzi et al3 to data obtained
from an investigation which explored the relation
between lifestyle factors and risk of breast cancer.4
Patients, methods, and results
Briefly, we recruited 1996 married women aged 25-
59 between 1980 and 1984 from seven hospitals. They
comprised 998 patients with newly diagnosed breast
cancer that had been confirmed histologically and 998
controls, who had been admitted electively with
conditions not originally related to breast cancer.
British Columbia (Anthony Chiu, Gershon Growe, David
Landsberg, John Price, John Shepherd, Roger Sutton,*
Linda Vickers, Ronald Werb). *=Principal investigators.
We thank the study nurses, data coordinators, and
pharmacists at each of the centres; the support staff of Robarts
Research Institute; Ulrike Kuprath, Patricia Laplante,
Angela Mongul, David Slaughter, and other staff at Ortho
Pharmaceutical (Canada) Ltd; and the patients. We also
thank Janice Coffey for secretarial help.
1 Eschbach JW, Egrie JC, Downing MR, Browne JK, Adamson JW
Correction of thc anicmia of end-stage renal disease with recombinant human
crythropoictin. Engl.ViMed 1987;316:73-8.
2 Winearls C(G, P'ippard AIj, t)owning MR, Oliver DO, Reid C, Cotes PMN.
Ef'fect of htuman erythropoietin derived from recombinant DNA on the
anaemia of patients tnaintaitted by chronic haemodialysis. Latncet
1986;ii: 1 175-7.
3 Canadian Ersthropoietin Study Group. A prospective, randomized,
douLble-blind stuLds of recombittant erythropoietin (r-Hu-EPO)0 in chronic
hemodialv sis [Abstract]. Kidsne Iti 1988;33:218A.
4 Eschbach JVI. 'rhc ancmia ot' chronic renal failure: pathophvsiologv and the
effect of recombinant crsthropoietin. Kidnicesv Int 1989;35:134-48.
5 Ltindin AP. Qualits ot' life: subjective and objectise improvements with
recombinant htuman crs thropoictin therapy. Semin Nephrol 1989;9:22-9.
6 Schacfer RAM, Kokot F, Wlcrnze H, Geiger H, Heidland A. Impros-ed sexual
funictioni in hcmodialssis patients on rccombinant erythropoietiit: a possible
role for prolactin. Clin .\' phrol 1989;31:1-5.
7 Guvatt GH, Sullisan M1J, Fhompson P1, ct al. The 6-minute walk: a new
measure of excrcise capacits in patients with chronic hcart failure. Can Med
Assocj 1985;132:919-23.
8 Guyatt GH, Bombardier C, I'ugwell P. Measuring disease-specific quality of
life in clinical trials. Can Med .Assoc7 1986;134:889-95.
9 Bergner M, Bobbitt RA, Carter WB, Gilson BS. The sickniess impact profile:
deselopment anid final res-ision of a health status measure. Med Care
1981;19:787-805.
10 Hart LG, Evans RW. The functional status of ESRD patients as measured by
the sickness impact profile. ] Chronic Dis 1987;40(suppl 1): I 17-30S.
11 Churchill DN, Torrance GW, Taylor DX', et al. ieasurement of quality of'lifc
in end-stage renal disease: the time trade-off approach. Clin Invest Med
1987;10: 14-20.
12 Naughton J, Sevelius G, Balke B. Physiological responses otf' normal and
pathological subjects to a modified work capacity test. J Sports Med P/s's
Fitness 1963;3:201-7.
13 Schwartz D, Flamant R, Lellouch J. Clinical trio/ls. New York: Academic
Press, 1980:137.
14 Winer BS. Statistitcal prnciples in experimental design. New York: MSc(Graw-
Hill, 1971:518-39.
15 Gibbons J[). NnparantietrIc statistical inference. Ncw Ytork: MIcGraw-Hill,
1971:2209-26.
16 Mohiide EA, Torrance GW, Streitter D)t, ltrittgle l)Ni, Gilbert R. MNeasuring
the wcllbeing of family caregis-crs usitag thc timc trade-otff technique.
Cl/inical Epidemiots/s 1988;41:475-82.
17 Jacschkc R, Singer J, Guyatt G, Health status mcasuremcnt: ascertaining the
minimal clinically important diff'ercncc. CAontrolled C/in Irials 1989;10:407-
15.
(Accepte7d 19 IJecenther /989
Page 84
Patients and controls were matched for admitting
hospital and within five year age groups but not for
parity. The present analysis aimed at detecting any
increase in the risk of breast cancer shortly after a full
term pregnancy. The interval between the date of
diagnosis of breast cancer and the last term birth was
studied. As this is related to age, age at first term birth,
and parity these variables were adjusted for in the
analysis.
Only women under 50 with two or more children
were included to accord with the analysis by Bruzzi et
al,' and this resulted in the study becoming unmatched
(422 cases and 447 controls). The generalised interactive
modelling (GLIM) package was used to estimate
the maximum likelihood of effects. Graduated levels of
exposure were assessed by linear trend tests.
The two groups were comparable in terms of centre
of recruitment, which was ignored in subsequent
analyses. An increased risk of breast cancer was
associated with decreasing interval since last term birth
(p=0-021), increasing age at first term birth (p=0 006),
and decreasing parity (p=0 -002) (table). When women
aged under 40 and 40-49 were considered separately
the trends were broadly similar, although not all were
significant.
Comment
Our results suggest that a transient increase in the
risk of breast cancer occurs after full term pregnancy.
578 BMJ VOLUME 300 3 MARCH 1990

Page 85

Page 86

Page 87

Page 88

Page 89

Page 90

Page 91

Page 92

Page 93

Nephrol Dial Transplant (2009) 24: 3138–3143
doi: 10.1093/ndt/gfp213
Advance Access publication 18 May 2009
Blood transfusion use in non-dialysis-dependent chronic kidney
disease patients aged 65 years and older
Hassan N. Ibrahim 1,2 , Areef Ishani 1,2,3 , Haifeng Guo 1 and David T. Gilbertson 1
1 Chronic Disease Research Group, Minneapolis Medical Research Foundation, 2 Department of Medicine, University of Minnesota
and 3 Section of Nephrology, Department of Medicine, Minneapolis Veterans Affairs Medical Center, Minneapolis, MN, USA
Correspondence and offprint requests to: Hassan N. Ibrahim; E-mail: ibrah007@umn.edu
Abstract
Background. Erythropoiesis stimulating agents (ESA)
have alleviated the need for blood transfusions in dialysis
patients. Their impact on transfusion frequency in
elderly chronic kidney disease (CKD) patients aged
65 years and older with non-dialysis-dependent CKD has
not been studied.
Methods. We conducted a retrospective cohort study of
Medicare beneficiaries with CKD, point prevalent as of
1 January of each calendar year 1992–2004 (n = 301 000),
and a concurrent group of beneficiaries without CKD (n =
15 772 039). During the entry year, we used administrative
claim data to assemble a comorbidity profile for each participant.
Transfusion event rates, ESA use and parenteral
iron use were of primary interest.
Results. CKD patients were at least four times more likely
than non-CKD patients to receive transfusions. Transfusion
rates adjusted for case mix fell from 194.2 transfusions per
1000 patient-years in 1992 to 112.2 in 2004. The decline
in transfusions was highest for anaemic CKD patients receiving
ESAs, which were given to 0.8% of CKD patients
in 1992 and to 7.5% in 2004.
Conclusions. Transfusion use in non-dialysis-dependent
CKD patients has decreased considerably but continues
to be common. ESA use and possibly changed attitudes
towards transfusion use explain most of the reduction
noted.
Keywords: chronic kidney disease; erythropoiesis stimulating agents;
transfusions.
Introduction
Anaemia is a classic consequence of reduced renal function,
and its prevalence is almost universal by the time
chronic kidney disease (CKD) progresses to end-stage renal
disease (ESRD) (1–2). The introduction of erythropoiesis
stimulating agents (ESAs) has improved the quality of life
for CKD patients, including those on dialysis (3–6). We
recently showed reduced transfusion frequency in a large
cohort of dialysis patients studied between 1992 and 2004
(7). Surprisingly, little is known regarding the effect of ESA
use on the frequency of blood transfusions in CKD patients
who are not on dialysis.
Therefore, we compared the transfusion experience of a
large cohort of Medicare beneficiaries with non-dialysisdependent
CKD to the experience of beneficiaries without
CKD. We also compared ESA use and parenteral iron use in
beneficiaries with and without CKD. We tested the hypothesis
that ESA use is responsible for most of the observed
reduction in transfusion frequency in elderly anaemic CKD
patients.
Subjects and methods
C○ The Author [2009]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
For Permissions, please e-mail: journals.permissions@oxfordjournals.org
Page 94
Study cohorts
We studied 301 000 general Medicare beneficiaries who entered the Part
A and Part B programmes from 1992 through 2004. Medicare Part A reimburses
health care providers for services to Medicare beneficiaries during
hospitalizations, and it covers facility costs. Medicare Part B covers services
provided by hospital outpatient departments, rural health clinics,
dialysis facilities, outpatient rehabilitation services and community mental
health centres. Most citizens or permanent residents of the United
States aged 65 years or older are eligible for Part A coverage if they are
eligible for Social Security benefits. Before age 65 years, Medicare coverage
is available only to people with disabilities or with ESRD. Anyone
eligible for Medicare Part A can enrol in Part B by paying a monthly
premium.
Included patients entered the Medicare programme before 1 January
of each year, were aged 65 years or older at the beginning of each year,
survived and remained in the programme through 31 December of each
year and had non-dialysis-dependent CKD diagnosed in each year (entry
year). Henceforth in this report, ‘CKD’ means ‘non-dialysis-dependent
CKD’. Patients enrolled in health maintenance organizations, with Medicare
as secondary payer, diagnosed with ESRD, or not residing in the
50 US states, the District of Columbia, Puerto Rico or the Territories
during each year were excluded. Patients with evidence of cancer or
blood dyscrasias during the 12-month entry period for each year were
also excluded, as these patients can be hyporesponsive to ESA therapy
and may require multiple transfusions not directly related to anaemia
of CKD. The relevant exclusions with International Classification of
Diseases, Ninth Edition, Clinical Modification (ICD-9-CM) codes are
shown in Table 1. Patients who did not survive the entry period were also
excluded.
CKD and comorbid conditions were defined during the 12-month entry
period. Conditions were defined by the presence of one inpatient Part A
claim or two outpatient Part B or one Part A and one Part B claims.
ICD-9-CM codes used to define CKD and relevant comorbid conditions
used to describe the cohort are listed in Table 1.

Transfusions and chronic kidney disease 3139
Table 1. International Classification of Diseases, Ninth Edition, Clinical
Modification (ICD-9-CM) codes used to define chronic kidney disease
and identify excluded and comorbid conditions
Condition ICD-9-CM codes
CKD 016.0, 095.4, 189.0, 189.9, 223.0,
236.91, 250.4, 271.4, 274.1, 283.11,
403.x1, 404.x2, 404.x3, 440.1, 442.1,
447.3, 572.4, 580–588, 591, 642.1,
646.2, 753.12–753.17, 753.19, 753.2
and 794.4
Excluded conditions
Cancer 140.xx–172.xx, 174.xx–208.xx,
230.xx–231.xx
Haemolytic anaemias 282.x, 283.x
Aplastic anaemia 284.x
Coagulation defects 286.x
Myeloma 203.0x
Myelofibrosis 289.89
Myelodysplasia 238.7
Comorbid conditions
ASHD 410.xx–414.xx, V45.18–V45.82
CHF 398.91, 422.xx, 425.xx, 428.xx, 402.x1,
404.x1, 404.x3, V42.1
CVA 430.xx–438.xx
PVD 440.xx–444.xx, 447.xx, 451.xx-453.xx,
557.xx
Other CVD 420.xx–421.xx, 423.xx–424.xx, 429.xx,
785.0–785.3, V42.2, V43.3
COPD 491.xx–494.xx, 496.xx, 510.xx
GI disorders 456.0–456.2, 530.7, 531.xx–534.xx,
569.84, 569.85, 578
Liver disease 570.xx, 571.xx, 572.1, 572.4,
573.1–573.3, V42.7
Dysrhythmia 426.xx–427.xx, V45.0, V53.30
Diabetes 250.xx, 357.2, 362.0x, 366.41
Hypertension 362.11, 401.x–405.x, 437.2
Anaemia 280.xx–285.xx
ASHD, atherosclerotic heart disease; CHF, congestive heart failure; CKD,
chronic kidney disease; CVA, cerebrovascular accident; PVD, peripheral
vascular disease; CVD, cardiovascular disease; COPD, chronic obstructive
pulmonary disease; GI, gastrointestinal.
Based on claim source (inpatient or outpatient/Part B claims), CKD
and comorbid conditions were further classified as follows: inpatient
claim source if the inpatient claim occurred before the second Part B
or outpatient claim, or in the absence of a Part B or outpatient claim;
outpatient claim source if the second outpatient or Part B claim occurred
before an inpatient claim or in the absence of an inpatient claim. Transfusions
billed by outpatient facilities and those administered during inpatient
hospital stays were obtained from Part A files using code files for both
whole and packed cell transfusions. Those obtained from Part B files included
whole and packed red blood cell transfusions, regardless of whether
these were leukocyte reduced, irradiated or deglycerolized. ESA use was
defined during the entry period by Healthcare Common Procedure Coding
System (HCPCS) codes. Anaemia was defined the same way that
CKD was defined, using ICD-9-CM codes 280.xx–285.xx, encompassing
all haemolytic, non-haemolytic and bone-marrow-failure-related anaemia
(see Table 1).
A non-CKD cohort including 15 772 039 patients was also studied from
1992 through 2004, applying similar methodology and exclusions.
Analysis
Categorical variables are presented as percentages. Patients were followed
from 1 January of each year to 31 December of the next year for blood
transfusions and were censored at death, ESRD diagnosis and payer status
change. Raw transfusion rates were calculated by the total number of
transfusions over total follow-up time. Adjusted transfusion rates were
calculated by the Poisson model. Seven models were fitted by adjusting
for the following covariates:
Model 1: age, sex, race.
Model 2: age, sex, race and ESA use in the entry year.
Model 3: age, sex, race and ESA use and hospital stays in the entry
year.
Model 4: age, sex, race and comorbid conditions in the entry year.
Model 5: age, sex, race and ESA use, hospital stays and comorbid
conditions in the entry year.
Model 6: age, sex, race and ESA use, hospital stays and CKD and
comorbid condition claim source (inpatient or outpatient) in the entry
year. Adjustment for some of these comorbid conditions was done
due to their potential to cause anaemia directly (e.g. gastrointestinal
disorders) or to be associated with anaemia of CKD. The 1992 cohort
was used as the reference group for these six models. An additional
model, model 7, was similar to model 6 but used 2004 as the reference
year. As the adjusted models gave comparable results, only results
from model 7 are presented.
Results
Page 95
General Characteristics of CKD and Non-CKD cohorts
We studied 301 000 Medicare patients with CKD, 1992–
2004. The number of patients rose steadily, from 14 057
in 1992 to 42 225 in 2004 (Table 2), and the increase was
uniform across all age groups (ages 65–74, 75–84 and ≥85
years). In 2004, women constituted 55.2% of the CKD cohort
and white patients 82.6%. Approximately a third of
CKD patients were hospitalized at least once for an average
of 1–7 days during the observation period. Interestingly,
the proportion of CKD patients hospitalized for 8
or more days fell from 35.8% in 1992 to 27.2% in 2004.
Claims related to ASHD were present for approximately
half the CKD cohort during the observation period, claims
related to CHF for 40% and claims related to CVA/TIA
for 20%. The proportions of patients with these three conditions
changed little from 1992 to 2004. The prevalence
of diabetes, PVD, hypertension and COPD rose steadily
over the study period. In 2004, the reported diabetes prevalence
was 48.4% and the hypertension prevalence 89.0%.
The anaemia prevalence increased from 31.1% in 1992 to
44.0% in 2004, paralleled by a steady increase in ESA use
by almost 10-fold from 0.8% in 1992 to 7.5% in 2004.
We also found a steady, though small, increase in parenteral
iron use in the entry year. In 1992, only 0.1% of
CKD patients received parenteral iron, rising to 1.0% in
2004.
The non-CKD cohort included 15 772 039 Medicare patients
(Table 3). In this cohort, 13.2% were aged older than
85 years compared with 20.6% in the CKD cohort. The
proportion of African Americans was also lower, 7.3%
compared with 12.9% in the CKD cohort. A substantially
lower proportion (17.5%) than in the CKD cohort
were hospitalized in the entry period, and the comorbidity
burden, especially ASHD, CHF, CVA/TIA, diabetes
and hypertension was lower. The anaemia prevalence was
also much lower, at 9.0% versus 39.5% in the CKD cohort.
ESA and iron use was negligible in the non-CKD
cohort.

3140 H. N. Ibrahim et al.
Table 2. Chronic kidney disease patient characteristics, 1992–2004
Cohort year
All 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
N 301 000 14 057 14 381 15 539 15 976 17 147 18 315 19 588 21 209 24 187 28 252 32 748 37 376 42 225
Age (years)
65–74 36.5 37.9 38.3 38.4 38.3 38.1 37.0 36.9 36.5 36.4 36.1 35.4 35.5 35.1
75–84 42.9 41.6 41.4 40.6 41.4 41.6 42.5 42.1 42.6 43.1 43.9 44.1 44.1 44.0
≥85 20.6 20.5 20.3 21.0 20.4 20.2 20.5 21.0 20.9 20.5 20.0 20.5 20.5 20.9
Women 55.1 54.9 54.6 55.7 55.5 55.3 55.6 55.2 54.8 54.9 55.2 55.1 54.5 55.2
White 82.1 82.2 81.6 81.3 81.6 81.6 81.7 81.7 81.9 82.2 82.0 82.6 82.7 82.6
African American 12.9 14.1 14.7 14.4 14.0 13.6 13.5 13.0 12.8 12.4 12.6 12.3 12.2 12.1
Other race 5.0 3.6 3.9 4.3 4.4 4.8 4.9 5.4 5.3 5.4 5.4 5.2 5.1 5.3
ASHD 47.8 44.5 44.0 46.5 47.3 47.0 47.6 48.1 48.5 48.6 49.0 49.0 48.5 48.2
CHF 41.7 42.1 42.5 43.5 43.8 43.1 43.3 43.5 42.9 42.5 41.7 40.7 39.6 39.0
CVA/TIA 21.3 21.2 20.4 21.4 23.2 22.7 22.2 21.7 21.6 21.5 21.4 21.2 20.4 20.2
Dysrhythmia 33.8 30.5 30.4 33.8 33.9 33.0 33.7 33.9 33.8 34.0 34.1 34.7 34.7 34.3
Other cardiac 29.3 26.0 26.3 28.8 29.4 29.1 30.0 29.7 29.6 29.4 29.7 30.1 29.8 29.6
Diabetes 44.3 35.6 36.8 37.5 39.3 40.6 42.0 43.9 44.9 45.9 46.9 47.7 47.5 48.4
PVD 31.0 25.6 26.6 29.2 30.3 31.2 31.0 30.6 31.6 31.3 32.1 32.0 32.2 32.2
Hypertension 80.4 65.3 67.4 70.5 72.7 74.5 76.2 77.6 79.7 82.1 84.2 85.4 87.2 89.0
COPD 24.4 21.6 22.4 23.4 24.2 24.9 24.6 23.9 24.5 24.7 24.8 24.8 25.0 25.3
Hospital stay, entry year
No 38.9 30.4 31.7 31.6 32.7 35.2 37.3 38.3 39.6 39.5 40.3 42.1 43.1 43.8
1–7 days 28.7 33.7 35.5 25.3 27.2 27.2 27.2 28.2 27.3 27.6 29.2 28.6 28.9 29.1
≥8 days 32.4 35.8 32.8 43.1 40.2 37.6 35.6 33.5 33.2 32.9 30.5 29.4 28.0 27.2
Anaemia 39.5 31.1 32.2 35.0 35.0 35.7 37.6 39.1 39.2 39.7 41.0 42.9 43.1 44.0
ESA use, entry year 3.9 0.8 0.9 1.3 1.8 1.8 2.1 2.5 3.1 3.8 4.5 5.0 6.3 7.5
Iron use, entry year 0.5 0.1 0.2 0.1 0.2 0.2 0.2 0.3 0.3 0.3 0.5 0.6 0.7 1.0
Results are expressed as percentage of the cohort in each year.
ASHD, atherosclerotic heart disease; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; CVA/TIA, cerebrovascular accident/transient
ischaemic attack; PVD, peripheral vascular disease.
Transfusion rates
Raw transfusion rates for CKD patients fell from 139.8 per
1000 patient-years in 1992 to 112.2 in 2004 (Table 4), a
decline that was observed in both inpatient and outpatient
claims. Results from the adjusted models echoed the decline
seen with raw rates. In the most comprehensive model that
adjusted for age, sex, race, ESA use, hospital stay and CKD
and comorbid condition claim source, the transfusion rate
fell from 194.2 transfusions per 1000 patient-years in 1992
to 112.2 in 2004. Non-CKD patients were four to five times
less likely than CKD patients to receive transfusions, and
these rates also decreased from 46.5 transfusions per 1000
patient-years in 1992 to 25.7 in 2004 (Table 4).
ESA use and transfusions
In 1992, 0.8% of CKD patients received ESAs, rising to
7.5% in 2004 (Table 5). The raw transfusion rate for CKD
patients who received ESAs was 538.6 transfusions per
1000 patient-years in 1992 and 198.6 in 2004. In contrast,
for CKD patients who did not receive ESAs, the raw transfusion
rate was 137.4 transfusions per 1000 patient-years in
1992 and 105.5 in 2004. Adjusted rates followed a similar
pattern, with most of the decline in transfusions occurring
for patients who received ESAs in the entry period. A similar
analysis of non-CKD patients showed no ESA use in
1992, 1993 or 1994; 0.3% of the population received ESAs
in 1995 and 0.7% in 2004. In 1992, the raw transfusion rate
for non-CKD patients receiving ESAs was 2066.9 transfu-
Page 96
sions per 1000 patient-years, declining to 262.1 by 2004.
For non-CKD patients with no ESA use in the entry period,
the rate changed minimally from 1992 to 2004. The adjusted
models, particularly for non-CKD patients receiving
ESAs, revealed similar results. Expressed differently, we
found that relative risk of receiving a transfusion for patients
receiving ESAs decreased from 2.47 (95% confidence interval
1.78–3.41) in 1992 to 1.35 (95% CI 1.22–1.48) in
2004. This reduction was observed for both inpatient and
outpatient transfusion risk (data not shown).
We next assessed transfusion exposure according to the
presence of anaemia and ESA use. Comparing transfusion
rates for CKD patients with and without anaemia in the
same entry period, we found that adjusted transfusion rates
for CKD patients with anaemia fell from 199.1 transfusions
per 1000 patient-years in 1992 to 129.3 in 2004. For CKD
patients without anaemia in the entry period, rates fell from
101.2 transfusions per 1000 patient-years in 1992 to 63.1 in
2004. Assessing raw transfusion rates for CKD patients by
both presence of anaemia and ESA use in the entry period,
we found that the rates for CKD patients with anaemia
who received ESAs in the entry period fell from 572.9
transfusions per 1000 patient-years in 1992 to 200.1 in
2004 (Table 6). For CKD patients with anaemia but no
prior ESA use, the rates fell from 249.2 transfusions per
1000 patient-years in 1992 to 174.7 in 2004. The rates for
non-anaemic CKD patients who received ESAs fell from
313.3 transfusions per 1000 patient-years in 1992 to 113.0
in 2004. Transfusion rates for non-anaemic CKD patients
followed a similar trend.

Table 3. Non-chronic kidney disease patient characteristics, 1992–2004
Cohort year
All 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
N 15 772 039 1 275 841 1 268 213 1 281 813 1 258 494 1 220 702 1 186 056 1 153 571 1 142 879 1 146 071 1 176 330 1 208 940 1 225 370 1 227 759
Age (years)
65–74 49.1 52.6 52.4 51.9 51.0 49.9 49.0 48.1 47.3 46.9 47.0 47.0 47.1 47.3
75–84 37.7 35.6 35.7 35.8 36.4 37.2 37.7 38.3 38.8 39.1 39.0 39.0 38.9 38.6
≥85 13.2 11.9 11.9 12.2 12.6 12.9 13.3 13.7 13.9 14.0 14.0 13.9 14.1 14.1
Women 60.4 60.9 60.4 60.8 60.8 60.6 60.8 60.8 60.6 60.4 60.1 59.7 59.5 59.3
White 88.8 88.6 89.5 89.1 89.1 89.1 89.0 88.9 88.8 88.7 88.5 88.4 88.2 88.2
African American 7.3 7.3 7.4 7.4 7.4 7.2 7.2 7.2 7.2 7.2 7.3 7.4 7.4 7.4
Other race 3.9 4.1 3.1 3.4 3.6 3.7 3.9 3.9 4.0 4.1 4.2 4.3 4.4 4.5
ASHD 17.5 15.4 15.6 16.2 16.7 17.3 17.8 17.9 18.0 18.2 18.4 18.5 18.6 18.8
CHF 8.5 7.2 7.5 7.8 8.1 8.6 9.0 9.2 9.1 9.0 8.9 8.7 8.6 8.5
CVA/TIA 7.3 6.0 6.2 6.5 7.0 7.5 7.7 7.6 7.6 7.7 7.7 7.7 7.7 7.6
Dysrhythmia 12.2 9.5 9.8 10.5 11.1 11.8 12.5 12.8 12.9 13.2 13.4 13.6 13.8 14.0
Other cardiac 9.1 7.1 7.3 7.8 8.2 8.8 9.3 9.4 9.6 9.7 9.9 10.3 10.4 10.6
Diabetes 14.5 11.2 11.6 12.1 12.5 13.0 13.8 14.4 15.0 15.7 16.4 17.1 17.7 18.5
PVD 8.2 6.6 6.9 7.5 7.8 8.3 8.4 8.5 8.4 8.5 8.7 8.9 9.1 9.4
Hypertension 40.5 29.9 31.4 33.2 34.8 36.6 38.2 40.0 41.7 44.0 46.3 48.4 50.6 52.8
COPD 10.2 8.4 8.9 9.1 9.5 9.9 10.2 10.4 10.7 10.6 10.9 11.1 11.2 11.4
Hospital stay, entry year
No 82.5 83.0 83.0 82.9 82.7 82.6 82.4 82.3 82.1 82.1 82.0 82.3 82.5 82.8
1–7 days 11.6 10.8 11.2 10.1 10.7 11.2 11.5 11.9 12.2 12.1 12.6 12.4 12.4 12.3
≥ 8 days 5.8 6.2 5.8 7.0 6.6 6.3 6.0 5.8 5.8 5.8 5.4 5.3 5.1 4.9
Anaemia 9.0 6.3 6.6 8.1 7.4 8.1 9.1 9.7 9.9 10.0 10.4 10.9 10.9 11.3
ESA use, entry year 0.3 0.0 0.0 0.0 0.3 0.2 0.1 0.2 0.3 0.4 0.5 0.5 0.6 0.7
Iron use, entry year 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1
Results are expressed as percentage of the cohort in each year.
ASHD, atherosclerotic heart disease; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; CVA/TIA, cerebrovascular accident/transient ischaemic attack; PVD, peripheral vascular
disease.
Transfusions and chronic kidney disease 3141
Page 97

3142 H. N. Ibrahim et al.
Discussion
Our data demonstrate that transfusion rates for non-dialysisdependent
CKD patients are 4- to 5-fold higher than rates
for non-CKD patients. Transfusion rates declined markedly
between 1992 and 2004, with the largest decline for anaemic
CKD patients who received ESAs. Paralleling the decrease
in blood transfusions, increases in the frequency of ESA
use and, to a lesser extent, parenteral iron use were noted.
CKD patients who received transfusions were more likely
to receive ESAs, reflecting a high comorbidity burden.
While blood transfusion use has declined, transfusions
remain frequent. Undoubtedly, ESA use accounts for a
proportion of the reduction; however, transfusion avoidance
during the study period, as suggested by the decline in
transfusions in the non-CKD population, may play a role.
Transfusion rates in 2004 of 200.1 per 1000 patient-years
Table 4. Transfusion rates a for chronic kidney disease and non-chronic
kidney disease patients
Raw rates Adjusted rates
Year Non-CKD CKD Non-CKD CKD
1992 26.9 139.8 46.5 194.2
1993 25.9 132.3 41.8 199.0
1994 24.7 128.8 37.4 152.7
1995 25.4 126.1 31.8 153.6
1996 27.9 133.1 35.7 149.5
1997 29.1 138.3 35.9 158.0
1998 30.0 146.4 34.7 160.2
1999 30.6 145.5 34.5 158.0
2000 31.6 142.5 34.5 152.1
2001 32.2 130.9 34.0 132.5
2002 31.7 137.7 32.7 139.5
2003 29.5 121.4 30.0 122.5
2004 25.7 112.2 25.7 112.2
CKD, chronic kidney disease.
a Per 1000 patient-years, including inpatient and outpatient transfusions.
Table 5. Transfusion rates a by erythropoiesis stimulating agent use
CKD patients Non-CKD patients
for anaemic CKD patients receiving ESAs and 174.7 transfusions
per 1000 patient-years for anaemic CKD patients
not receiving ESAs demonstrate how common transfusions
remain, despite the encouraging reductions in transfusions
over time. Why these anaemic CKD patients were not receiving
ESAs is unclear from this analysis. Possibly, the
transfusions were given for an indication other than anaemia
of CKD. A transfusion rate for non-anaemic CKD patients
of only 63 per 1000 patient-years (in contrast to 174.7 for
anaemic CKD patients not on ESAs) suggests but does not
prove that anaemia, in addition to the many comorbid conditions
borne by CKD patients, was the major reason for these
transfusions. Possibly, patients with CKD and mild anaemia
may receive transfusions during acute events when their
haemoglobin levels are close to the threshold level physicians
generally employ to transfuse patients. Comparing
transfusion rates in this cohort of CKD patients to a cohort
Table 6. Raw transfusion rates a for chronic kidney disease patients by
anaemia status and erythropoiesis stimulating agent use
Anaemia No Anaemia
ESA No ESA ESA No ESA
Year
1992 572.9 249.2 313.3 91.5
1993 739.7 242.8 180.7 79.6
1994 345.5 237.4 401.8 73.1
1995 466.7 222.9 231.9 73.2
1996 367.5 239.2 0.0 77.0
1997 505.7 235.4 163.0 77.8
1998 487.1 254.2 282.6 75.4
1999 391.4 241.7 152.3 82.7
2000 445.5 238.8 143.6 73.9
2001 266.6 216.1 128.7 72.7
2002 277.6 220.2 135.5 75.8
2003 217.5 192.1 153.8 68.8
2004 200.1 174.7 113.0 63.0
ESA, erythropoiesis stimulating agent
a Per 1000 patient-years, including inpatient and outpatient transfusions.
Raw rates Adjusted rates Raw rates Adjusted rates
Year ESA use, % ESA No ESA ESA No ESA ESA use, % ESA No ESA ESA No ESA
1992 0.8 538.6 137.4 399.2 161.8 0 2066.9 26.8 656.3 32.0
1993 0.9 668.4 128.6 471.1 150.3 0 1862.5 25.6 511.0 29.9
1994 1.3 347.7 126.8 201.7 137.5 0 1406.4 24.4 399.0 28.2
1995 1.7 436.4 121.6 286.7 127.4 0.3 205.2 25.0 151.3 28.4
1996 1.8 339.6 129.9 218.2 134.3 0.2 540.1 27.2 262.3 30.0
1997 2.1 481.9 132.5 308.3 134.0 0.1 1079.7 28.2 326.9 29.5
1998 2.5 474.7 139.5 300.2 136.9 0.2 871.7 28.8 264.8 29.6
1999 3.1 381.8 139.1 240.6 141.3 0.3 841.0 28.9 261.5 29.6
2000 3.8 433.5 132.6 268.9 133.1 0.4 755.5 29.6 232.3 30.3
2001 4.5 262.4 125.4 170.2 124.1 0.4 734.3 29.7 227.3 30.0
2002 5.0 274.5 131.1 184.5 129.7 0.5 580.3 29.4 182.0 29.5
2003 6.3 216.2 115.4 152.9 115.4 0.6 443.0 27.4 150.4 27.5
2004 7.5 198.6 105.5 142.0 105.5 0.7 262.1 24.3 97.7 24.3
CKD, chronic kidney disease; ESA, erythropoiesis stimulating agent.
a Per 1000 patient-years, including inpatient and outpatient transfusions.
Page 98

Transfusions and chronic kidney disease 3143
of dialysis patients we previously reported on, covering
the same study period (1992–2005), reveals that transfusion
rates were 2.7-fold higher for haemodialysis than for
non-dialysis-dependent CKD patients in 1992 and 2.1-fold
higher in 2005 (7).
Red blood cell transfusions are not without risks, including
transfusion reactions, transmission of infectious
agents and iron overload (8). Before the widespread adoption
of ESA therapy, 41% of dialysis patients were estimated
to show evidence of iron overload incurred from
multiple blood transfusions; this era has fortunately passed
(9). Other concerns include associated costs and the potential
adverse effect of blood transfusions on the development
of allo-antibodies, which in turn can prolong time on the
kidney transplant waiting list and possibly jeopardize the
kidney transplant outcome (10). Continued efforts to minimize
transfusion use, in accordance with the US Food and
Drug Administration guidelines, will likely minimize the
risks associated with transfusions. Identifying correctable
causes of anaemia, using ESAs thoughtfully, providing adequate
iron stores and improved understanding of reasons
for blood transfusion use may help reduce that use.
Our study is inherently limited by its retrospective design.
Furthermore, we had no information on the numbers
of or reasons for transfusions given to individual patients,
as indications for giving transfusions are not available in
claim data. Importantly, some non-anaemic CKD patients
not receiving ESAs received transfusions. This most likely
represents transfusions given due to trauma, surgical procedure
or gastrointestinal bleeding. It may also reflect a
general feeling in the medical community that sick patients,
particularly those with heart disease, benefit from
higher haemoglobin levels, despite lack of evidence to
support such practice (11). This latter possibility needs
to be substantiated. Also relevant is the unavailability of
haemoglobin levels at which transfusions took place; this
limits the ability to draw conclusions regarding whether the
haemoglobin threshold at which clinicians transfuse CKD
patients changed during this study period.
Our study used administrative claim data to define blood
transfusion occurrences. The accuracy of billing data for
identifying blood transfusions is unknown, but is likely
similar to the accuracy for other procedures, given that
reimbursement is tied to accurate claim submission and
erroneous claims are considered fraud. Finally, the temporal
trends in blood transfusion use may not be causally
related to increasing ESA use, but to other factors, such as
a decreased role of blood transfusions in medical practice,
as suggested by the decline in transfusion rates for nonanaemic
non-CKD patients between 1992 and 2004. The
former possibility, however, remains more likely.
In conclusion, these data indicate an encouraging decline
in transfusion rates for both CKD and non-CKD Medicare
patients, and ESA use is largely responsible for the observed
reduction in transfusion use in the setting of CKD. More
research is needed to better identify potentially avoidable
transfusions, as understanding the reasons for transfusion
use may lead to thoughtful avoidance of those that are not
absolutely necessary.
Acknowledgements. The authors thank Chronic Disease Research Group
colleagues Shane Nygaard, BA, for manuscript preparation, and Nan
Booth, MSW, MPH, for manuscript editing.
Conflicts of interest statement. This study was supported by a research
contract from Amgen Inc., Thousand Oaks, CA, USA. The contract provides
for the authors to have final determination of the content of this
manuscript. D.T.G. has received consulting fees from Amgen. H.N.I. and
A.I. consult for, and H.G. is employed by, the Chronic Disease Research
Group.
References
Page 99
1. Higgins MR, Grace M, Ulan RA et al. Anemia in hemodialysis patients.
Arch Intern Med 1977; 137: 172–176
2. Goodnough LT, Monk TG, Andriole GL. Erythropoietin therapy. N
EnglJMed1997; 336: 933–938
3. Jones M, Ibels L, Schenkel B et al. Impact of epoetin alfa on clinical
end points in patients with chronic renal failure: a meta-analysis.
Kidney Int 2004; 65: 757–767
4. Ross SD, Fahrbach K, Frame D et al. The effect of anemia treatment on
selected health-related quality-of-life domains: a systematic review.
Clin Ther 2003; 25: 1786–1805
5. Delano BG. Improvements in quality of life following treatment with
r-HuEPO in anemic hemodialysis patients. Am J Kidney Dis 1989;
14(Suppl 1): 14–18
6. Beusterien KM, Nissenson AR, Port FK et al. The effects of recombinant
human erythropoietin on functional health and wellbeing
in chronic dialysis patients. J Am Soc Nephrol 1996; 7: 763–
773
7. Ibrahim H, Ishani A, Foley RN et al. Temporal trends in red blood
transfusions among US dialysis patients, 1992–2005. Am J Kidney
Dis 2008; 52: 1115–1121
8. Kleinman S, Chan P, Robillard P. Risks associated with transfusion
of cellular blood components in Canada. Transfus Med Rev 2003; 17:
120–162
9. Hakim RM, Stivelman JC, Schulman G et al. Iron overload and mobilization
in long-term hemodialysis patients. AmJKidneyDis1987;
10: 293–299
10. Cardarelli F, Pascual M, Tolkoff-Rubin N et al. Prevalence and significance
of anti-HLA and donor-specific antibodies long-term after
renal transplantation. Transpl Int 2005; 18: 532–540
11. Besarab A, Bolton WK, Browne JK et al. The effects of normal as
compared with low hematocrit values in patients with cardiac disease
who are receiving hemodialysis and epoetin. N Engl J Med 1998; 339:
584–590
Received for publication: 20.2.09; Accepted in revised form: 17.4.09

Temporal Trends in Red Blood Transfusion Among US Dialysis
Patients, 1992-2005
Hassan N. Ibrahim, MD, MS, 1,2 Areef Ishani, MD, MS, 1,2 Robert N. Foley, MB, 1 Haifeng Guo, MS, 1
Jiannong Liu, PhD, 1 and Allan J. Collins, MD, FACP 1,2
Background: Studies addressing patterns and trends in red blood cell transfusion use in US
hemodialysis patients surprisingly have received little attention in the last decade.
Study Design: Retrospective cohort study.
Setting & Participants: Point prevalent (as of January 1 of each calendar year 1992 to 2005) dialysis
patients with Medicare Part A and Part B as primary insurance (n � 77,347 in 1992, n � 164,933 in
2005). The 6 months preceding January 1 of each year were used to assemble a comorbidity profile
based on administrative claims data.
Predictors: Hemoglobin levels, patient characteristics, comorbid conditions.
Outcomes: Blood transfusion events obtained from Part A and Part B files using code files for both
whole and packed red blood cell transfusions and hemoglobin levels.
Measurements: Comorbid conditions were defined by the presence of 1 or more inpatient/outpatient
institutional claims (inpatient hospitalization, skilled nursing facility, or home health agency), 2 or more
outpatient or physician/supplier claims, or 1 or more outpatient and 1 or more physician/supplier claims
for atherosclerotic heart disease, congestive heart failure, cerebrovascular accidents/transient ischemic
attacks, peripheral vascular disease, other cardiovascular diseases, chronic obstructive pulmonary
disease, gastrointestinal disorders, liver disease, arrhythmia, and diabetes mellitus.
Results: Raw transfusion rates decreased in both outpatient and inpatient settings from 535.33/1,000
patient-years for 1992 prevalent dialysis patients to 263.65/1,000 patient-years in 2005 (P for trend �
0.001, 1992 versus 1999 and 1999 versus 2005). Adjusted rates decreased similarly. This phenomenon
could not be explained by changes in case mix.
Limitations: Cause, effect, and confounding cannot be separated in this observational study. The
accuracy of blood transfusion billing data is unknown. Temporal trends may be related to factors other
than erythropoiesis-stimulating agent use.
Conclusion: Transfusion events in hemodialysis patients decreased more than 2-fold from 1992 to
2005; most of the decrease occurred in the first 5 years after erythropoietin was introduced.
Am J Kidney Dis 52:1115-1121. © 2008 by the National Kidney Foundation, Inc.
INDEX WORDS: Erythropoiesis-stimulating agents; hemodialysis; transfusions.
Anemia, almost a universal feature of endstage
renal disease (ESRD), is caused
largely by lack of erythropoietin production from
the failed kidneys. 1,2 In 1989, epoetin alfa (EPO),
the first erythropoiesis-stimulating agent (ESA),
was approved by the US Food and Drug Administration
for use in dialysis patients. The introduction
of ESAs to the care of dialysis patients has
been linked to improved quality of life and
well-being. 3-6 However, concerns have emerged
recently regarding the safety of exceeding recommended
hemoglobin levels (ie, �13 g/dL) because
ESAs have been linked to increased risk of
cardiovascular events. 7,8 This concern prompted
the Food and Drug Administration to include a
black box warning on ESA package inserts stating
“Patients experienced greater risk of death
and serious cardiovascular events when administered
erythropoiesis stimulating agents to target
higher versus lower hemoglobin level . . .” and,
“. . . the lowest dose needed to avoid red blood
Page 100
cell transfusions” should be used. 9,10 Although it
commonly is held that the introduction of ESAs
to the care of the ESRD population has resulted
in decreased need for blood transfusions, there is
no information about the impact of ESA therapy
on the rate of blood transfusions in US dialysis
patients beyond a single publication from 1 center.
11 The 1995 US Renal Data System Annual
From the 1 Chronic Disease Research Group, Minneapolis
Medical Research Foundation; and 2 University of Minnesota,
Minneapolis, MN.
Received February 25, 2008. Accepted in revised form
July 7, 2008. Originally published online as doi:
10.1053/j.ajkd.2008.07.022 on September 29, 2008.
Address correspondence to Hassan N. Ibrahim, MD, MS,
Division of Renal Diseases and Hypertension, University of
Minnesota, 717 Delaware St SE, Ste 353, MMC 1932,
Minneapolis, MN 55414. E-mail: ibrah007@umn.edu
© 2008 by the National Kidney Foundation, Inc.
0272-6386/08/5206-0012$34.00/0
doi:10.1053/j.ajkd.2008.07.022
American Journal of Kidney Diseases, Vol 52, No 6 (December), 2008: pp 1115-1121 1115

1116
Data Report provides the only large published
evidence documenting a decrease in the fraction
of dialysis patients receiving transfusions nationally
in the outpatient dialysis setting during the
period spanning the introduction of EPO. 12
Considering the paucity of information quantifying
the impact of ESAs on the frequency of
blood transfusions in the dialysis population, we
used US Renal Data System data to address this
issue between 1992 and 2005.
METHODS
Study Population
Point-prevalent dialysis patients with Medicare Part A and
Part B as their primary insurance source were studied as of
January 1 of each calendar year. The preceding 6 months
were used to assemble a comorbidity profile based on
administrative claims data. The first cohort consisted of
point prevalent dialysis patients on January 1, 1992, and the
last cohort consisted of point prevalent dialysis patients on
January 1, 2005. Each annual cohort was followed up for
transfusion events for 1 year.
Excluded Patients
Patients without Medicare Part A and Part B as their
primary insurance at the beginning of each cohort year were
excluded. We also excluded dialysis patients with evidence
of cancer or blood dyscrasias during the 6-month entry
period for each year because these patients can be hyporesponsive
to ESA therapy and may require multiple transfusions
not directly related to ESRD. Relevant diagnoses with
International Classification of Diseases, Ninth Edition, Clinical
Modification codes included cancer (140.xx to 172.xx,
174.xx to 208.xx, 230.xx to 231.xx, and 233.xx to 234.xx),
hereditary hemolytic anemia (282.x), acquired hemolytic
anemia (codes 283.x), aplastic anemia and other bone marrow
failure syndromes (284.x), coagulation defects (286.x),
myeloma (203.0x), myelofibrosis (289.89), and myelodysplasia
(238.7). Patients who did not survive the entry period
were also excluded. After these exclusions, approximately
85% of patients from each year constituted the final cohort.
Including the excluded patients in the analysis did not alter
results (data not shown).
Measurements
Comorbid conditions were defined by the presence of 1
Part A or Part B claim for atherosclerotic heart disease
(410.xx to 414.xx and V45.18 to V45.82), congestive heart
failure (398.91, 422.xx, 425.xx, 428.xx, 402.x1, 404.x1,
404.x3, and V42.1), cerebrovascular accident/transient ischemic
attack (430.xx to 438.xx), peripheral vascular disease
(440.xx to 44.xx, 447.xx, 451.xx to 453.xx, and 557.xx),
other cardiovascular diseases (420.xx to 421.xx, 423.xx to
424.xx, 429.xx, 785.0 to 785.3, V42.2, and V43.3), chronic
obstructive pulmonary disease (491.xx to 494.xx, 496.xx,
and 510.xx), gastrointestinal disorders (456.0 to 456.2, 530.7,
531.xx to 534.xx, 569.84, 569.85, and 578), liver disease
Page 101
Ibrahim et al
(570.xx, 571.xx, 572.1, 572.4, 573.1 to 573.3, and V42.7),
arrhythmia (426.xx to 427.xx, V45.0, and V53.30), or diabetes
mellitus (250.xx, 357.2, 362.0x, and 366.41).
Outcomes
Transfusions billed by dialysis facilities and those performed
during inpatient hospital stays were obtained from
Part A files by using code files for both whole and packed
blood cell transfusions. Those obtained from Part B files
included whole and packed red blood cell transfusions
regardless of whether these were leukocyte reduced, irradiated,
or deglycerolized. ESA use was defined by using
Healthcare Common Procedure Coding System codes. For
patients with more than 1 ESA claim in 1 month, average
hemoglobin levels were obtained for that month. Hemoglobin
levels greater than 18.3 g/dL or less than 3.3 g/dL were
considered invalid and coded as missing (hemoglobin in
g/dL may be converted to g/L by multiplying by 10).
Analysis
The transfusion event raw rate was calculated from total
number of transfusion events divided by total follow-up
time. Patients were censored at death, transplantation, loss to
follow-up, or end of Medicare coverage. Poisson regression
was used to determine the adjusted transfusion rates. The
prevalent cohort from calendar year 1992 was used as the
reference group. The Poisson model included the covariates
age, sex, race, primary cause of ESRD, dialysis vintage, and
the variables determined during the entry period for each
cohort: average hemoglobin level, parenteral iron use, length
of stay during hospital admissions, and comorbid conditions.
Because some patients may appear in more than 1 year
in the Poisson model, we compared transfusion rates in 1992
with rates in 1999 and 1999 to 2005. Overlap thus was less
than 10%, but we nevertheless accounted for the correlation
of transfusion rates over time for the same patient.
RESULTS
The 1992 cohort included 77,347 prevalent
dialysis patients (Table 1). The composition of
the prevalent cohort changed considerably over
time, including increases in mean age, proportion
of the population older than 75 years, and
number of patients initiating dialysis therapy
because of diabetes. However, the most striking
change was the increase in proportion of dialysis
patients with multiple comorbid conditions, particularly
cardiovascular disease.
The percentage of patients with hemoglobin
values less than 11 g/dL decreased steadily, from
64.0% in 1992 to 11.9% in 2004 (Table 2). In
parallel, the proportion of patients with hemoglobin
values greater than 12 g/dL increased from
0.5% in 1992 to 38.2% in 2004. Concurrent with
the increasing fraction of patients with hemoglobin
values greater than 12 g/dL, the proportion of

Table 1. Characteristics of Prevalent Dialysis Cohorts, 1992 to 2004 (even years shown)
patients receiving more than 58,000 U/mo of
ESA increased by almost 10-fold. Similarly, the
fraction of dialysis patients receiving parenteral
iron increased substantially, from just 0.3% in
1992 to 71.6% in 2004.
Figure 1 shows average hemoglobin levels in
the 3 months preceding and 3 months after the
transfusion month. The transfusion month, as
expected, coincided with the lowest hemoglobin
level of the study period in all cohorts studied.
Interestingly, the hemoglobin threshold at which
prevalent dialysis patients receive transfusions
has steadily increased over the years. This threshold
increased from 9.0 g/dL in 1992 to almost
Cohort Year
1992 1994 1996 1998 2000 2002 2004
Page 102
ESAs and Blood Transfusions 1117
Characteristics (n � 77,347) (n � 89,815) (n � 100,540) (n � 109,685) (n � 121,133) (n � 140,227) (n � 157,960)
Age (y) 57.9 � 16.0 58.4 � 16.1 58.9 � 16.0 59.3 � 15.9 59.8 � 15.8 60.2 � 15.7 60.6 � 15.6
0-44 17,364 (22.5) 19,455 (21.7) 20,517 (20.4) 21,476 (19.6) 22,563 (18.6) 24,472 (17.5) 25,895 (16.4)
44-64 28,256 (36.5) 32,180 (35.8) 36,636 (36.4) 40,468 (36.9) 44,999 (37.2) 53,420 (38.1) 61,684 (39.1)
65-74 20,588 (26.6) 24,064 (26.8) 26,485 (26.3) 28,119 (25.6) 30,281 (25.0) 34,051 (24.3) 37,548 (23.8)
�75 11,139 (14.4) 14,116 (15.7) 16,902 (16.8) 19,622 (17.9) 23,290 (19.2) 28,284 (20.2) 32,833 (20.8)
Men 39,769 (51.4) 46,142 (51.4) 51,808 (51.5) 56,756 (51.7) 62,809 (51.9) 74,018 (52.8) 83,919 (53.1)
Race
White 43,759 (56.6) 49,310 (54.9) 55,542 (55.2) 58,240 (53.1) 63,136 (52.1) 74,697 (53.3) 83,588 (52.9)
African American 31,045 (40.1) 37,056 (41.3) 40,626 (40.4) 45,790 (41.8) 51,382 (42.4) 57,182 (40.8) 64,657 (40.9)
Asian/other 2,543 (3.3) 3,449 (3.8) 4,372 (4.4) 5,655 (5.2) 6,615 (5.5) 8,348 (6.0) 9,715 (6.2)
Primary cause of ESRD
Diabetes 21,316 (27.6) 27,244 (30.3) 33,487 (33.3) 39,579 (36.1) 46,402 (38.3) 56,434 (40.2) 66,279 (42.0)
Hypertension 24,254 (31.4) 29,037 (32.3) 32,029 (31.9) 33,454 (30.5) 36,077 (29.8) 40,835 (29.1) 45,786 (29.0)
Glomerulonephritis 14,076 (18.2) 15,355 (17.1) 16,397 (16.3) 17,021 (15.5) 17,912 (14.8) 19,528 (13.9) 20,559I (13.0)
Cystic kidney disease 3,810 (4.9) 3,853 (4.3) 4,029 (4.0) 3,899 (3.6) 3,878 (3.2) 4,249 (3.0) 4,484 (2.8)
Other urological 2,347 (3.0) 2,464 (2.7) 2,544 (2.5) 2,923 (2.7) 3,221 (2.7) 3,824 (2.7) 4,116 (2.6)
Other cause 7,562 (9.8) 8,207 (9.1) 8,507 (8.5) 8,908 (8.1) 9,413 (7.8) 10,501 (7.5) 11,282 (7.1)
Unknown cause 3,294 (4.3) 3,332 (3.7) 3,361 (3.3) 3,773 (3.4) 4,139 (3.4) 4,771 (3.4) 5,336 (3.4)
Missing 688 (0.9) 323 (0.4) 186 (0.2) 128 (0.1) 91 (0.1) 85 (0.1) 118 (0.1)
Vintage (y)
�1 17,449 (22.6) 20,352 (22.7) 22,984 (22.9) 22,728 (20.7) 23,943 (19.8) 27,504 (19.6) 30,415 (19.3)
1-5 31,554 (40.8) 38,258 (42.6) 43,301 (43.1) 48,935 (44.6) 54,251 (44.8) 63,179 (45.1) 72,183 (45.7)
�5 28,344 (36.7) 31,205 (34.7) 34,255 (34.1) 38,022 (34.7) 42,939 (35.5) 49,544 (35.3) 55,362 (35.1)
Hospital stay (d)
0 47,779 (61.8) 56,360 (62.8) 62,654 (62.3) 68,822 (62.8) 75,143 (62.0) 86,494 (61.7) 97,731 (61.9)
1-3 10,531 (13.6) 12,907 (14.4) 15,735 (15.7) 17,830 (16.3) 20,171 (16.7) 23,567 (16.8) 25,746 (16.3)
4-6 7,316 (9.5) 8,189 (9.1) 9,475 (9.4) 10,137 (9.2) 11,201 (9.3) 13,171 (9.4) 15,418 (9.8)
�6 11,721 (15.2) 12,359 (13.8) 12,676 (12.6) 12,896 (11.8) 14,618 (12.1) 16,995 (12.1) 19,065 (12.1)
Comorbid conditions
ASHD/CHF/dys/other 35,908 (46.4) 42,731 (47.6) 63,263 (62.9) 58,920 (53.7) 65,405 (54.0) 92,266 (65.8) 92,435 (58.5)
CVA/TIA/PVD 21,129 (27.3) 28,437 (31.7) 34,662 (34.5) 38,280 (34.9) 42,412 (35.0) 50,902 (36.3) 60,000 (38.0)
GI/liver disease 14,547 (18.8) 17,778 (21.2) 20,404 (22.7) 25,835 (27.6) 31,515 (26.0) 23,245 (16.6) 24,615 (15.6)
COPD 5,930 (7.7) 8,046 (9.0) 10,073 (10.0) 11,958 (10.9) 14,009 (11.6) 17,826 (12.7) 23,052 (14.6)
Diabetes 23,306 (30.1) 30,637 (34.1) 40,718 (40.5) 48,841 (44.5) 56,077 (46.3) 71,390 (50.9) 86,748 (54.9)
Note: Values expressed as mean � SE or number (percent).
Abbreviations: ASHD, atherosclerotic heart disease; CHF, congestive heart failure; COPD, chronic obstructive pulmonary
disease; CVA, cerebrovascular accident; dys, dysrhythmia; ESRD, end-stage renal disease; GI, gastrointestinal; PVD,
peripheral vascular disease; TIA, transient ischemic attack.
10.3 g/dL in 2004. Most of the increase occurred
between 1992 and 2000, from 9.0 to 10.0 g/dL.
To simplify comparisons, we show trends in
blood transfusions in relation to patient age, sex,
race, primary cause of kidney disease, dialysis
vintage, and comorbid conditions for 1992, 1999,
and 2005 only (Table 3). Blood transfusion use
decreased steadily for patients aged 0 to 44 years
while increasing steadily for patients older than
75 years. Women received more transfusions
than men in each of the 3 years, and white
patients received about 55% of all transfusions.
In 2005, 45.2% of patients with diabetes as
primary cause of ESRD received transfusions

1118
Table 2. Hemoglobin, Erythropoiesis-Stimulating Agent, and Iron Use Trends, 1992 to 2004 (even years shown)
compared with 30.4% in 1992. Conversely,
only 16.2% of patients with glomerulonephritis
as primary cause of ESRD received transfusions
in 1992 compared with even fewer,
11.4%, in 2005. Dialysis vintage in relation to
transfusion use was fairly constant throughout
the observation period. The largest increase in
blood transfusion use was in patients with
atherosclerotic heart disease or congestive heart
failure.
Raw transfusion rates decreased considerably
over time, from 535.33 (95% confidence interval,
528.48 to 542.27) transfusions/1,000 patientyears
in 1992 to 263.65 (95% confidence interval,
261.04 to 266.29) transfusions/1,000 patientyears
in 2005 (Table 4). Multivariate analysis
excluding hemoglobin level yielded similar results.
However, additional adjustment for hemo-
Cohort Year
1992 1994 1996 1998 2000 2002 2004
(n � 77,347) (n � 89,815) (n � 100,540) (n � 109,685) (n � 121,133) (n � 140,227) (n � 157,960)
Hemoglobin (g/dL) 9.7 � 1.0 10.1 � 1.0 10.5 � 1.0 10.9 � 0.8 11.5 � 1.0 11.7 � 1.0 11.8 � 0.9
0-�11 49,471 (64.0) 58,160 (64.8) 54,345 (54.1) 48,704 (44.4) 27,398 (22.6) 23,507 (16.8) 18,775 (11.9)
�11-�12 4,453 (5.8) 10,411 (11.6) 23,452 (23.3) 40,642 (37.1) 49,005 (40.5) 60,484 (43.1) 65,070 (41.2)
�12 404 (0.5) 1,071 (1.2) 3,194 (3.2) 4,599 (4.2) 30,482 (25.2) 42,073 (30.0) 60,336 (38.2)
Missing/unknown
Total ESA use/mo
23,019 (29.8) 20,173 (22.5) 19,549 (19.4) 15,740 (14.4) 14,248 (11.8) 14,163 (10.1) 13,779 (8.7)
None 23,125 (29.9) 19,988 (22.3) 18,973 (18.9) 15,506 (14.1) 14,244 (11.8) 14,085 (10.0) 13,631 (8.6)
0-�28,000 units 31,676 (41.0) 31,786 (35.4) 29,336 (29.2) 32,791 (29.9) 28,866 (23.8) 33,586 (24.0) 36,514 (23.1)
28,000-�58,000 units 18,884 (24.4) 27,234 (30.3) 31,298 (31.1) 35,219 (32.1) 36,086 (29.8) 41,957 (29.9) 46,233 (29.3)
�58,000 units 3,662 (4.7) 10,807 (12.0) 20,933 (20.8) 26,169 (23.9) 41,937 (34.6) 50,599 (36.1) 61,582 (39.0)
Iron use (yes) 258 (0.3) 22,601 (25.2) 36,781 (36.6) 63,678 (58.1) 75,385 (62.2) 83,718 (59.7) 113,03 (71.6)
Note: Values expressed as mean � SE or number (percent). Hemoglobin in g/dL may be converted to g/L by
multiplying by 10.
Abbreviation: ESA, erythropoiesis-stimulating agent.
Mean hemoglobin level in g/dL
12.0
11.5
11.0
10.5
10.0
9.5
9.0
1992 1994 1996 1998 2000 2002 2004
-3 -2 -1 0 1 2 3
Month
Page 103
Ibrahim et al
globin levels showed that in more recent years,
transfusion rates at any hemoglobin level had
increased again. The difference between rates in
1992 and 1999 and 1999 and 2005 and after
accounting for patient correlation at these 2 times
was significant; P for trend � 0.001 for both
tests. Transfusion rates from the raw and the 2
adjusted models are shown in Fig 2. The major
decrease in transfusions occurred between 1992
and 1997, but rates reached a plateau thereafter.
DISCUSSION
Our results show a dramatic decrease in blood
transfusions in prevalent dialysis patients spanning
the study period from 1992 to 2005. Associated
with this decrease in blood transfusions was
a simultaneous increase in average ESA dose
administered to dialysis patients and average
Figure 1. Mean hemoglobin
levels at month of transfusion � 3
months in prevalent dialysis patients
from 1992 to 2004. Hemoglobin
in g/dL may be converted to
g/L by multiplying by 10.

Table 3. Characteristics of Prevalent Dialysis Patients Receiving Blood Transfusions, 1992, 1999, and 2005
Characteristics
reported hemoglobin concentration. Patients with
atherosclerotic heart disease, congestive heart
failure, or diabetes and those older than 75 years
experienced the greatest increase in blood transfusion
frequency. However, these temporal trends
persisted in models accounting for differences in
case mix.
EPO was first introduced in 1989, with a
primary indication for blood transfusion minimization
in dialysis patients. Before the widespread
adoption of ESA therapy in dialysis patients,
anemia, blood transfusions, and consequently
iron overload were common occurrences. Since
Transfusions 1992 Transfusions 1999 Transfusions 2005
No Yes No Yes No Yes
No. of
Patients %
No. of
Patients %
No. of
Patients %
No. of
Patients %
No. of
Patients %
Page 104
ESAs and Blood Transfusions 1119
No. of
Patients %
All 57,869 74.8 19,478 25.2 96,490 84.0 18,340 16.0 139,114 84.4 25,819 15.7
Age (y)
0-44 13,298 76.6 4,066 23.4 18,654 85.5 3,174 14.5 22,734 85.8 3,771 14.2
45-64 21,219 75.1 7,037 24.9 35,583 84.4 6,603 15.7 55,341 85.0 9,760 15.0
65-74 15,173 73.7 5,415 26.3 24,346 83.1 4,946 16.9 32,258 83.1 6,557 16.9
�75 8,179 73.4 2,960 26.6 17,907 83.2 3,617 16.8 28,781 83.4 5,731 16.6
Sex
Men 30,806 77.5 8,963 22.5 50,907 85.7 8526 14.4 75,456 85.6 12,693 14.4
Women 27,063 72.0 10,515 28.0 45,583 82.3 9814 17.7 63,658 82.9 13,126 17.1
Race
White 32,970 75.3 10,789 24.7 50,569 83.9 9,685 16.1 73,077 83.7 14,219 16.3
African American 22,941 73.9 8,104 26.1 40,823 84.1 7,694 15.9 57,206 85.2 9,966 14.8
Asian 1,232 77.1 365 22.9 2,878 84.7 521 15.3 5,177 84.8 930 15.2
Other 726 76.7 220 23.3 2,220 83.5 440 16.5 3,654 83.9 704 16.2
Primary cause of ESRD
Diabetes 15,393 72.2 5,923 27.8 35,384 82.4 7,576 17.6 58,644 83.4 11,678 16.6
Hypertension 18,013 74.3 6,241 25.7 29,272 84.8 5,254 15.2 40,665 85.0 7,158 15.0
Glomerulonephritis 10,927 77.6 3,149 22.4 14,980 85.9 2,466 14.1 17,709 85.8 2,944 14.3
Other/missing/unknown 13,536 76.5 4,165 23.5 16,854 84.7 3,044 15.3 22,096 84.6 4,039 15.5
Vintage (y)
�1 12,911 74.0 4,538 26.0 19,544 84.3 3,643 15.7 26,531 85.0 4,695 15.0
1-�5 23,492 74.5 8,062 25.6 42,975 83.9 8,272 16.1 63,609 84.2 11,953 15.8
�5 21,466 75.7 6,878 24.3 33,971 84.1 6,425 15.9 48,974 84.2 9,171 15.8
Comorbid conditions
ASHD 11,524 69.0 5,174 31.0 37,254 80.7 8,927 19.3 64,840 81.5 14,686 18.5
CHF 10,992 66.5 5,529 33.5 38,484 80.2 9,507 19.8 61,557 81.2 14,263 18.8
CVA/TIA 4,343 66.8 2,160 33.2 11,418 79.0 3,029 21.0 19,983 79.5 5,151 20.5
PVD 11,862 68.3 5,516 31.7 26,754 79.6 6,842 20.4 43,748 80.5 10,618 19.5
Other cardiac 11,011 67.2 5,387 32.9 25,385 79.4 6,596 20.6 43,968 80.5 10,651 19.5
COPD 4,032 68.0 1,898 32.0 9,988 78.7 2,699 21.3 19,862 79.7 5,054 20.3
Gastrointestinal 3,690 61.2 2,340 38.8 7,188 74.7 2,429 25.3 10,080 75.6 3,251 24.4
Liver disease 6,862 72.5 2,608 27.5 22,162 82.7 4,629 17.3 13,092 81.7 2,933 18.3
Arrhythmia 8,879 68.8 4,023 31.2 18,597 80.0 4,645 20.0 31,492 80.3 7,744 19.7
Diabetes 16,312 70.0 6,994 30.0 42,746 81.4 9,741 18.6 77,990 82.9 16,142 17.2
Note: � 2 tests for comparing patients who received versus did not receive transfusions for all variables are significant
except for race and vintage for 1999 (gray shaded cells).
Abbreviations: ASHD, atherosclerotic heart disease; CHF, congestive heart failure; COPD, chronic obstructive pulmonary
disease; CVA, cerebrovascular accident; ESRD, end-stage renal disease; PVD, peripheral vascular disease; TIA, transient
ischemic attack.
the introduction of ESAs, several observational
and randomized studies have assessed different
aspects of the risks and benefits of these therapies.
3-8 Although firm conclusions regarding the
risk-benefit ratio of ESA use are far from established,
recent evidence has generated concern
over targeting normal hemoglobin levels in patients
with chronic kidney disease. 7,8
Surprisingly few observational studies of transfusion
use in dialysis populations have been performed,
especially in more recent times. We are
aware of only one previous study that assessed the
impact of the introduction of EPO therapy on the

1120
Table 4. Raw and Adjusted Transfusion Rates (per 1,000 patient-years), Prevalent Dialysis Patients, 1992 to 1995
Year
No. of
Patients
Raw Rates
rate of blood transfusions in dialysis patients. 11
This single-center study showed a substantial decrease
in the rate of blood transfusions with the
introduction of EPO (rates, 13.4/100 in 1988 and
4.1/100 in 1991; P � 0.01). Our results, although
observational, confirm the dramatic decrease in
rate of blood transfusion during the past 14 years,
with substantial increases in EPO therapy and average
hemoglobin concentrations.
Although blood transfusion use has decreased,
it continues to be a frequent occurrence and, in
recent years, has increased at any given hemoglobin
concentration. The hemoglobin threshold at
Transfusion Rates/1,000 Patient-Years
Adjusted Rates, 1992 as Reference Year
Page 105
Ibrahim et al
Hemoglobin Level Excluded Hemoglobin Level Included
All Inpatient Outpatient All Inpatient Outpatient All Inpatient Outpatient
1992 77,347 535.33 309.70 225.63 535.33 309.70 225.63 535.33 309.70 225.63
1993 84,066 465.81 275.08 190.74 448.73 265.40 183.38 447.65 264.66 183.10
1994 89,815 396.97 235.25 161.72 380.47 227.15 153.27 388.33 230.63 157.71
1995 93,585 365.96 217.72 148.24 343.55 207.11 136.33 358.75 213.53 145.16
1996 100,540 347.82 208.94 138.88 316.01 187.21 128.83 343.09 198.12 145.37
1997 105,122 312.01 201.72 110.29 297.27 191.33 105.90 330.98 207.33 123.91
1998 109,685 297.10 206.18 90.92 282.85 195.72 87.09 320.73 214.34 106.83
1999 114,830 276.28 204.78 71.50 246.59 178.51 68.14 301.33 208.91 93.96
2000 121,133 265.04 195.75 69.30 250.23 180.52 69.74 323.48 218.74 106.85
2001 128,961 287.77 199.00 88.77 271.59 183.29 88.32 363.20 225.17 141.00
2002 140,227 284.60 197.09 87.51 263.13 175.28 88.08 357.41 218.24 143.59
2003 150,482 270.81 194.29 76.52 255.18 178.46 76.77 358.91 225.78 136.58
2004 157,960 270.86 202.30 68.56 260.73 188.73 72.19 375.95 242.90 137.79
2005 164,933 263.65 196.99 66.66 245.92 177.53 68.51 382.48 242.63 144.97
Note: Prevalent dialysis patients in 1992 to 2005 with 90-day rule, initiated therapy at least 6 months before January 1 of
each year, with Medicare Part A and Part B as primary payer as of December 31 of previous year. Patients are followed up
until December 31 of each year and censored at death, transplantation, loss to follow-up, or payer status change.
which transfusions occur has increased by about
0.1 g/dL/y. This may be a direct consequence of
the increasing comorbidity burden in dialysis
patients and a general feeling in the medical
community that sick patients, particularly those
with heart disease, may benefit from higher hemoglobin
levels despite lack of evidence to support
such practice. 13 Red blood cell transfusions have
intrinsic risks, including transfusion reactions,
transmission of infectious agents, and iron overload.
Before the widespread adoption of ESA
therapy, 41% of dialysis patients were estimated
to have iron overload, evidenced by ferritin level
Figure 2. Transfusion event
rates, prevalent dialysis patients,
1992 to 2005; raw rates and adjusted
rates including and excluding
hemoglobin levels.

greater than 2,000 ng/mL (ferritin levels expressed
in ng/mL and �g/L are equivalent). 14
Other concerns include the costs associated with
blood transfusions and the potential adverse effect
of blood transfusions on the development of
alloantibodies, which can prolong time on the
kidney transplant waiting list and jeopardize the
outcome of kidney transplantation. 15
Our study has limitations. It is observational,
with all the limitations associated with observational
studies. In particular, cause, effect, and
confounding cannot be separated. In addition,
our study used administrative claims data to
define the occurrence of blood transfusions. The
accuracy of billing data for blood transfusions
currently is unknown. However, claims data for
blood transfusions likely are similar in accuracy
to other procedures, given that reimbursement is
tied to the accurate submission of claims and
erroneous claims are considered fraud. Finally,
the temporal trends in blood transfusions may
not be causally related to increasing ESA use, but
to other factors, such as a decreased role of blood
transfusions in medical practice. However, if
transfusions were decreasing solely because of a
greater reluctance to administer transfusions, one
would expect to observe a decrease in average
hemoglobin values over time, but our results
show the opposite phenomenon; average hemoglobin
concentrations have increased during the
past 13 years, and patients are administered transfusions
at higher hemoglobin levels.
Transfusion events in hemodialysis patients
decreased more than 2-fold between 1992 and
2005, with most of the decrease occurring in the
first 5 years after ESA introduction.
ACKNOWLEDGEMENTS
The authors thank Chronic Disease Research Group colleagues
Shane Nygaard, BA, for manuscript preparation,
and Nan Booth, MSW, MPH, for manuscript editing.
Support: Information for funding sources is listed in the
financial disclosure.
Financial Disclosure: This study was supported by a
research contract from Amgen Inc, Thousand Oaks, CA, the
manufacturer of a proprietary epoetin alfa. The contract
provides for the authors to have final determination of the
content of this manuscript. Dr Foley and Dr Collins have
received consulting fees from Amgen and Roche.
REFERENCES
Page 106
ESAs and Blood Transfusions 1121
1. Goodnough LT, Monk TG, Andriole GL: Erythropoietin
therapy. N Engl J Med 336:933-938, 1997
2. Higgins MR, Grace M, Ulan RA, et al: Anemia in
hemodialysis patients. Arch Intern Med 137:172-176, 1977
3. Jones M, Ibels L, Schenkel B, et al: Impact of epoetin
alfa on clinical end points in patients with chronic renal
failure: A meta-analysis. Kidney Int 65:757-767, 2004
4. Ross SD, Fahrbach K, Frame D, et al: The effect of
anemia treatment on selected health-related quality-of-life domains:
A systematic review. Clin Ther 25:1786-1805, 2003
5. Delano BG: Improvements in quality of life following
treatment with r-HuEPO in anemic hemodialysis patients.
Am J Kidney Dis 14:14-18, 1989
6. Beusterien KM, Nissenson AR, Port FK, et al: The
effects of recombinant human erythropoietin on functional
health and well-being in chronic dialysis patients. J Am Soc
Nephrol 7:763-773, 1996
7. Singh AK, Szczech L, Tang KL, et al: Correction of
anemia with epoetin alfa in chronic kidney disease. N Engl
J Med 355:2085-2098, 2006
8. Drueke TB, Locatelli F, Clyne N, et al: Normalization
of hemoglobin level in patients with chronic kidney disease
and anemia. N Engl J Med 355:2071-2084, 2006
9. Fishbane S, Nissenson AR: The new FDA label for
erythropoietin treatment: How does it affect hemoglobin
target? Kidney Int 72:806-813, 2007
10. Amgen: Aranesp [package insert]. Thousand Oaks,
CA, Amgen, 2007
11. Goodnough LT, Strasburg D, Riddell J, et al: Has
recombinant human erythropoietin therapy minimized redcell
transfusions in hemodialysis patients? Clin Nephrol
41:303-307, 1994
12. US Renal Data System: USRDS 1995 Annual Data
Report. The National Institutes of Health, National Institute
of Diabetes and Digestive and Kidney Diseases, Bethesda,
MD, 1995
13. Besarbe A, Bolton K, Browne J, et al: The effects of
normal as compared with low hematocrit values in patients
with cardiac disease who are receiving hemodialysis and
epoetin. N Engl J Med 339:584-590, 1989
14. Hakim RM, Stivelman JC, Schulman G: Iron overload
and mobilization in long-term hemodialysis patients.
Am J Kidney Dis 10:293-299, 1987
15. Cardarelli F, Pascual M, Tolkoff-Rubin N, et al:
Prevalence and significance of anti-HLA and donor-specific
antibodies long-term after renal transplantation. Transpl Int
18:532-540, 2005

Kidney International, Vol. 60 (2001), pp. 741–747
Novel erythropoiesis stimulating protein for treatment of
anemia in chronic renal insufficiency
FRANCESCO LOCATELLI, JESÚS OLIVARES, ROWAN WALKER, MARTIN WILKIE, BARBARA JENKINS,
CLAIRE DEWEY, and STEPHEN J. GRAY, on behalf of the EUROPEAN/AUSTRALIAN
NESP 980202 STUDY GROUP 1
Divisione di Nefrologia e Dialisi, Ospedale A. Manzoni, Lecco, Italy; Servicio de Nefrología, Hospital General Universitario de
Alicante, Alicante, Spain; Department of Nephrology, Royal Melbourne Hospital, Parkville Victoria, Australia; Sheffield Kidney
Institute, Northern General Hospital NHS Trust, Sheffield, and Amgen Limited, Cambridge, England, United Kingdom
1 In addition to the authors, the European/Australian NESP 980202
Study Group consists of the following individuals (in alphabetical or-
der): J. Braun (KfH Kuratorium für Dialyse und Neirentransplantation
Novel erythropoiesis stimulating protein for treatment of
anemia in chronic renal insufficiency.
Background. Novel erythropoiesis stimulating protein (NESP)
is a glycoprotein with a threefold longer terminal half-life than
recombinant human erythropoietin (rHuEPO) in humans. The
aim of this study was to determine whether NESP is effective
for the treatment of anemia at a reduced dosing frequency
relative to rHuEPO in patients with chronic renal failure not
e.V., Nürnberg, Germany), Jacques Chanard (Hôpital Maison Blanche, yet on dialysis [chronic renal insufficiency (CRI)].
Reims, France), Naomi Clyne (Karolinska Hospital, Stockholm, Sweden),
Gerry A. Coles (University Hospital of Wales, Cardiff, Wales, UK),
José M. Cruz (Hospital Universitario La Fe, Valencia, Spain), Marc
De Broe (UZ Antwerpen, Edegen, Belgium), Angel M. Luis de Fran-
cisco (Hospital de Valdecilla, Santander, Spain), Max Dratwa (Hôpital
Universitaire Brugmann, Brussels, Belgium), Ram Gokal (Manchester
Royal Infirmary, Manchester, England, UK ), Carola Grönhagen-Riska
Methods. This was a multicenter, randomized, open-label
study. A total of 166 rHuEPO-naive patients with CRI were
randomized in a 3:1 ratio to receive NESP (0.45 �g/kg once
weekly) or rHuEPO (50 U/kg twice weekly) administered
subcutaneously for up to 24 weeks. Dose adjustments were
made as necessary to achieve a hemoglobin response, defined
(Helsinki University Hospital, Helsinki, Finland), Helmut Graf (Kran- as an increase �1.0 g/dL from baseline and a concentration
kenanstalt Rudolfstiftung, Vienna, Austria), Levi Guerra (Hospital de
São João, Porto, Portugal), Thierry Hannedouche (Hospices Civils,
Strasbourg, France), David Harris (Westmead Hospital, Westmead, New
South Wales, Australia), Carmel Hawley (Princess Alexandra Hospital,
Woolloongabba, Queensland, Australia), Herwig Holzer (Medizinische
Universitatklinik Graz, Graz, Austria), Michel Jadoul (UCL St. Luc,
�11.0 g/dL.
Results. During the 24-week treatment period, 93% (95%
CI, 87 to 97%) of patients receiving NESP and 92% (95% CI, 78
to 98%) of patients receiving rHuEPO achieved a hemoglobin
response. The median time to response was seven weeks (range
Brussels, Belgium), Peter Kerr (Monash Medical Centre, Clayton, Vic- of 3 to 25 weeks) in both groups. After correction of anemia,
toria, Australia), W. Kösters (Dialyseninstitut, Vielingen-Schwenningen,
Germany), Jean-Louis Lacombe (Clinique Saint-Exupéry, Toulouse,
France), Norbert Lameire (UZ Ghent, Ghent, Belgium), J. Mann
(Schwabing General Hospital, Munich, Germany), Paolo Marai (Ospe-
dale A. Manzoni, Lecco, Italy), Maria Joao Pais (Hospital de Santa Cruz,
Carnaxide, Portugal), Luis Piera (Hospital Universitario Vall D’Hebrón,
mean hemoglobin concentrations were maintained within the
target range of 11.0 to 13.0 g/dL for the remainder of the 24week
treatment period. The safety profiles of NESP and rHuEPO
were similar, and no antibodies were detected to either drug.
Conclusions. These results demonstrate that NESP safely
Barcelona, Spain), Pedro Ponce (Hospital Garcia da Orta, Almada, Por- and effectively corrects and maintains hemoglobin concentratugal),
E. Quellhorst (Nephrologisches Zentrum Niedersachsen, Hannoversch-Münden,
Germany), Emilio Rivolta (IRCCS Ospedale Maggiore
di Milano, Milano, Italy), R.M. Schäfer (Medizinische Poliklinik, Mun-
ster, Germany), G. Stein (Klinikum der Friedrich-Schiller-Universität,
Jena, Germany), Marinus A. van den Dorpel (St. Clara Hospital, Rotterdam,
The Netherlands), Christopher Winearls (Oxford Renal Unit, Ox-
tions at a reduced dosing frequency relative to rHuEPO in
patients with CRI, providing a potential benefit to patients and
health care providers.
ford, England, UK), and Carmine Zoccali (Azienda Ospedaliera Bianchi,
Melacrino e Morelli, Reggio Calabria, Italy).
Patients with chronic renal failure (CRF) frequently
experience anemia, which can significantly affect their
Key words: anemia, novel erythropoiesis stimulating factor, hemoglo-
bin, recombinant human erythropoietin, red blood cell, and chronic
renal insufficiency.
morbidity, mortality, and quality of life. Recombinant
human erythropoietin (rHuEPO) is effective for the treat-
Received for publication December 19, 2000
and in revised form February 23, 2001
ment of anemia in CRF based on its ability to increase
hemoglobin, virtually eliminate the need for red blood
Accepted for publication March 1, 2001 cell transfusions, mitigate symptoms associated with ane-
© 2001 by the International Society of Nephrology
mia, and improve energy and physical functioning [1, 2].
741
Page 107

742
Locatelli et al: NESP treatment in CRI
Page 108
However, in most patients, administration is required androgen therapy within eight weeks of the first planned
two or three times per week.
dose of study drug.
Previous research on rHuEPO has indicated that its
carbohydrate content significantly affects its half-life and Study design
biological activity in vivo (abstract; Egrie, Glycoconj J This was a multicenter, randomized, open-label study.
10:263, 1993). Novel erythropoiesis stimulating protein All patients completed a screening period within two
(NESP; ARANESP�) is a glycoprotein that was de- weeks before the first dose of study drug. Eligible patients
signed by introducing five amino acid changes into the were then randomized by a centralized system to receive
primary sequence of rHuEPO to create two extra con- NESP or rHuEPO in a 3:1 ratio, respectively. NESP was
sensus N-linked glycosylation sites. Consequently, NESP administered subcutaneously at a starting dose of 0.45
has five N-linked carbohydrate chains, whereas rHuEPO �g/kg once weekly, and rHuEPO was administered sub-
has only three. The increased carbohydrate content of cutaneously at a starting dose of 50 U/kg twice weekly.
NESP results in a terminal half-life in humans that is The dose and frequency of NESP were chosen based on
approximately threefold longer than that of rHuEPO experience in treating anemia in dialysis patients [5]. The
after intravenous administration (25.3 hours vs. 8.5 hours, dose of rHuEPO was chosen to approximate the starting
respectively) and at least twofold longer after subcutane- dose of NESP, based on a peptide mass conversion factor
ous administration (48.8 hours vs. 18 to 24 hours, respec- (1.0 �g/kg NESP � 200 U/kg rHuEPO), and the fre-
tively) [3, 4]. This unique pharmacokinetic profile allows quency of rHuEPO administration was chosen to reflect
for a reduced dosing frequency, representing a potential the most commonly used frequency at the participating
clinical advantage over rHuEPO [5].
centers. Study drug dose was adjusted by 25% of the
Preliminary studies have indicated that NESP, at a starting dose as necessary to achieve a hemoglobin instarting
dose of 0.45 to 0.75 �g/kg, is safe and effective for crease of �1.0 g/dL from baseline and to maintain hemo-
the treatment of anemia in patients with CRF receiving globin concentrations within a range of 11.0 to 13.0 g/dL.
dialysis [5]. However, the effects of NESP have not been Additionally, if a patient’s hemoglobin increased by �2.0
evaluated in patients with CRF not yet on dialysis [chronic g/dL during any four-week period, study drug doses were
renal insufficiency (CRI)]. This multicenter, randomized, reduced by 25% of the starting dose. If a patient’s hemo-
open-label study was designed to evaluate the efficacy globin concentration increased to �14.0 g/dL, study drug
and safety of NESP in this patient population. The was withheld until the hemoglobin concentration fell
rHuEPO treatment group was included to provide a below 12.0 g/dL and was restarted at a dose 25% lower
reference reflecting current medical practice.
than the previous dose. Patients received the first dose
of study drug within seven days of randomization and
METHODS
Patients
continued treatment for 24 weeks. After 24 weeks, patients
receiving NESP could continue treatment, and
patients receiving rHuEPO ended the study.
Our study was approved by each institution’s indepen- The primary measure of efficacy was the proportion
dent ethics committee, and all patients gave written in- of patients achieving a hemoglobin response during the
formed consent before participation. Patients 18 years 24-week treatment period, defined as an increase in he-
of age or older with a diagnosis of CRI were enrolled moglobin of �1.0 g/dL from baseline and a hemoglobin
at 32 dialysis centers in Europe and 4 in Australia. Pa- concentration of �11.0 g/dL. For this assessment, hemotients
were not receiving dialysis and were rHuEPO na- globin measurements were taken at two-week intervals
ive (that is, had not received treatment with rHuEPO through the end of the 24-week treatment period. The
within 12 weeks before the first planned dose of study efficacy of NESP was also evaluated by assessing the time
drug). Additional entry criteria included a hemoglobin required to achieve a hemoglobin response, hemoglobin
concentration �11.0 g/dL, adequate iron stores (as deter- concentration over time, the dose of study drug at the
mined by serum ferritin �100 �g/L), serum vitamin B12 time of hemoglobin response and at week 24, and the
and folate levels above the lower limit of the normal number of patients receiving red blood cell transfusions.
range, and a creatinine clearance of �30 mL/min (as Measurements to evaluate hemoglobin concentration
estimated by the Cockcroft-Gault equation). Exclusion over time were taken at four-week intervals (that is, at
criteria included uncontrolled hypertension (diastolic weeks 5, 9, 13, 17, 21, and 25).
blood pressure �100 mm Hg); congestive heart failure Safety was assessed by monitoring adverse events, lab-
(New York Heart Association Class III or IV); hematooratory variables (hematology, biochemistry, and iron
logic disorders that could cause anemia, systemic infec- status), vital signs, and antibody formation to NESP or
tion or inflammatory disease; or other disorders that rHuEPO. In addition, the ability to increase hemoglobin
could interfere with the response to NESP or rHuEPO. levels safely was evaluated by the maximum increase in
Patients had not received a red blood cell transfusion or hemoglobin within any four-week period and the num-

Page 109
Locatelli et al: NESP treatment in CRI 743
ber of increases in hemoglobin �2.0, �2.5, or �3.0 g/dL
within any four-week period. Hematology parameters
Table 1. Patient demographic and baseline characteristics in novel
erythropoiesis stimulating protein (NESP)- or recombinant human
erythropoietin (rHuEP)-treated patientsa and vital signs were evaluated every two weeks throughout
the study. Biochemistry parameters and antibody
formation were evaluated every 12 weeks, and iron status
was evaluated every four weeks.
NESP rHuEPO
(N � 129) (N � 37)
Sex
Men 70 (54%) 19 (51%)
Study medications
Women
Race
59 (46%) 18 (49%)
Novel erythropoiesis stimulating protein was supplied
by Amgen Inc. (Thousand Oaks, CA, USA). Recombi-
Asian
Black
White
3 (2%)
3 (2%)
123 (95%)
1 (3%)
0 (0%)
36 (97%)
Age years b nant human erythropoietin (Epoetin alfa) was obtained
from commercial sources and provided by the pharmacy
at each study center. To ensure adequate support of the
Primary cause of renal failure
Diabetes
Hypertension
60.4 (15.0)
32 (25%)
15 (12%)
60.6 (15.7)
8 (22%)
1 (3%)
erythropoietic response to study drug, intravenous iron
therapy was required to be administered to patients with
serum ferritin values �100 �g/L. The intravenous iron
Glomerulonephritis
Polycystic kidney disease
Other urologic
Other
24 (19%)
6 (5%)
5 (4%)
32 (25%)
10 (27%)
2 (5%)
0 (0%)
11 (30%)
dosing regimen used for patients with serum ferritin val- Unknown
Hemoglobin g/dL
15 (12%) 5 (14%)
b ues �100 �g/L or �100 �g/L was determined by the Serum ferritin lg/L
9.3 (1.0) 9.8 (1.1)
c Creatinine clearance mL/min
168 (30–1420) 151 (31–899)
b individual center’s treatment policy.
15.7 (6.6) 15.7 (6.4)
Statistical analysis
a Baseline is defined as the closest measurement taken before the first dose
of study drug
All patients who received at least one dose of the
study drug were included in the analyses of efficacy and
safety. An additional, prospectively defined “per-protocol”
analysis of efficacy was conducted in those patients
b Mean (SD)
c Median (range)
who completed 24 weeks of treatment, had at least 75% tients receiving NESP or rHuEPO were diabetes mellitus
of the postbaseline hemoglobin measurements, and re- and glomerulonephritis. Hypertension was the most
ceived within �25% of their total prescribed dose of common medical condition among patients in both the
study drug, no more than one incorrect dose of study NESP and rHuEPO groups at baseline (91 and 84%,
drug, and no red blood cell transfusions before achieving respectively). The mean baseline hemoglobin concentra-
a hemoglobin response. tion was lower in the NESP group (9.3 g/dL, range of
Descriptive statistics were summarized for all end 6.6 to 11.0) than in the rHuEPO group (9.8 g/dL, range
points according to the treatment group. This study was of 6.1 to 11.5; Table 1). Median baseline serum ferritin
powered to evaluate whether a clinically meaningful pro- levels (168 and 151 �g/L in the NESP and rHuEPO
portion of patients receiving NESP achieved a hemoglobin
response, which was considered to be 50% of patients,
but was not designed to compare NESP and
rHuEPO. The sample size of at least 120 NESP patients
was chosen to give a power of �99% for demonstrating
groups, respectively) indicated that most patients were
iron replete [6, 7]. Mean creatinine clearance was similar
in the NESP (15.7 mL/min; range of 7 to 44) and rHuEPO
(15.7 mL/min, range of 4 to 33) groups.
that the response rate for NESP exceeds 50% (that is, Efficacy
the lower limit of the confidence interval for the NESP
response rate is �50%).
A hemoglobin response (defined as an increase of
�1.0 g/dL from baseline and a concentration �11.0 g/dL)
was achieved by 93% (95% CI, 87 to 97%) of patients
RESULTS receiving NESP during the 24-week treatment period.
Patient characteristics
Similarly, 92% (95% CI, 78 to 98%) of patients achieved
Patients were randomized to receive NESP (129 pa-
tients) or rHuEPO (37 patients). All 166 patients were
included in the efficacy and safety analyses. The per-
protocol analysis set included 105 NESP patients and 29
a response in the rHuEPO group. These findings were
supported by the per-protocol analysis, in which 94% of
the NESP patients (95% CI, 88 to 98%) and 100% of
the rHuEPO patients (95% CI, 88 to 100%) achieved a
rHuEPO patients. Overall, 54% of patients were men, hemoglobin response. The percentage of patients achiev-
and most were white (96%). The mean age was 61 years, ing a hemoglobin increase of �1.0 g/dL from baseline
with a range of 19 to 87 years. Demographic characteris- also was comparable between the NESP (99%; 95% CI,
tics were similar between the two treatment groups (Ta- 96 to 100%) and rHuEPO (95%; 95% CI, 82 to 99%)
ble 1). The most frequent causes of renal failure in pa- groups. In both treatment groups, the median time to

744
Locatelli et al: NESP treatment in CRI
Page 110
Fig. 1. Mean (95% CI) hemoglobin concentrations
at four-week intervals. The shaded
area indicates target range. Symbols are the patients
receiving: (�) recombinant human erythropoietin
(rHuEPO); (�) novel erythropoiesis
stimulating protein (NESP). Abbreviations are:
E, number of patients receiving rHuEPO; N,
number of patients receiving NESP.
achieve a hemoglobin response was seven weeks (range The percentage of patients with dose reductions to levels
of 3 to 25 weeks).
below their starting dose up to week 24 was similar in
Mean hemoglobin concentrations increased over the the NESP (68%) and rHuEPO (70%) groups. In patients
initial 12 weeks for patients in both groups and were who continued to receive NESP after the 24-week treat-
maintained within the target range of 11.0 to 13.0 g/dL ment period, the median weekly weight-adjusted dose
for the remainder of the 24-week treatment period. In at week 48 was 0.30 �g/kg (range of 0.1 to 1.7).
patients continuing to receive NESP after the 24-week Few patients required a red blood cell transfusion in
treatment period, the mean hemoglobin concentration the NESP or rHuEPO treatment groups (5 and 8%,
remained within the target range for up to 48 weeks respectively). Only three NESP-treated patients and two
(Fig. 1). The difference in hemoglobin levels observed rHuEPO-treated patients received a red blood cell transbetween
the NESP and rHuEPO patients at baseline fusion before achieving a hemoglobin response.
appeared to remain constant throughout the study; thus,
no difference in the change in hemoglobin from baseline Safety
between the two treatment groups was evident. After The safety profiles of NESP and rHuEPO were similar.
the initial four weeks of treatment, the two groups had A total of 107 of 129 NESP patients (83%) and 24 of 37
comparable increases in mean hemoglobin concentra- rHuEPO patients (65%) experienced at least one ad-
tions: 1.38 g/dL (95% CI, 1.21 to 1.55) in the NESP group verse event during the study, most of which were mild
and 1.40 g/dL (95% CI, 1.07 to 1.72) in the rHuEPO to moderate in severity. The interpretation of differences
group. Mean changes in hemoglobin from baseline were between the treatment groups is confounded by the
very similar between the treatment groups up to 24 weeks. open-label design of the study and the potential over-
Fifty-eight percent of NESP patients and 59% of attribution of adverse events to a novel agent (NESP)
rHuEPO patients did not require any dose adjustment compared with an established therapy. Adverse events
before achieving a hemoglobin response. A similar per- occurring in more than 10% of patients in either treatcentage
of patients receiving NESP or rHuEPO had a ment group are presented in Table 2. The most frequently
dose increase (35 and 32%, respectively) or a dose de- reported adverse events in the NESP and rHuEPO groups
crease (7 and 9%, respectively) before achieving a hemo- were hypertension (32 and 22%, respectively) and pe-
globin response. At the time of hemoglobin response, ripheral edema (13 and 11%, respectively). Most of these
the median weekly weight-adjusted dose of NESP was events were attributed to concurrent medical conditions
0.46 �g/kg (range of 0.3 to 2.3 �g/kg), and the corre- and were consistent with events expected in this patient
sponding dose of rHuEPO was 100 U/kg (range of 75 population. Adverse events that were considered by the
to 175 U/kg). Both doses were nearly identical to those investigator to be treatment related were reported with
at the beginning of the study. At week 24, median study the same frequency in each treatment group (30% NESP
drug doses had decreased to 0.34 �g/kg (range of 0.0 to and 27% rHuEPO). Hypertension was the most common
1.3 �g/kg) in patients receiving NESP and to 56.9 U/kg of these events (21% NESP and 19% rHuEPO).
(range of 19 to 250 U/kg) in patients receiving rHuEPO. There were six reported deaths (4% NESP and 3%

Page 111
Locatelli et al: NESP treatment in CRI 745
Table 2. Adverse events occurring in �10% of novel erythropoiesis Table 3. Deaths during the study or within 28 days after withdrawal
stimulating protein (NESP)- or recombinant human erythropoietin in novel erythropoiesis stimulating protein (NESP)- or
(rHuEPO)-treated patients recombinant human erythropoietin (rHuEPO)-treated patients
a
NESP rHuEPO Treatment
Adverse event (N � 129) (N � 37) group Age/sex Study week Primary cause of death
Hypertension 41 (32%) 8 (22%) rHuEPO 75/M 7 Cardiac arrest
Peripheral edema 17 (13%) 4 (11%) NESP 64/M 16 Cardiac arrest
Fatigue 16 (12%) 0 (0%) NESP 75/F 13a Cachexia
NESP 53/M 10a Diarrhea 14 (11%) 3 (8%)
Multi-organ failure
Headache 14 (11%) 4 (11%) NESP 64/M 22 Cardiac arrest
Nausea 11 (9%) 5 (14%) NESP 64/F 23 Hematemesis
Pruritis 7 (5%) 4 (11%) a Occurred within 28 days after withdrawal from the study
a Data are presented as the number of patients (%) experiencing each adverse
event.
rHuEPO) during the 24-week treatment period or within
28 days after withdrawal from the study (Table 3). The
deaths were largely due to cardiovascular failure and
occurred in patients with a history of cardiovascular disease.
Increases in hemoglobin were well controlled in pa-
tients receiving NESP or rHuEPO. The mean of the
maximum increase in hemoglobin over any four-week
interval was 1.99 g/dL (95% CI, 1.88 to 2.11) in the
NESP group and 2.13 g/dL (95% CI, 1.88 to 2.39) in
the rHuEPO group. The percentage of patients with a
hemoglobin increase of �2.0 g/dL over any four-week
interval was 51% in the NESP group and 60% in the
rHuEPO group. In addition, a lower percentage of pavalues
were 155 � 19 and 81 � 11 mm Hg, respectively,
at baseline and 146 � 20 and 79 � 12 mm Hg, respec-
tively, at week 25. In the rHuEPO group, values were
141 � 21 and 79 � 10 mm Hg at baseline and 143 � 23
and 79 � 8 mm Hg at week 25.
No clinically meaningful changes in laboratory values
were evident, and values were similar between the NESP
and rHuEPO groups. Mean � SD creatinine clearance
values were 15.7 � 6.6 mL/min in the NESP group and
15.7 � 6.4 mL/min in the rHuEPO group at baseline,
and 14.1 � 6.7 mL/min and 13.7 � 7.6 mL/min, respectively,
after 24 weeks of treatment. No antibody forma-
tion to NESP or rHuEPO was detected for any patient
during the study.
tients in the NESP group relative to the rHuEPO group DISCUSSION
had a hemoglobin increase of �2.5 g/dL (21 and 32%,
respectively) or �3.0 g/dL (9 and 14%, respectively) over
any four-week period. In patients who had study drug
In patients with chronic renal failure, anemia is associ-
ated with an increased risk of cardiovascular morbidity
and mortality and a reduction in quality of life. Correcwithheld
for hemoglobin concentrations �14.0 g/dL tion of hemoglobin and hematocrit levels has been shown
(24% NESP and 35% rHuEPO), the median time re- to reduce cardiovascular risk factors such as left ventricuquired
for hemoglobin levels to return to �12.0 g/dL lar hypertrophy and improve energy and physical funcwas
similar in the NESP (7 weeks, range of 2 to 13) and tioning, emphasizing the importance of treating anemia
rHuEPO (9 weeks, range of 6 to 13) groups (Fig. 2). in this population [2, 8, 9].
After four weeks of treatment, median serum ferritin This is the first report of the administration of NESP
concentrations had decreased in the NESP and rHuEPO to patients with CRI. The results of this study demongroups
from baseline values of 168 and 151 �g/L to 80 strate that NESP, administered once weekly at a starting
and 63 �g/L, respectively. Median serum ferritin concen- dose of 0.45 �g/kg, is safe and effective for the correction
trations remained �100 �g/L until week 17, when values of anemia and maintenance of hemoglobin concentrations
increased to 108 and 101 �g/L in the NESP and rHuEPO in this patient population. Of the 129 patients randomgroups,
respectively. Median concentrations remained ized to receive NESP, 93% achieved a hemoglobin reabove
100 �g/L for the rest of the 24-week treatment sponse during the 24-week treatment period. Although
period. Intravenous iron was administered to 77% of hemoglobin concentrations were lower in the NESP
patients receiving NESP and 81% of patients receiving group than in the rHuEPO group at baseline, both the
rHuEPO, at a total mean patient dose of 685 mg and proportion of patients achieving a hemoglobin response
678 mg, respectively. In addition, oral iron was adminis- and the time to achieve this response were the same
tered to a similar proportion of patients in the NESP for patients receiving NESP once weekly or rHuEPO
(52%) and rHuEPO (51%) groups. administered twice weekly. The degree of anemia correc-
Vital signs were stable during the 24-week treatment tion with NESP treatment was also similar to that obperiod
in both treatment groups. In patients receiving served in previous studies of rHuEPO administered
NESP, mean � SD systolic and diastolic blood pressure three times weekly in patients with CRI [10–12].

746
Locatelli et al: NESP treatment in CRI
Page 112
Fig. 2. Time required for hemoglobin concentrations
to return to �12.0 g/dL after withheld
doses for hemoglobin concentrations �14.0
g/dL. Symbols are: (�) rHuEPO (N � 13);
(�) NESP (N � 31). Means and 95% CIs are
shown.
Consistent with previous observations of NESP ad- laboratory or vital sign measures, and changes in mean
ministered intravenously or SC in patients with CRF re- creatinine clearance were similar between treatment
ceiving dialysis [5], a NESP starting dose of 0.45 �g/kg/ groups.
week was effective for the correction of anemia in the The correction of anemia with NESP treatment was
CRI population. The median dose of NESP at the time well controlled, with mean hemoglobin concentrations
patients achieved the target hemoglobin range was simi- maintained within the target range of 11.0 to 13.0 g/dL
lar to the starting dose. However, the median NESP dose for up to 48 weeks. During the initial 24-week treatment
was reduced to 0.34 �g/kg after 24 weeks of treatment, phase, increases in hemoglobin over any four-week pe-
potentially reflecting titration to maintain hemoglobin riod were �2.5 g/dL in most patients, consistent with
within the target range. Reduced maintenance doses current recommendations for the management of anemia
were also observed for patients receiving rHuEPO in in CRI [6, 7]. In patients who had doses withheld for
this study and have been reported in other studies of hemoglobin concentrations �14.0 g/dL during the study,
patients with CRF receiving rHuEPO [1, 13]. the time required for levels to return to �12.0 g/dL was
Adverse events were consistent with those expected similar in the NESP and rHuEPO treatment groups. This
in a population of patients with CRI and were generally result suggests that the rate of decline in hemoglobin
considered unrelated to the study drug. While a higher after withholding NESP or rHuEPO therapy is primarily
overall incidence of adverse events was reported for the determined by destruction of circulating red blood cells
NESP group than for the rHuEPO group, this difference as they reach the end of their lifespan, and not by clearmay
have been a consequence of the open-label design ance of the study drug.
of the study and the potential over-attribution of adverse Median serum ferritin concentrations in both treat-
events to a novel agent compared to an established therapy. ment groups decreased below the recommended level
The 3:1 randomization of patients to NESP or rHuEPO of 100 �g/L [6, 7] during the initial correction of anemia
and the resulting small number of patients evaluated and subsequently increased following supplemental in-
in the rHuEPO group may have also contributed to the travenous iron therapy. Other studies in patients with
observed difference. Patients in the NESP group had lower CRI receiving rHuEPO therapy have also reported a
baseline hemoglobin, a higher incidence of hypertension high incidence of absolute or functional iron deficiency
at baseline (91% NESP and 84% rHuEPO), and higher requiring intravenous or oral supplementation [10, 11,
baseline systolic blood pressure values (155 mm Hg 14]. As iron deficiency is the most common cause of
NESP and 141 mm Hg rHuEPO). These results suggest resistance to rHuEPO therapy, iron stores should be
that the NESP group had a greater burden of comorbid- monitored regularly in patients receiving NESP to enity
and may partially explain the difference in adverse sure adequate support of the erythropoietic response.
events between treatment groups, particularly since hy- Further research into optimal iron supplementation in
pertension was the most common adverse event reported this population is warranted.
in this study.
Although rHuEPO most commonly is administered
No clinically significant changes were observed for any two or three times weekly for the treatment of anemia

Page 113
Locatelli et al: NESP treatment in CRI 747
in CRF, it has also been used once weekly in some
patients. As a result of the extended half-life of NESP
relative to rHuEPO, even longer dosing intervals are
ease: Results of a phase III multicenter clinical trial. Ann Intern
Med 111:992–1000, 1989
2. Revicki DA, Brown RE, Feeny DH, et al: Health-related quality
of life associated with recombinant human erythropoietin therapy
possible with NESP therapy [3]. In a randomized, comparative
study of NESP and rHuEPO in dialysis patients,
95% of patients who were receiving rHuEPO once
for predialysis chronic renal disease patients. Am J Kidney Dis
25:548–554, 1995
3. Macdougall IC, Gray SJ, Elston O, et al: Pharmacokinetics of
novel erythropoiesis stimulating protein compared with Epoetin
weekly at baseline successfully maintained stable hemoglobin
concentrations when switched to NESP adminisalfa
in dialysis patients. J Am Soc Nephrol 10:2392–2395, 1999
4. Macdougall IC, Roberts DE, Coles GA, Williams JD: Clinical
pharmacokinetics of Epoetin (recombinant human erythropoie-
tered once every two weeks [5]. The longer dosing interval tin). Clin Pharmacokinet 20:99–113, 1991
with NESP offers the possibility of greater convenience
and less patient discomfort by reducing the number of
5. Macdougall IC: Novel erythropoiesis stimulating protein. Semin
Nephrol 20:375–381, 2000
6. NKF-DOQI clinical practice guidelines for the treatment of anemia
physician office visits (where standard practice) and sub- of chronic renal failure. Am J Kidney Dis 30(Suppl):S192–S240,
cutaneous injections. With dose administration once
weekly or once every two weeks, patients can avoid up
1997
7. Working Party for European Best Practice Guidelines for
the Management of Anemia in Patients with Chronic Renal
to 104 injections per year. Further clinical studies to
evaluate the efficacy and safety of NESP at extended
dosing intervals are currently underway.
Failure: European best practice guidelines for the management
of anemia in patients with chronic renal failure. Nephrol Dialysis
Transplant 14(Suppl 5):1–50, 1999
8. Hayashi T, Suzuki A, Shoji T, et al: Cardiovascular effect of
In conclusion, these results demonstrate that NESP
safely and effectively corrects and maintains hemoglobin
concentrations at a reduced dosing frequency relative to
normalizing the hematocrit level during erythropoietin therapy in
predialysis patients with chronic renal failure. Am J Kidney Dis
35:250–256, 2000
9. Moreno F, Sanz-Guajardo D, López-Gómez JM, et al: Increasing
rHuEPO in patients with CRI, providing a potential
benefit to patients and health care providers.
the hematocrit has a beneficial effect on quality of life and is safe
in selected hemodialysis patients. J Am Soc Nephrol 11:335–342,
2000
10. Eschbach JW, Kelly MR, Haley NR, et al: Treatment of the
ACKNOWLEDGMENTS
anemia of progressive renal failure with recombinant human erythropoietin.
N Engl J Med 321:158–163, 1989
This study was supported by Amgen Inc. Results from this study
were presented in abstract form at the annual meeting of the American
Society of Nephrology in Toronto, Canada, October, 2000. The authors
thank Ms. Janis Schaap for assistance in writing this manuscript.
11. Watson AJ, Gimenez LF, Cotton S, et al: Treatment of the anemia
of chronic renal failure with subcutaneous recombinant human
erythropoietin. Am J Med 89:432–435, 1990
12. US Recombinant Human Erythropoietin Predialysis Study
Group: Double-blind, placebo-controlled study of the therapeutic
Reprint requests to Francesco Locatelli, M.D., Divisione di Nefro-
logia e Dialisi, Azienda Ospedaliera di Lecco, Ospedale Alessandro
Manzoni, Via dell’Eremo 9-11, 23900 Lecco, Italy.
E-mail: nefrologia@ospedale.lecco.it
use of recombinant human erythropoietin for anemia associated
with chronic renal failure in predialysis patients. Am J Kidney Dis
18:50–59, 1991
13. Eschbach JW, Egrie JC, Downing MR, et al: Correction of the
anemia of end-stage renal disease with recombinant human erythropoietin:
Results of a combined phase I and II clinical trial. N Engl
REFERENCES
J Med 316:73–78, 1987
14. Macdougall IC, Hörl WH, Jacobs C, et al: European Best Prac-
1. Eschbach JW, Abdulhadi MH, Browne JK, et al: Recombinant tice Guidelines 6–8: Assessing and optimizing iron stores. Nephrol
human erythropoietin in anemic patients with end-stage renal dis- Dial Transplant 15(Suppl 4):20–32, 2000

Double-Blind Comparison of Full and Partial Anemia
Correction in Incident Hemodialysis Patients without
Symptomatic Heart Disease
Patrick S. Parfrey,* Robert N. Foley, † Barbara H. Wittreich, ‡ Daniel J. Sullivan, §
Martin J. Zagari, ‡ and Dieter Frei; ‡ for the Canadian European Study Group
*Memorial University of Newfoundland, St. John’s, Newfoundland, Canada; † Chronic Disease Research Group,
Minneapolis, Minnesota; ‡ Ortho Biotech, Bridgewater, New Jersey; and § Johnson and Johnson, Pharmaceutical Research,
LLC, Raritan, New Jersey
It is unclear whether physiologic hemoglobin targets lead to cardiac benefit in incident hemodialysis patients without
symptomatic heart disease and left ventricular dilation. In this randomized, double-blind study, lower (9.5 to 11.5 g/dl) and
higher (13.5 to 14.5 g/dl) hemoglobin targets were generated with epoetin � over 24 wk and maintained for an additional 72
wk. Major eligibility criteria included recent hemodialysis initiation and absence of symptomatic cardiac disease and left
ventricular dilation. The primary outcome measure was left ventricular volume index (LVVI). The study enrolled 596 patients.
Mean age, duration of dialysis therapy, baseline predialysis hemoglobin, and LVVI were 50.8 yr, 0.8 yr, 11.0 g/dl, and 69 ml/m 2 ,
respectively; 18% had diabetic nephropathy. Mean hemoglobin levels in the higher and lower target groups were 13.3 and 10.9
g/dl, respectively, at 24 wk. Percentage changes in LVVI between baseline and last value were similar (7.6% in the higher and
8.3% in the lower target group) as were the changes in left ventricular mass index (16.8 versus 14.2%). For the secondary
outcomes, the only between-group difference was an improved SF-36 Vitality score in the higher versus the lower target group
(1.21 versus �2.31; P � 0.036). Overall adverse event rates were similar in both target groups; higher (P < 0.05) rates of skeletal
pain, surgery, and dizziness were seen in the lower target group, and headache and cerebrovascular events were seen in the
higher target group. Normalization of hemoglobin in incident hemodialysis patients does not have a beneficial effect on
cardiac structure, compared with partial correction.
J Am Soc Nephrol 16: 2180–2189, 2005. doi: 10.1681/ASN.2004121039
Left ventricular hypertrophy (LVH) is present in
nearly 80% of dialysis patients (1). Morphologically,
left ventricular dilation and concentric hypertrophy
are approximately equally represented, and both conditions
are associated with higher rates of cardiovascular events
(1–7). In dialysis patients, anemia is associated with LVH,
left ventricular dilation, congestive heart failure, hospitalization,
and mortality (8–11). Controlled studies with quality of
life (QOL) and left ventricular mass as end points support
partial correction of hemoglobin in dialysis patients (12–14).
Received December 3, 2004. Accepted April 6, 2005.
Published online ahead of print. Publication date available at www.jasn.org.
Address correspondence to: Dr. Patrick Parfrey, Division of Nephrology, Health
Sciences Center, Memorial University of Newfoundland, Prince Philip Drive, St.
John’s, Newfoundland A1B 3V6, Canada. Phone: 709-777-7261; Fax: 709-777-6995;
E-mail: pparfrey@mun.ca
Conflict of interest statement: P.S.P. has received research support and is an
academic advisor to companies that make erythropoietin products—Ortho Biotech,
Amgen, and Roche. P.S.P. declares that he had full access to all of the data
in the study and had final responsibility for the decision to submit for publication.
R.N.F. has received research support and honoraria from Ortho Biotech. D.J.S. is
an employee of Johnson & Johnson. B.H.W., M.J.Z., and D.F. are employees of
Ortho Biotech.
P.S.P. and R.N.F. contributed equally to this work.
Page 114
These findings underpin guidelines from the United States,
Europe, and Canada, which suggest hemoglobin levels between
11 and 12 g/dl in dialysis patients (15–17). The potential
risks and benefits of higher hemoglobin targets in dialysis
patients have generated considerable discussion.
Evaluation of studies reported to date (18–21), with the
benefit of hindsight, led to concerns in one or more of the
following areas: Late intervention, small sample size, short
follow-up, and nonblinded participants. The current study
was conducted to compare the impact of higher versus lower
hemoglobin targets on left ventricular size in incident hemodialysis
patients without symptomatic cardiac disease. Secondary
outcomes included left ventricular mass index
(LVMI), 6-min walking test, QOL, and the occurrence of
adverse events.
Materials and Methods
Among incident hemodialysis patients without symptomatic cardiac
disease and left ventricular dilation, we compared the effects of higher
(13.5 to 14.5 g/dl) and lower (9.5 to 11.5 g/dl) hemoglobin targets
maintained with epoetin � on (1) left ventricular volume index (LVVI),
the primary outcome; (2) LVMI; (3) the incidence of de novo congestive
heart failure; (4) QOL, as revealed by vitality, fatigue, and quality of
social interaction scores; and (5) 6-min walking test performance.
Copyright © 2005 by the American Society of Nephrology ISSN: 1046-6673/1607-2180

J Am Soc Nephrol 16: 2180–2189, 2005 Anemia Correction in Hemodialysis Patients 2181
Study Population
Adult (�18 yr of age) hemodialysis patients were selected using the
following inclusion criteria: (1) Predialysis hemoglobin between 8 and
12 g/dl; (2) hemodialysis started in the preceding 3 to 18 mo (this
relatively broad definition for early intervention was chosen to facilitate
enrollment); (3) LVVI �100 ml/m 2 on screening echocardiography,
with normal being �90 ml/m 2 ; and (4) predialysis diastolic BP �100
mmHg. The exclusion criteria were (1) symptomatic ischemic heart
disease or heart failure; (2) angiographic critical coronary artery disease;
(3) current treatment �10 mg of prednisone daily for a failed renal
transplant; (4) surgery for pericardial or valvular disease within the last
2 yr; (5) medical conditions that are likely to reduce the response to
epoetin �, including uncorrected iron deficiency; (6) concurrent malignancy;
(7) blood transfusion within the preceding month; (8) therapy
with cytotoxic agents; (9) seizure within the previous year; (10) hypersensitivity
to intravenous iron; and (11) current pregnancy or breastfeeding.
Design
A randomized, double-blind design was used with patients and
outcome assessors but not treating physicians, who were blinded to
assigned hemoglobin target. The study was divided into a 24-wk titration
phase, which was used to achieve hemoglobin targets, and a 72-wk
maintenance period.
Local independent ethics committees or institutional review boards
approved the study protocol form before study initiation. The study
was conducted in accordance with Good Clinical Practice guidelines
Table 1. Baseline characteristics for the lower and higher target groups a
Total (n � 596;
Mean � SD or %)
and the Declaration of Helsinki. All patients gave informed consent
before study enrollment.
Study Procedures
The study was centrally monitored from St. John’s, Canada (P.S.P.),
for Canadian patients and Manchester, England (R.N.F.), for European
patients. Thirty percent of patients enrolled were from Canada, and
70% were from Europe. Screening echocardiograms were sent to a
central site (Cardialysis B.V., Rotterdam, The Netherlands) to determine
whether LVVI was �100 ml/m 2 . Randomization was performed
at the two coordinating centers with an interactive voice randomization
telephone system, using permuted blocks, stratified by gender and
concurrent epoetin use.
Laboratory tests were measured by Quest Diagnostics (Van Nuys,
CA, and Heston, UK). Baseline evaluations included medical history,
vital signs, height and weight, blood chemistry, M-mode echocardiography,
electrocardiography, Kidney Disease Quality of Life (KDQOL)
(22), Functional Assessment of Chronic Illness Therapy (FACIT) fatigue
scale (23), and a 6-min walking test (24). As part of the sponsor’s effort
to investigate the occurrence of pure red-cell aplasia in patients who
received epoetin �, the protocol was amended after 36 mo to mandate
measurement of anti-erythropoietin antibodies and serum erythropoietin.
The following treatment targets were used: (1) Predialysis hemoglobin
levels of 9.5 to 11.5 g/dl throughout in the lower target group and,
in the higher target group, increments of 0.5 to 1.0 g/dl biweekly, until
achieving stability between 13.5 and 14.5 g/dl (measured weekly for 24
Lower (9.5 to 11.5 g/dl;
n � 300;
Mean � SD or %)
Higher (13.5 to 14.5 g/dl;
n � 296;
Mean � SD or %)
Age (yr) 50.8 � 15.4 49.4 � 15.2 52.2 � 15.6
Duration of dialysis (mo) 10.1 � 5.0 10.2 � 5.1 10.0 � 4.9
Weight (kg) 75 � 16 74 � 17 75 � 16
Height (m) 1.68 � 0.10 1.68 � 0.10 1.68 � 0.09
Hemoglobin (g/dl) 11.0 � 1.2 11.0 � 1.2 11.0 � 1.2
Transferrin saturation (%) 36.2 � 17.2 36.8 � 17.8 35.7 � 16.7
Urea reduction ratio (%) 65.9 � 10.7 66.0 � 11.3 65.7 � 10.1
Serum albumin (g/dl) 3.9 � 0.3 4.0 � 0.3 3.9 � 0.3
Male 60 60 60
Previous epoetin use 92 91 93
Race
white 89 88 91
black 5 6 4
Primary cause of renal failure
glomerulonephritis 29 29 28
diabetic nephropathy 18 17 19
polycystic kidney disease 9 8 10
hypertension 8 9 7
unknown 10 10 9
other 27 27 27
Dialysis access
fistula 84 83 86
graft 6 5 6
catheter 10 12 8
a P � 0.05 for all comparisons between lower and higher target groups, except age (P � 0.02).
Page 115

wk, then biweekly); (2) predialysis diastolic BP between 70 and 90
mmHg (measured weekly for 24 wk, then every 4 wk); (3) urea reduction
ratio �67% (measured every 4 wk for 24 wk, then every 12 wk);
and (4) transferrin saturation �20% (measured biweekly for 24 wk,
then every 4 wk).
Between February 2000 and June 2001, 596 patients were enrolled in
95 treatment centers in 10 European countries and Canada. For each
patient, a form was faxed weekly to the coordinating center displaying
hemoglobin, epoetin � regimen, BP, and transferrin saturation levels.
Treatment recommendations were faxed back, weekly. Initially, epoetin
� could be administered subcutaneously or intravenously. A study
amendment, effective August 22, 2002, limited administration to the
intravenous route. The last patient completed the study in May 2003.
A predefined algorithm was used for epoetin � therapy. Epoetinnaïve
patients in the higher target group were assigned a dose of 150
IU/kg per wk. For hemoglobin levels that deviated from target, epoetin
doses were changed by 25% of the previous dose or 25 IU/kg.
Outcome Measures
Echocardiograms were performed at 24, 48, and 96 wk after study
start. Standard echocardiographic recordings were performed within 1
kg of dry weight, usually within 24 h after the midweek dialysis. Dry
weight was the optimal postdialysis weight for BP control without
symptoms of blood volume contraction. American Society of Echocardiography
criteria (25) were used, and echocardiograms were read by
an independent central reader, Cardialysis. Left ventricular volume, in
milliliters, was calculated as 0.001047 (left ventricular end-diastolic
diameter mm) 3 (26). Left ventricular mass, in grams, was calculated as
0.00083 � [(end-diastolic diameter mm � interventricular septum
thickness mm � posterior wall thickness mm) 3 � (end-diastolic diameter
mm) 3 � 0.6] (27). LVVI and LVMI were calculated as left ventricular
volume/M 2 body surface area and left ventricular mass/M 2 body
surface area, respectively (28).
The 6-min walking test was performed at 24, 48, and 96 wk after
study start, and QOL scales were recorded at 24, 36, 48, 60, 72, 84, and
96 wk. The FACIT fatigue scale and two subscales from the KDQOL,
the SF-36 Vitality and KDQOL Quality of Social Interaction, were
chosen for analysis before initiating the study, with higher values
representing better QOL. The 6-min walking test measured the distance
walked in 6 min, change in heart rate from pre- to postexercise, and
perceived exertion (24). De novo heart failure was defined as dyspnea at
rest with two of the following: Increased jugular venous pressure,
bilateral basal crackles, radiographic pulmonary hypertension, and
radiographic interstitial edema. Safety evaluations included adverse
events as defined by the investigator, vital signs, and laboratory measures.
Statistical Analyses
Data from Canadian Normalization of Hemoglobin trial were used to
guide sample size estimation (19). The sample size needed to detect a
15% difference between treatment groups was calculated as 166 per
treatment group, given a two-tailed significance of 0.05, a power of 0.90,
and an SD of the percentage change in left ventricular cavity volume
index of 42%. With an expected dropout rate of 40%, primarily as a
result of transplantation, 277 patients were required for each treatment
group.
The intention-to-treat philosophy was used. Analysis of covariance
(adjusted for age, epoetin use at randomization, and baseline LVH) was
used to adjust the effect of hemoglobin target on percentage change in
LVVI and LVMI. De novo heart failure rates were compared using
time-to-event analyses. QOL outcomes were compared in patients with
Page 116
2182 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2180–2189, 2005
at least one QOL score, using piecewise linear regression, where QOL
growth curves were allowed to change slopes at week 24, i.e., one slope
was used from weeks 0 to 24, and another slope was used from week
24 and beyond. The main QOL analysis assumed that missing QOL
data were missing at random; however, a sensitivity analysis, which
did not assume that data were missing at random, yielded similar
findings. The primary QOL assessment was mean follow-up minus
baseline QOL score, with P values adjusted for multiple comparisons
using the Hochberg procedure (29).
The proportions with at least one treatment-emergent adverse event
were compared by treatment assignment; because the number of component
conditions was very large, any adverse event that occurred in
�10% of patients is reported here, as are vascular events, access loss
events, cardiac events, and death when the occurrence rates exceeded
2%. Parametric, two-sided statistical tests were used throughout, with
a type I error of 0.05, using SAS Version 6.12.
Role of the Funding Source
The study sponsor, Johnson & Johnson Pharmaceutical Research and
Development, LLC, identified the participating centers, monitored the
data collection, and entered the data in a central database.
Results
Patients
Baseline characteristics of the 596 patients enrolled are shown
in Table 1 and were similar in the two randomly assigned
treatment groups, apart from age, which was greater in the
higher target group. As anticipated, 324 (54%) patients remained
in the study for 96 wk, 160 (54%) in the higher and 164
(55%) in the lower target groups. The reasons for study exit—
renal transplantation (n � 133, 67 in the higher and 66 in the
lower target group), adverse events (n � 76, 39 and 37), patient
choice (n � 28, 9 and 19), loss to follow-up (n � 2, 1 and 1), and
other (n � 36, 21 and 15)—were similar in the two target
Figure 1. Hemoglobin levels (top) and epoetin � doses (bottom),
presented as means � 2 SE in the higher and lower target
groups.

Table 2. Clinical indicators for the lower and higher target groups
groups. The mean times in the study were similar for both
target groups: 73.4 � 31.1 wk for the higher compared with
74.0 � 31.8 for the lower target group. In total, 339 patients had
echocardiograms performed after 72 wk of follow-up.
Figure 1 shows hemoglobin levels and epoetin � doses in the
two target groups. Toward the end of the titration phase, at 24
wk, the mean hemoglobin levels in the higher and lower target
groups were 13.3 � 1.5 and 10.9 � 1.2 g/dl, respectively.
During the maintenance phase, from weeks 24 to 96, mean
hemoglobin levels were 13.1 � 0.9 g/dl in the higher and 10.8 �
0.7 g/dl in the lower target group. Predialysis hemoglobin
levels, averaged for the whole study, including both the titration
and the maintenance phases, were 12.6 � 1.0 g/dl in the
higher and 10.9 � 0.7 g/dl in the lower target group, whereas
the corresponding postdialysis levels were 13.5 � 1.2 and
11.8 � 1.1 g/dl, respectively. The proportionate increase in
post- to predialysis hemoglobin, which reflects interdialytic
weight gain, were similar in both groups (7 versus 8%).
Table 2 summarizes clinical indicators at the beginning and
Lower (9.5 to 11.5 g/dl) Higher (13.5 to 14.5 g/dl)
N
Mean � SD
or n (%)
N
Mean � SD
or n (%)
Transferrin saturation (%)
baseline 294 36.8 � 17.5 293 35.7 � 16.72
last value
Systolic BP (mmHg)
294 34.2 � 16.53 293 34.6 � 14.91
baseline a
300 140 � 20 295 144 � 21.65
last value
Diastolic BP (mmHg)
299 139 � 22 295 141 � 22.14
baseline 298 80 � 12 295 81 � 11.48
last value
No. of antihypertensive drugs
299 79 � 14 295 80 � 12.57
baseline 300 1.8 � 1.4 296 2.0 � 1.5
last value a
Urea reduction ratio (%)
300 1.7 � 1.5 296 2.0 � 1.7
baseline 281 66 � 11 281 66 � 10.10
last value a
292 68 � 10 285 66 � 10.87
Serum albumin (g/dl)
baseline 296 4.0 � 0.3 290 3.9 � 0.29
last value
No. with transferrin saturation �20%
279 4.0 � 0.4 279 3.9 � 0.35
baseline 294 40 (14) 293 35 (12)
last value
No. receiving intravenous iron
294 42 (14) 293 33 (11)
baseline 300 129 (43) 296 139 (47)
last value a
Page 117
J Am Soc Nephrol 16: 2180–2189, 2005 Anemia Correction in Hemodialysis Patients 2183
300 75 (25) 296 110 (37)
No. receiving antihypertensives
baseline 300 242 (81) 296 244 (82)
last value 300 222 (74) 296 226 (76)
a The following comparisons were statistically significant: Baseline systolic BP (P � 0.019), last value for number of
antihypertensive drugs (P � 0.020), last value for urea reduction ratio (P � 0.044), and last value for number receiving
intravenous iron (P � 0.001).
at the end of the study. The last observed values were similar in
the two target groups except for a higher proportion receiving
intravenous iron, number of antihypertensive drugs, and lower
urea reduction ratios in the higher target group (all P � 0.05).
There was no difference in the change in number of antihypertensive
drugs from baseline to last value, comparing the two
groups.
Table 3 shows the major cardiac outcomes of the study,
compared by assigned hemoglobin target. LVVI did not change
significantly over time in the overall study population, and
changes in LVVI were similar in both target groups. Although
LVMI increased over time in the overall study population (P �
0.001), the changes observed were unaffected by target hemoglobin
assignment. These conclusions were unchanged when
only those who had echocardiogram performed after week 80
were included in the analysis. Rates of de novo heart failure
were similar in both target groups, as were the changes in
6-min walking distance.
Table 4 shows the QOL outcomes, by assigned hemoglobin.

Table 3. Changes in LVVI, LVMI, 6-min walking test, and incidence of heart failure for the lower and higher
target groups a
SF-36 Vitality scores improved more in the higher than in the
lower target group (P � 0.036), and differences between target
groups, adjusted for multiple comparisons, were significant at
weeks 24, 36, 48, 60, and 72 (Figure 2). No significance betweengroup
differences were seen in the time course of FACIT Fatigue
scores and the KDQOL Quality of Social Interaction
scores.
Thirty-three patients, 20 in the lower and 13 in the higher
target group, died during the study (P � 0.28). Table 5 shows
rates of treatment-emergent adverse events, vascular events,
access loss events, cardiac events, and death. The proportions
with at least one treatment-emergent adverse event were similar
in both target groups; more patients in the lower target
group experienced skeletal pain (P � 0.009), surgery (P �
0.013), and dizziness (P � 0.001), whereas more patients in the
higher target group experienced headache (P � 0.030) and
cerebrovascular events (P � 0.045).
Discussion
We studied patients who had recently started hemodialysis
therapy without overt cardiac disease. The fraction of incident
Lower
(9.5 to 11.5 g/dl)
Higher
(13.5 to 14.5 g/dl)
Between Group
Effect b
Page 118
2184 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2180–2189, 2005
LVVI (ml/m 2 ; n �mean � SD�)
baseline 300 (68.6 � 21.1) 296 (68.8 � 20.9)
week 24 254 (70.6 � 21.7) 255 (69.4 � 22.4)
week 48 222 (70.4 � 25.5) 223 (67.5 � 23.4)
week 96 169 (68.5 � 25.2) 170 (66.7 � 26.1)
last value 256 (69.0 � 25.2) 264 (68.4 � 25.7)
% change from baseline c (n �Est � SE�) 256 (8.3 � 4.7) 264 (7.6 � 4.7) �0.7 0.8726
LVMI (g/m 2 ; n �mean� SD�)
baseline 300 (111.9 � 33.2) 296 (116.6 � 35.5)
week 24 254 (121.3 � 37.2) 255 (121.7 � 40.1)
week 48 222 (122.8 � 42.5) 223 (124.2 � 38.9)
week 96 169 (128.2 (40.3) 166 (128.4 � 43.2)
last value 256 (126.3 � 40.9) 260 (126.8 � 42.4)
% change from baseline c (n �Est� SE�) 256 (16.8 � 3.5) 260 (14.2 � 3.5) �2.6 0.4353
Six-min walking test (m; N �mean� SD�)
baseline 277 (399 � 136) 284 (385 � 136)
week 24 233 (416 � 133) 242 (407 � 141)
week 48 200 (416 � 138) 203 (416 � 142)
week 96 143 (415 � 144) 142 (412 � 142)
last value 242 (416 � 141) 254 (410 � 144)
change from baseline d (n �Est� SE�) 237 (26 � 9) 249 (32 � 9) 6.6 0.4462
Congestive heart failure (N; n �%�) 300 (12 �4%�) 296 (11 �4%�) 0.8687 e
a
LVVI, left ventricular volume index; LVMI, left ventricular mass index.
b
Estimated treatment difference (higher target group minus lower target group).
c
Compared with analysis of covariance, adjusted for baseline left ventricular hypertrophy, gender, and epoetin � usage at
randomization.
d
Compared with analysis of covariance, adjusted for baseline gender and epoetin � usage at randomization.
e
Log-rank statistic from the time-to-event analysis.
dialysis patients in the Canadian echocardiographic cohort
study (1), with manifestations similar to the inclusion criteria of
the current study (no symptomatic heart failure or ischemic
heart disease and LVVI � 100 ml/m 2 ) was 43%. Hemoglobin
targets higher than the current guidelines had no effect on
cardiac dimensions, the incidence of congestive heart failure, or
performance on a 6-min walking test. QOL (in terms of vitality)
was improved. Adverse event rates were comparable in both
groups, except for skeletal pain, surgery, and dizziness being
significantly higher in the lower target group and headaches
and cerebrovascular events being higher in the higher target
group.
Several observational studies have shown associations between
anemia and adverse outcomes in dialysis patients (8–11).
Both uncontrolled (30,31) and controlled (14) studies have
shown that partial correction of anemia leads to partial regression
of LVH in dialysis patients, in addition to improving QOL
and physical performance (12,13). Three controlled studies of
normalization of hemoglobin with erythropoietin in hemodialysis
populations have been reported to date. The United States
Normal Hematocrit Trial differed from our study in several
P

Table 4. Quality-of-life outcomes for the lower and higher target groups a
Lower
(9.5 to 11.5 g/dl)
respects, including the requirement of pre-existing symptomatic
cardiac disease, a higher proportion of patients with diabetes,
patients on dialysis therapy an average of 3 yr or more,
the predominance of synthetic grafts for vascular access, and
the administration of substantially larger epoetin � doses in
both target groups. Primary outcome rates (death or first nonfatal
myocardial infarction) were higher in the higher target
groups, although this was not statistically significant. Although
rates of access loss were significantly higher in the normalized
group, QOL was better (18). The study was stopped early
because of significantly higher rates of vascular access thrombosis
and loss, together with 30% higher mortality in the high
hematocrit group. The Canadian Normalization of Hemoglobin
trial included patients with asymptomatic LVH or LV dilation.
The higher hemoglobin target failed to induce regression of
pre-existing LVH or dilation. Secondary outcome analysis suggested
that higher hemoglobin levels could prevent de novo left
ventricular dilation; in addition, QOL was improved without
Higher
(13.5 to 14.5 g/dl)
Page 119
J Am Soc Nephrol 16: 2180–2189, 2005 Anemia Correction in Hemodialysis Patients 2185
Between Group
N Est � SE N Est � SE Est � SE P b
FACIT-Fatigue d
week 24 291 69.7 � 1.0 291 70.9 � 1.0 1.3 � 1.2 0.28 0.56
week 36 291 68.9 � 1.0 291 70.5 � 1.0 1.6 � 1.1 0.16 0.32
week 48 291 68.1 � 1.0 291 70.0 � 1.0 1.9 � 1.2 1.00 0.20
week 60 291 67.3 � 1.0 291 69.5 � 0.1 2.3 � 1.3 0.08 0.16
week 72 291 66.4 � 1.2 291 69.1 � 1.2 2.6 � 1.5 0.08 0.15
week 84 291 65.6 � 1.3 291 68.6 � 1.3 3.0 � 1.7 0.08 0.16
week 96 291 64.8 � 1.4 291 68.1 � 1.5 3.3 � 1.9 0.09 0.17
mean follow-up minus baseline 291 �3.21 � 1.0 291 �1.0 � 1.0 2.3 � 1.3 0.08 0.16
SF-36 Vitality d
week 24 282 55.1 � 1.1 282 59.2 � 1.1 4.1 � 1.4 0.002 0.007
week 36 282 54.6 � 1.1 282 58.5 � 1.1 3.9 � 1.3 0.003 0.008
week 48 282 54.2 � 1.1 282 57.9 � 1.1 3.7 � 1.3 0.005 0.014
week 60 282 53.7 � 1.1 282 57.2 � 1.1 3.5 � 1.4 0.011 0.035
week 72 282 53.3 � 1.2 282 56.6 � 1.2 3.3 � 1.6 0.032 0.096
week 84 282 52.8 � 1.3 282 56.0 � 1.4 3.1 � 1.8 0.073 0.220
week 96 282 52.4 � 1.5 282 55.3 � 1.5 2.9 � 2.0 0.147 0.275
mean follow-up minus baseline 282 �2.31 � 1.08 282 1.21 � 1.08 3.5 � 1.4 0.012 0.036
KDQOL quality of social interaction d
week 24 282 67.6 � 1.4 282 68.2 � 1.4 0.5 � 1.3 0.68 0.68
week 36 282 67.4 � 1.3 282 68.0 � 1.3 0.6 � 1.2 0.62 0.62
week 48 282 67.1 � 1.3 282 67.8 � 1.3 0.7 � 1.2 0.58 0.56
week 60 282 66.9 � 1.3 282 67.7 � 1.3 0.8 � 1.3 0.56 0.58
week 72 282 66.7 � 1.4 282 67.5 � 1.4 0.8 � 1.4 0.56 0.56
week 84 282 66.4 � 1.4 282 67.3 � 1.5 0.9 � 1.6 0.56 0.56
week 96 282 66.2 � 1.5 282 67.1 � 1.6 1.0 � 1.7 0.57 0.58
mean follow-up minus baseline 282 0.03 � 1.0 282 0.8 � 1.0 0.8 � 1.30 0.56 0.56
a
FACIT, Functional Assessment of Chronic Illness Therapy; KDQOL, Kidney Disease Quality of Life.
b
Unadjusted P value.
c
The P value is adjusted for multiple comparisons across the three primary QOL scales using the Hochberg procedure.
d
Results from piecewise linear regression, where one slope was used from weeks 0 to 24 and another slope was used from
week 24 and beyond. This analysis assumed that missing QOL data were missing at random and unstructured covariance.
Figure 2. Mean of change in SF-36 Vitality scores from baseline in
the lower and higher target groups, presented � 2 SE. *P � 0.05.
P c

Table 5. Treatment-emergent adverse events that occurred in �10% of patients; vascular, access loss, and cardiac
events that occurred in �2% of patients; and death in lower and higher target groups a
Lower (9.5 to 11.5 g/dl;
n � 300)
Higher (13.5 to 14.5 g/dl;
n � 296)
Treatment emergent AE in �10% of patients
any AE
cardiovascular
281 (94%) 284 (96%)
hypertension 110 (37) 120 (41)
hypotension
platelet, bleeding, and clotting
105 (35) 85 (29)
arteriovenous fistula thrombosis 23 (8) 30 (10)
hematoma
vascular (extracardiac) disorders
36 (12) 45 (15)
arteriovenous fistula loss
gastrointestinal
26 (9) 30 (10)
vomiting 52 (17) 54 (18)
diarrhea 53 (18) 50 (17)
nausea 53 (18) 47 (16)
abdominal pain
respiratory
46 (15) 45 (15)
upper respiratory tract infection 69 (23) 72 (24)
dyspnea 42 (14) 35 (12)
cough 36 (12) 35 (12)
pharyngitis
musculoskeletal
31 (10) 29 (10)
myalgia 85 (28) 81 (27)
skeletal pain b
64 (21) 39 (13)
arthralgia
nervous system
43 (14) 36 (12)
64 (21) 86 (29)
headache b
dizziness b
40 (13) 16 (5)
skin
skin disorder 39 (13) 41 (14)
pruritus
other
33 (11) 23 (8)
infection 32 (11) 34 (11)
urinary tract infection 27 (9) 29 (10)
hyperparathyroidism 30 (10) 19 (6)
pain 47 (16) 41 (14)
back pain 40 (13) 35 (12)
fever 42 (14) 30 (10)
influenza-like symptoms 37 (12) 30 (10)
device complications 27 (9) 42 (14)
surgery b
39 (13) 20 (7)
Vascular, access loss, and cardiac events in �2% of patients
vascular
arteriovenous fistula thrombosis 36 (12) 45 (15)
Non–site-specific embolism thrombosis 12 (4) 14 (5)
permanent catheter thrombosis 9 (3) 8 (3)
cerebrovascular disorder b
4 (1) 12 (4)
peripheral ischemia 7 (2) 8 (3)
angina pectoris 8 (3) 9 (3)
myocardial infarction 4 (1) 7 (2)
chest pain 7 (2) 4 (1)
access loss
permanent catheter loss 6 (2) 7 (2)
arteriovenous fistula loss 27 (9) 30 (10)
arteriovenous graft loss 9 (3) 9 (3)
cardiac
angina pectoris 8 (3) 9 (3)
myocardial infarction 4 (1) 7 (2)
chest pain 7 (2) 4 (1)
tachycardia 15 (5) 22 (7)
palpitations 9 (3) 6 (2)
atrial fibrillation 7 (2) 6 (2)
bradycardia 5 (2) 7 (2)
pulmonary edema 16 (5) 9 (3)
cardiac failure 6 (2) 2 (1)
pulmonary edema or heart failure 19 (61) 11 (4)
death 20 (7) 13 (4)
Page 120
2186 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2180–2189, 2005
a AE, adverse events.
b P � 0.05 for all comparisons except headache (P � 0.030), skeletal pain (P � 0.009), surgery (P � 0.013), dizziness (P �
0.001), and cerebrovascular disorders (P � 0.045).

compromising vascular access (arteriovenous fistulas, predominantly)
(19). Another study used a double-blind crossover
design to compare hemoglobin targets of 10 and 14 g/dl, each
applied for 6 wk; the higher target led to reduced cardiac
output and improved QOL (21).
Two trials of higher hemoglobin targets, not restricted to
hemodialysis patients, have been reported. A Scandinavian
study with target hemoglobin levels of 9.0 to 12.0 or 13.5 to 16.0
g/dl included dialysis and nondialysis patients with chronic
kidney disease. QOL improved with the higher target, whereas
thrombovascular event rates, including those involving vascular
access, were similar (20). An Australian study compared
strategies of early or delayed anemia therapy targets in patients
with chronic kidney disease. Eligible patients had hemoglobin
levels of 11.0 to 12.0 g/dl (women) or 11.0 to 13.0 g/dl (men).
One group received epoetin � immediately, to maintain hemoglobin
between 12.0 and 13.0 g/L. In the other group, hemoglobin
was allowed to decline below 9.0 g/dl; epoetin � then
was used to maintain levels between 9.0 and 10.0 g/dl. The
mean hemoglobin levels achieved, in practice, were 12.1 and
10.8 g/dl, respectively. In this 2-yr study, changes in LVMI
were similar in both target groups (32). Thus, the available
literature suggests, as in our study, that enhanced QOL is the
only consistent benefit conferred by normalizing hemoglobin in
patients with chronic kidney disease.
Assessment of different adverse event rates should take account
of the fact that multiple comparisons were made. The
observation that a higher number of cerebrovascular events
occurred in the higher (n � 12) compared with the lower (n �
4) target group requires further study. Previous trials in chronic
kidney disease did not report a higher incidence of stroke in the
higher target group (18–21), but concern has been expressed
regarding thrombotic events during erythropoietin therapy in
oncology studies (33). The higher cerebrovascular event rate
was not supported by a higher death rate in the higher target
(n � 13) compared with the lower (n � 20) target group, and
other cardiovascular events rates were similar. The BP levels
were similar in both groups. Although a higher number of
antihypertensive agents were prescribed in the higher group
(compared with the lower group) at study end, the change in
number of antihypertensive agents from baseline was similar in
both groups.
Certain limitations in our study need to be considered. (1)
The study was not powered to examine major cardiovascular
events and death. This being said, our study used a randomized,
double-blind design; a reasonable duration of follow-up;
and sequential measurements in an incident cohort that started
hemodialysis. (2) The positive findings of our study—that
higher hemoglobin levels increase vitality—were generated
from secondary outcome analyses. The changes observed were
probably clinically important (34). (3) Our study, by design,
selected hemodialysis patients without symptomatic cardiac
disease and severe left ventricular dilation, and its findings can
be truly generalized only to such patients (approximately 40 to
50% of incident dialysis patients). As a result of the inclusion
criteria, the average age of the patients studied (51 yr) was
younger than the median age of incident patients in Canada in
Page 121
J Am Soc Nephrol 16: 2180–2189, 2005 Anemia Correction in Hemodialysis Patients 2187
2000 (65 yr), and the proportion with diabetes (18%) was lower
(44%) (35). This is reflected in the low mortality rate during the
study (5.5%). (4) Despite central monitoring with real-time
treatment recommendations, we approached but failed to reach
intended hemoglobin targets in the higher target group. However,
the separation by target group was approximately 2.3
g/dl throughout the maintenance phase of the study and was
compatible with the sample size assumption that a 15% difference
in LVVI changes would be associated with a 2-g/dl difference
in hemoglobin levels (19). Therefore, it seems unlikely
that the failure to achieve hemoglobin targets could fully account
for the similarity seen in the primary outcome comparisons.
(5) Although we observed no difference in cardiovascular,
access, and malignancy rates, the actual event rates were low,
and an untreated placebo group was not used. Thus, small or
moderate differences in event rates cannot be excluded. (6) The
high dropout rate was anticipated because patients such as
those in our study are usually referred and then wait for renal
transplantation. No differences in the primary outcome were
observed whether analyzed by changes from baseline to last
value or by analysis of serial measurements.
Despite its limitations, we believe that this randomized, controlled
trial, in enrolling a large number of incidence hemodialysis
patients, in achieving a clinically important difference in
hemoglobin levels, and in serial follow-up of several clinically
important outcomes over 2 yr, provides important information
regarding the question of normalizing hemoglobin in dialysis
patients. It suggests that normalization of hemoglobin in patients
who start hemodialysis, although improving vitality, has
no effect on left ventricular size when compared with patients
who are maintained with hemoglobin levels between 9.5 and
11.5 g/dl.
Acknowledgments
This paper was presented at the American Society of Nephrology
Clinical Trials Symposium, November 1, 2004, St. Louis, MO (J AmSoc
Nephrol 15: 726A, 2004).
We are grateful to Janet Morgan in Canada and to Aileen Foley in
England for coordinating patient enrollment and management.
P.S.P. and R.N.F., the co-principal investigators, designed the trial,
applied for funding to Johnson and Johnson, coordinated the study,
and wrote the report. D.J.S. performed the statistical analyses and
contributed to preparation of the manuscript. B.H.W. contributed to the
study design, collection of data, and preparation of the manuscript.
M.J.Z. contributed to the study design with respect to QOL analysis, to
analysis and interpretation of the data, and to the manuscript. D.F.
contributed to the study design, to the collection and analysis of the
data, and to the manuscript.
Members of the Canadian European Study Group
EPO-INT-68 Independent Data Monitoring Committee Members:
L.J. Wei, Boston, MA; M.-M. Samama, Paris, France; P. Ivanovich,
Chicago, IL; M.A. Pfeffer, Boston, MA
Principal Investigator/Site: W. Hoerl, Wien, Austria; H.-K. Stummvoll,
Linz, Austria; G. Mayer, Innsbruck, Austria; H. Graf, Wien,
Austria; H. Holzer, Graz, Austria; Y. Vanrenterghem, Leuven, Belgium;
W. Lornoy, Aalst, Belgium; C. Tielemans, Bruxelles, Belgium; M.
Jadoul, Bruxelles, Belgium; P. Parfrey, St. John’s, Canada; P. Barre,

Montréal, Canada; A. Levin, Vancouver, Canada; P. Cartier, Montréal,
Canada; N. Muirhead, London, Canada; A. Fine, Winnipeg, Canada; B.
Murphy, Calgary, Canada; S. Handa, St. John’s, Canada; P. Campbell,
Edmonton, Canada; V. Pichette, Montreal, Canada; S. Tobe, Toronto,
Canada; C. Lok, Toronto, Canada; D. Kates, Kelowna, Canada; D.
Holland, Kingston, Canada; G. Karr, Penticton, Canada; G. Pylpchuk,
Saskatoon, Canada; G. Wu, Mississauga, Canada; M. Vasilevsky, Montreal,
Canada; E. Carisle, Hamilton, Canada; E.R. Gagne, Fleurimont,
Canada; W. Callaghan, Windsor, Canada; G. Soltys, Greenfield Park,
Canada; P. Tam, Scarborough, Canada; R. Turcot, Trios-Rivieres, Canada;
M. Berrall, Toronto, Canada; J. Zacharias, Winnipeg, Canada; S.
Donnelly, Toronto, Canada; G. London, Fleury-Merogis, France; A.
London, Aulnay sous Bois, France; F.P. Wambergue, Lille, France; H.
Geiger, Frankfurt, Germany; V. Kliem, Hann Muenden, Germany; R.
Winkler, Rostock, Germany; B. Kraemer, Regensburg, Germany; H.
Schiffl, Munich, Germany; R. Brunkhorst, Hannover, Germany; D. Seybold,
Bayreuth, Germany; M. Hilfenhaus, Langenhagen, Germany; D.
Schaumann, Hameln, Germany; R. Goetz, Bad Windsheim, Germany;
P. Roch, Regensburg, Germany; H.-P. Brasche, Ludwigshafen, Germany;
V. Wizemann, Giessen, Germany; K. Bittner, Ansbach, Germany;
K. Appen, Hamburg, Germany; B. Schroeder, Bad Toelz, Germany; W.
Schropp, Munich, Germany; D. O’Donoghue, Salford, England; I. Mac-
Dougall, London, England; G. Warwick, Leicester, England; M. Raftery,
London, England; K. Farrington, Stevenage, England; J. Kwan, Carshalton,
England; P. Conlon, Dublin, Ireland; G. Mellotte, Dublin, Ireland;
K. Siamopoulos, Ioannina, Greece; N. Tsaparas, Crete, Greece; D. Tsakiris,
Veria, Greece; V. Vargemezis, Alexandroupolis, Greece; S. Ferenczi,
Gyor, Hungary; S. Gorogh, Kisvarda, Hungary; I. Kulcsar,
Szombathely, Hungary; L. Locsey, Debrecen, Hungary; K. Akocsi,
Veszprem, Hungary; I. Solt, Szekesfehervar, Hungary; O. Arkossy,
Budapest, Hungary; E. Kiss, Szeged, Hungary; J. Manitius, Bydgoszcz,
Poland; B. Rutkowski, Gdansk, Poland; A. Wiecek, Katowice, Poland;
W. Sulowicz, Krakow, Poland; A. Ksiazek, Lublin, Poland; S. Czekalski,
Poznan, Poland; M. Klinger, Wroclaw, Poland; M. Mysliwiec, Bialystok,
Poland; H. Augustyniak-Bartosik, Milicz, Poland; J. Imiela, Warszawa-
Miedzylesie, Poland; A. Sydor, Tarnow, Poland; R. Rudka, Bytom,
Poland; M. Kiersztejn, Chrzanow, Poland; R. Wnuk, Oswiecim, Poland;
A. Milkowsk, Krakow, Poland; F. Valderrabano, Madrid, Spain; P.
Aljama, Córdoba, Spain; H. Alcocer, Valencia, Spain; A. Purroy, Pamplona,
Spain
References
1. Foley RN, Parfrey PS, Harnett JD, Kent GM, Martin CJ,
Murray DC, Barre PE: Clinical and echocardiographic disease
in patients starting end-stage renal disease therapy.
Kidney Int 47: 186–192, 1995
2. Acquatella H, Perez-Rojas M, Burger B, Guinand-Baldo A:
Left ventricular function in terminal uremia. A hemodynamic
and echocardiographic study. Nephron 22: 160–174,
1978
3. London GM, Pannier B, Guerin AP, Blacher J, Marchais SJ,
Darne B, Metivier F, Adda H, Safar ME: Alterations of left
ventricular hypertrophy in and survival of patients receiving
hemodialysis: Follow-up of an interventional study.
J Am Soc Nephrol 12: 2759–2767, 2001
4. Parfrey PS, Foley RN, Harnett JD, Kent GM, Murray DC,
Barre PE: Outcome and risk factors for left ventricular
disorders in chronic uraemia. Nephrol Dial Transplant 11:
1277–1285, 1996
5. Silberberg JS, Barre PE, Prichard SS, Sniderman AD: Im-
Page 122
2188 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2180–2189, 2005
pact of left ventricular hypertrophy on survival in endstage
renal disease. Kidney Int 36: 286–290, 1989
6. Stack AG, Saran R: Clinical correlates and mortality impact
of left ventricular hypertrophy among new ESRD patients
in the United States. Am J Kidney Dis 40: 1202–1210, 2002
7. Zoccali C, Benedetto FA, Mallamaci F, Tripepi G, Giacone
G, Cataliotti A, Seminara G, Stancanelli B, Malatino LS;
CREED Investigators: Prognostic impact of the indexation
of left ventricular mass in patients undergoing dialysis.
J Am Soc Nephrol 12: 2768–2774, 2004
8. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC,
Barre PE: The impact of anemia on cardiomyopathy, morbidity,
and mortality in end-stage renal disease. Am J Kidney
Dis 28: 53–61, 1996
9. Ma JZ, Ebben J, Xia H, Collins AJ: Hematocrit level and
associated mortality in hemodialysis patients. J Am Soc
Nephrol 10: 610–619, 1999
10. Madore F, Lowrie EG, Brugnara C, Lew NL, Lazarus JM,
Bridges K, Owen WF: Anemia in hemodialysis patients:
Variables affecting this outcome predictor. J Am Soc Nephrol
8: 1921–1929, 1997
11. Xia H, Ebben J, Ma JZ, Collins AJ: Hematocrit levels and
hospitalization risks in hemodialysis patients. J Am Soc
Nephrol 10: 1309–1316, 1999
12. Canadian Erythropoietin Study Group: Association between
recombinant human erythropoietin and quality of
life and exercise capacity of patients receiving haemodialysis.
BMJ 300: 573–578, 1990
13. McMahon LP, Johns JA, McKenzie A, Austin M, Fowler R,
Dawborn JK: Haemodynamic changes and physical performance
at comparative levels of haemoglobin after longterm
treatment with recombinant erythropoietin. Nephrol
Dial Transplant 7: 1199–1206, 1992
14. Sikole A, Polenakovic M, Spirovska V, Polenakovic B, Masin
G: Analysis of heart morphology and function following
erythropoietin treatment of anemic dialysis patients.
Artif Organs 17: 977–984, 1993
15. National Kidney Foundation Kidney Disease Outcomes
Quality Initiative (NKF K/DOQI): Guidelines for anemia
of chronic kidney disease. Available: www.kidney.org/
professionals/kdoqi/guidelines/doqiupan_ii.html. Accessed
March 2, 2005
16. Churchill DN, Blake PG, Jindal KK, Toffelmire EB, Goldstein
MB: Clinical practice guidelines for initiation of dialysis.
Canadian Society of Nephrology. J Am Soc Nephrol
10[Suppl 13]: S289–S291, 1999
17. Locatelli F: European best practice guidelines II: Targets
for anaemia treatment. Nephrol Dial Transplant 19[Suppl 2]:
ii6–ii15, 2004
18. Besarab A, Bolton WK, Browne JK, Egrie JC, Nissenson AR,
Okamoto DM, Schwab SJ, Goodkin DA: The effects of
normal as compared with low hematocrit values in patients
with cardiac disease who are receiving hemodialysis
and epoetin. N Engl J Med 339: 584–590, 1998
19. Foley RN, Parfrey PS, Morgan J, Barre PE, Campbell P,
Cartier P, Coyle D, Fine A, Handa P, Kingma I, Lau CY,
Levin A, Mendelssohn D, Muirhead N, Murphy B, Plante
RK, Posen G, Wells GA: Effect of hemoglobin levels in
hemodialysis patients with asymptomatic cardiomyopathy.
Kidney Int 58: 1325–1335, 2000
20. Furuland H, Linde T, Ahlmen J, Christensson A, Strombom
U, Danielson BG: A randomized controlled trial of

J Am Soc Nephrol 16: 2180–2189, 2005 Anemia Correction in Hemodialysis Patients 2189
haemoglobin normalization with epoetin alfa in pre-dialysis
and dialysis patients. Nephrol Dial Transplant 18: 353–
361, 2003
21. McMahon LP, Mason K, Skinner SL, Burge CM, Grigg
LE, Becker GJ: Effects of haemoglobin normalization on
quality of life and cardiovascular parameters in endstage
renal failure. Nephrol Dial Transplant 15: 1425–1430,
2000
22. Hays RD, Kallich JD, Mapes DL, Coons SJ, Carter WB:
Development of the kidney disease quality of life
(KDQOL) instrument. Qual Life Res 3: 329–338, 1994
23. Cella D: Manual of the Functional Assessment of Chronic
Illness Therapy (FACIT) Measurement System, Version 4,
Evanston, Center on Outcomes, Research and Education
(CORE), Evanston Northwestern Healthcare and Northwestern
University, 1997
24. Fitts SS, Guthrie MR: Six-minute walk by people with
chronic renal failure. Assessment of effort by perceived
exertion. Am J Phys Med Rehabil 74: 54–58, 1995
25. Sahn DJ, DeMaria A, Kisslo J, Weyman A: Recommendations
regarding quantitation in M-mode echocardiography:
Results of a survey of echocardiographic measurements.
Circulation 58: 1072–1083, 1978
26. Pombo JF, Troy BL, Russell RO Jr: Left ventricular volumes
and ejection fraction by echocardiography. Circulation 43:
480–490, 1971
27. Devereux RB: Detection of left ventricular hypertrophy by
M-mode echocardiography. Anatomic validation, standardization,
and comparison to other methods. Hypertension
9: II9–II26, 1987
28. Lin CL, Hsu PY, Yang CT, Yang HY, Yang TY, Huang WH,
Huang CC: LV mass index significantly impacts on patient
and renal outcomes in patients with coronary artery bypass
grafting and poor left-ventricular function. Ren Fail
25: 287–295, 2003
29. Proschan MA, Waclawiw MA: Practical guidelines for
multiplicity adjustment in clinical trials. Control Clin Trials
21: 527–539, 2000
30. London GM, Zins B, Pannier B, Naret C, Berthelot JM,
Jacquot C, Safar M, Drueke TB: Vascular changes in hemodialysis
patients in response to recombinant human erythropoietin.
Kidney Int 36: 878–882, 1989
31. Macdougall IC, Lewis NP, Saunders MJ, Cochlin DL, Davies
ME, Hutton RD, Fox KA, Coles GA, Williams JD:
Long-term cardiorespiratory effects of amelioration of renal
anaemia by erythropoietin. Lancet 335: 489–493, 1990
32. Roger SD, McMahon LP, Clarkson A, Disney A, Harris D,
Hawley C, Healy H, Kerr P, Lynn K, Parnham A, Pascoe R,
Voss D, Walker R, Levin A: Effects of early and late intervention
with epoetin alpha on left ventricular mass among patients
with chronic kidney disease (stage 3 or 4): Results of a
randomized clinical trial. J Am Soc Nephrol 15: 148–156, 2004
33. Leyland-Jones B: Breast cancer trial with erythropoietin
terminated unexpectedly. Lancet Oncol 4: 459–460, 2003
34. Hays RD, Morales LS: The RAND-36 measure of healthrelated
quality of life. Ann Med 33: 350–357, 2001
35. Canadian Organ Replacement Register: Preliminary report
for dialysis and transplantation 2002, Canadian
Institute for Health Information, Ottawa, Canada 2002.
Available: secure.cihi.ca/cihiweb/en/downloads/services_
corr_ e_2002prelimreport.pdf. Accessed March 2, 2005
Access to UpToDate on-line is available for additional clinical information
at http://www.jasn.org/
Page 123

The new england
journal of medicine
established in 1812 november 19, 2009 vol. 361 no. 21
A Trial of Darbepoetin Alfa in Type 2 Diabetes
and Chronic Kidney Disease
Marc A. Pfeffer, M.D., Ph.D., Emmanuel A. Burdmann, M.D., Ph.D., Chao-Yin Chen, Ph.D., Mark E. Cooper, M.D.,
Dick de Zeeuw, M.D., Ph.D., Kai-Uwe Eckardt, M.D., Jan M. Feyzi, M.S., Peter Ivanovich, M.D.,
Reshma Kewalramani, M.D., Andrew S. Levey, M.D., Eldrin F. Lewis, M.D., M.P.H., Janet B. McGill, M.D.,
John J.V. McMurray, M.D., Patrick Parfrey, M.D., Hans-Henrik Parving, M.D., Giuseppe Remuzzi, M.D.,
Ajay K. Singh, M.D., Scott D. Solomon, M.D., and Robert Toto, M.D., for the TREAT Investigators*
ABSTRACT
BACKGROUND
Anemia is associated with an increased risk of cardiovascular and renal events
among patients with type 2 diabetes and chronic kidney disease. Although darbepoetin
alfa can effectively increase hemoglobin levels, its effect on clinical outcomes
in these patients has not been adequately tested.
METHODS
In this study involving 4038 patients with diabetes, chronic kidney disease, and
anemia, we randomly assigned 2012 patients to darbepoetin alfa to achieve a hemoglobin
level of approximately 13 g per deciliter and 2026 patients to placebo, with
rescue darbepoetin alfa when the hemoglobin level was less than 9.0 g per deciliter.
The primary end points were the composite outcomes of death or a cardiovascular
event (nonfatal myocardial infarction, congestive heart failure, stroke, or hospitalization
for myocardial ischemia) and of death or end-stage renal disease.
RESULTS
Death or a cardiovascular event occurred in 632 patients assigned to darbepoetin alfa
and 602 patients assigned to placebo (hazard ratio for darbepoetin alfa vs. placebo,
1.05; 95% confidence interval [CI], 0.94 to 1.17; P = 0.41). Death or end-stage renal
disease occurred in 652 patients assigned to darbe poetin alfa and 618 patients assigned
to placebo (hazard ratio, 1.06; 95% CI, 0.95 to 1.19; P = 0.29). Fatal or nonfatal
stroke occurred in 101 patients assigned to darbepoetin alfa and 53 patients assigned
to placebo (hazard ratio, 1.92; 95% CI, 1.38 to 2.68; P

2020
Type 2 diabetes mellitus and chronic
kidney disease frequently coexist, and each
disease independently increases the risk of
cardiovascular events and end-stage renal disease.
1,2 Intensive treatment of concomitant conventional
risk factors such as hypertension and
elevated levels of low-density lipoprotein reduces
fatal and nonfatal cardiovascular complications
and slows the progression of kidney disease. 3-8
Observational studies suggest that anemia can
be considered another biomarker of risk, since a
lower hemoglobin level is independently associated
with a higher rate of cardiovascular and renal
events, 9-11 especially among patients with
diabetes. 12,13 Whether the use of erythropoiesisstimulating
agents (ESAs) to increase hemoglobin
levels lowers this risk is not known. 14
Early studies involving the use of recombinant
human erythropoietin in patients with severe
anemia who were undergoing dialysis showed a
reduced requirement for blood transfusions and
improved quality-of-life assessments, which were
considered major advances. 15,16 Extrapolations
from these favorable effects to other populations,
including patients with anemia and earlier stages
of chronic kidney disease, led to the wider use
of ESAs. The presumption of the benefits associated
with the use of ESAs was so pervasive that
major clinical trials considered the use of placebo
unnecessary or even unethical. 17-19 In fact,
there were no placebo-controlled trials in a recent
meta-analysis of the use of ESAs in patients with
chronic kidney disease. 20 Although ESAs have
been available for more than two decades, the
hypothesis that this treatment approach would
lower the risks associated with anemia and improve
the prognosis has not been examined. 21
We conducted the Trial to Reduce Cardiovascular
Events with Aranesp Therapy (TREAT) to test
the hypothesis that, in patients with type 2 diabetes,
chronic kidney disease that does not require
dialysis, and concomitant anemia, increasing
hemoglobin levels with the use of this agent
would lower the rates of death, cardiovascular
events, and end-stage renal disease. 22
Methods
The TREAT was a randomized, double-blind,
placebo-controlled trial conducted at 623 sites in
24 countries (see the Supplementary Appendix,
available with the full text of this article at NEJM.
org). The institutional review board or ethics
The new england journal of medicine
n engl j med 361;21 nejm.org november 19, 2009
committee at each site approved the protocol,
and all patients provided written informed consent.
Enrollment occurred from August 25, 2004,
through December 4, 2007.
The trial was sponsored by Amgen and designed
in collaboration with an academic executive
committee. Data collection and management
were the responsibility of the sponsor, with oversight
by the executive committee. The Statistical
Data Analysis Center at the University of Wisconsin
prepared the interim unblinded reports
for the independent data and safety monitoring
committee, which oversaw patient safety and the
quality of trial conduct. The independent academic
statistical center also performed all analyses
for this article and provided the writing
group with unrestricted access to the data. The
manuscript was prepared by the principal investigator
and subsequently revised and edited by
all the authors. All the authors decided to submit
the article for publication and assume responsibility
for the accuracy and completeness of the
reported data.
Study Population
Patients with type 2 diabetes, chronic kidney disease
(an estimated glomerular filtration rate [GFR]
of 20 to 60 ml per minute per 1.73 m 2 of bodysurface
area, calculated with the use of the fourvariable
Modification of Diet in Renal Disease
formula), anemia (hemoglobin level, ≤11.0 g per
deciliter), and a transferrin saturation of 15% or
more were eligible for enrollment. Patients with
any of the following factors were excluded: uncontrolled
hypertension; previous kidney transplantation
or scheduled receipt of a kidney transplant
from a living related donor; current use of
intravenous antibiotics, chemotherapy, or radiation
therapy; cancer (except basal-cell or squamouscell
carcinoma of the skin); diagnosed human
immunodeficiency virus infection; active bleeding;
a hematologic disease; or pregnancy. Patients
who had had a cardiovascular event or grand mal
seizure, had undergone major surgery, or had received
an ESA in the 12 weeks before randomization
were also ineligible.
Study Procedures
During a 2-week screening period, prestudy laboratory
assessments were obtained and the eligibility
of patients who had provided consent was
confirmed. Patients were randomly assigned with
the use of a computer-generated, permuted-block
Downloaded from www.nejm.org at Amgen, Inc. on November 20, 2009 .
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
Page 125

Darbepoetin Alfa Therapy for reduction of Cardiovascular Events
design, in a 1:1 ratio, to receive darbepoetin alfa
(Aranesp, Amgen) or placebo. Randomization was
stratified according to the study site, the baseline
level of proteinuria (with marked proteinuria defined
as a ratio of total protein to creatinine [calculated
in milligrams per deciliter] of ≥1.0 in a
spot urine sample), and an investigator-designated
history of cardiovascular disease. Darbepoetin
alfa and placebo were supplied in matching prefilled
syringes at 12 different strengths. A third
party used a point-of-care device to monitor hemoglobin
levels and enter the value into an interactive
voice-response system that selected the dosage
according to a computer algorithm (see the
Supplementary Appendix). This algorithm was
designed to adjust the dose in order to maintain
the hemoglobin level at approximately 13.0 g
per deciliter in the patients assigned to darbepoetin
alfa.
Patients in the placebo group were assigned
to receive darbepoetin alfa as a rescue agent if
the hemoglobin level fell below 9.0 g per deciliter,
with a return to placebo once the hemoglobin
level was 9.0 g per deciliter or higher. The site
investigator was to be notified if any patient had
a hemoglobin value of 7.0 g per deciliter or less, a
value of 16.0 g per deciliter or more, or a decrease
of 2.0 g per deciliter or more in a 4-week period.
Measurements of hemoglobin levels and vital
signs were performed every 2 weeks during the
study-drug–titration period and monthly thereafter.
Transferrin saturation and ferritin levels
were measured quarterly; other laboratory assessments
and spot urine collections were performed
at 24-week intervals. Patient-reported outcomes
and health resource utilization were assessed at
weeks 1, 13, and 25 and every 24 weeks thereafter.
At each visit, information was gathered regarding
adverse events, hospitalization, transfusions,
and the concomitant use of other medications.
Evaluation of Outcomes
The primary end points were the time to the
composite outcome of death from any cause or a
cardiovascular event (nonfatal myocardial infarction,
congestive heart failure, stroke, or hospitalization
for myocardial ischemia) and the time to
the composite outcome of death or end-stage renal
disease (see the Supplementary Appendix).
The overall alpha level for the primary end points
was split: 0.048 for the cardiovascular composite
outcome and 0.002 for the renal composite outcome.
All potential end points were adjudicated
by a clinical end-point committee whose members
were unaware of the treatment assignments,
and the hematocrit and hemoglobin values were
redacted from the documents under review. Important
secondary end points included time to
death, death from cardiovascular causes, and the
components of the primary end points, as well as
the rate of decline in the estimated GFR and
changes in patient-reported outcomes at week 25
measured with the use of the Functional Assessment
of Cancer Therapy–Fatigue (FACT-Fatigue)
instrument (on which scores range from 0 to 52,
with higher scores indicating less fatigue) and the
36-Item Short-Form General Health Survey questionnaire
(calculated with norm-based scoring so
that 50 is the average score, with higher scores
indicating a better quality of life) (added after the
second protocol amendment in May 2005).
Monitoring
The independent data and safety monitoring committee
reviewed safety reports monthly with formal
interim analyses scheduled when approximately
20, 40, 60, and 80% of the total expected
cardiovascular composite events had occurred.
Initial guidelines for early discontinuation included
a symmetric O’Brien–Fleming alpha-spending
function. In April 2006, data from a non–
placebo-controlled trial of a different ESA in
patients with chronic kidney disease showed that
treatment aimed at achieving a higher hemoglobin
level (13.5 g per deciliter vs. 11.3 g per deciliter)
was associated with increased rates of adverse
clinical events. 17 This information and more
equivocal results from a related, smaller trial 18
were proactively communicated to our patients,
investigators, and the data and safety monitoring
committee. The committee made no recommendation
to alter or terminate the study on the basis
of interim data that already reflected more clinical
events than the numbers of events in these
related trials. The consent form was modified
with the addition of the new information regarding
potential harm, and all study participants
were asked to provide written informed consent
again. The executive committee requested that
the data and safety monitoring committee adopt
a more cautious stopping rule, with the use of a
one-sided alpha of 0.05 for harm at each assessment
(without adjustment for multiple testing)
for either the cardiovascular composite outcome
or death. The boundaries for efficacy remained
unchanged, and with this new mandatory stop-
n engl j med 361;21 nejm.org november 19, 2009 2021
Downloaded from www.nejm.org at Amgen, Inc. on November 20, 2009 .
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
Page 126

2022
ping rule for harm, the committee made no recommendation
to terminate the trial before its
scheduled conclusion.
Statistical Analysis
The trial was event driven, with a total of 1203
cardiovascular composite events required to provide
80% statistical power to detect a 20% risk
reduction for this outcome, with a two-sided type
I error of 0.048 (unadjusted for interim analyses).
We aimed to enroll approximately 4000 patients;
assumptions included an annualized rate of events
in the placebo group of 12.5%, a 15% loss to
follow-up, and attenuation of the treatment effect
due to the anticipated use of ESAs in patients
who had progression to end-stage renal disease.
Time-to-event analyses were performed with
the use of life-table methods according to the
intention-to-treat principle. Kaplan–Meier cumulative
incidence curves were compared with the
use of a two-sided log-rank test, stratified according
to the baseline proteinuria level and the presence
or absence of a history of cardiovascular
disease. Estimated hazard ratios (for darbepoetin
alfa vs. placebo) and 95% confidence intervals
were obtained with the use of stratified Cox
proportional-hazards models. Patients who discontinued
either study drug early, including patients
who began treatment with dialysis or underwent
transplantation, were followed for study
end points.
Categorization of adverse events was based on
groupings of preferred terms defined by the
standardized queries in the Medical Dictionary
for Regulatory Activities. 23 Analyses of patientreported
outcomes were prespecified to impute
missing data with the last-observation-carriedforward
method and to use t-tests for treatment
comparisons, with means and standard deviations.
Reported P values are nominal and have
not been adjusted for multiple comparisons. Additional
statistical methods are described in the
Supplementary Appendix.
Results
Patients
We enrolled a total of 4047 patients, but before
unblinding, we excluded all information regarding
9 patients from two sites that did not adhere
to Good Clinical Practice guidelines (4 patients
who were randomly assigned to darbepoetin alfa
and 5 patients who were randomly assigned to
The new england journal of medicine
n engl j med 361;21 nejm.org november 19, 2009
placebo). Of the 4038 patients evaluated, 2012
were assigned to receive darbepoetin alfa and
2026 were assigned to receive placebo. The study
was event driven and was completed on March
28, 2009, with a median follow-up duration of
29.1 months. At that time, 3523 patients (87.2%)
were either still being followed for clinical end
points or had died; this group included 1761 patients
in the darbepoetin alfa group (87.5%) and
1762 patients in the placebo group (87.0%). The
vital status at study termination was unknown
for 153 patients in the darbepoetin alfa group
(7.6%) and 164 in the placebo group (8.1%) (see
the Consolidated Standards for the Reporting of
Trials [CONSORT] diagram in the Supplementary
Appendix).
The median age was 68 years, and 57.3% of
the patients were women; 65.4% of the patients
had a history of cardiovascular disease, with
similar percentages of patients in each group reporting
a history of coronary artery disease,
stroke, peripheral arterial disease, and myocardial
infarction. There was an imbalance in the
proportion of patients with a history of heart
failure (31.5% in the darbepoetin alfa group vs.
35.2% in the placebo group, unadjusted P = 0.01).
There were no clinically meaningful baseline imbalances
in vital signs, laboratory data, or the
use of medications for cardiovascular disease and
diabetes (Table 1).
Intervention and Hemoglobin Values
The overall median hemoglobin level at baseline
was 10.4 g per deciliter (interquartile range, 9.8 to
10.9). Between-group differences in hemoglobin
values were apparent less than 1 month after randomization,
and the differences were significant
by 1 month (Fig. 1). From 3 months to the end of
treatment, the median achieved hemoglobin level
(based on the area under the curve) was 12.5 g
per deciliter (interquartile range, 12.0 to 12.8) in
the darbepoetin alfa group and 10.6 g per deciliter
(interquartile range, 9.9 to 11.3) in the placebo
group (P

Darbepoetin Alfa Therapy for reduction of Cardiovascular Events
Table 1. Baseline Characteristics of the Patients.*
Characteristic
Darbepoetin Alfa
(N = 2012)
Placebo
(N = 2026) P Value†
Age (yr) 0.22
Median 68 68
Interquartile range 60–75 60–75
Female sex (%) 58.5 56.0 0.10
Race or ethnic group (%)‡ 0.84
White 63.1 64.2
Black 20.6 19.8
Hispanic 13.6 13.1
Other 2.7 3.0
Known duration of diabetes (yr) 0.94
Median 15.3 15.5
Interquartile range 8.2–21.8 8.4–21.7
Body-mass index§ 0.11
Median 30.5 30.1
Interquartile range 26.3–35.5 26.2–34.9
Retinopathy (%) 48.2 45.7 0.12
Laser treatment (%) 29.6 28.6 0.49
Diabetic neuropathy (%) 49.2 46.4 0.07
Amputation 5.7 6.1 0.58
History of cardiovascular disease (%)¶ 64.0 66.9 0.05
Coronary artery disease‖ 43.2 45.5 0.16
Heart failure 31.5 35.2 0.01
Myocardial infarction 18.4 18.3 0.95
Stroke 11.5 10.7 0.41
Peripheral arterial disease** 21.2 20.8 0.73
Use of pacemaker 4.5 6.1 0.02
Use of automatic implantable cardioverter–defibrillator 1.4 1.4 0.87
Atrial fibrillation (%) 11.0 10.0 0.29
Smoking status (%) 0.88
Current smoker 5.4 4.7
Former smoker 38.2 39.4
Blood pressure (mm Hg)
Systolic 0.25
Median 136 135
Interquartile range 122–149 123–148
Diastolic 0.55
Median 71 70
Interquartile range 64–80 64–80
Serum creatinine (mg/dl) 0.01
Median 1.8 1.9
Interquartile range 1.5–2.3 1.5–2.4
Estimated GFR (ml/min/1.73 m 2 ) 0.03
Median 34 33
Interquartile range 27–43 26–42
n engl j med 361;21 nejm.org november 19, 2009 2023
Downloaded from www.nejm.org at Amgen, Inc. on November 20, 2009 .
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
Page 128

2024
Table 1. (Continued.)
Characteristic
The new england journal of medicine
Darbepoetin Alfa
(N = 2012)
n engl j med 361;21 nejm.org november 19, 2009
Placebo
(N = 2026) P Value†
Ratio of total protein (in mg/dl) to creatinine (in mg/dl) in urine¶ 0.73
Median 0.4 0.4
Interquartile range 0.1–2.0 0.1–1.8

Table 1. (Continued.)
Characteristic
Medications (%)
Darbepoetin Alfa Therapy for reduction of Cardiovascular Events
poetin alfa. Among the patients who had not died
or progressed to end-stage renal disease, 93.9%
of the patients in the darbepoetin alfa group and
90.4% of those in the placebo group were receiving
the assigned treatment at 6 months; 87.4%
and 83.7%, respectively, at 1 year; and 74.3% and
69.3%, respectively, at 2 years. There was no significant
difference in the proportions of patients
receiving oral iron during the study (66.8% of
those in the darbepoetin alfa group and 68.6% of
those in the placebo group, P = 0.25); however,
more patients in the placebo group received intravenous
iron (20.4%, vs. 14.8% of patients assigned
to darbepoetin alfa; P

2026
Mean Hemoglobin (g/dl)
No. of Patients
Darbepoetin alfa
Placebo
13.5
13.0
12.5
12.0
11.5
11.0
10.5
10.0
9.5
between-group differences in the time to the first
occurrence of the following components, when
considered individually: death from any cause,
fatal or nonfatal congestive heart failure, fatal or
nonfatal myocardial infarction, or hospitalization
for myocardial ischemia (Table 2 and Fig. 2B, 2C,
and 2D). However, fatal or nonfatal stroke was
more likely to occur in the patients assigned to
darbepoetin alfa (101 patients [5.0%] vs. 53 patients
[2.6%]; hazard ratio, 1.92; 95% CI, 1.38 to
2.68; P

Darbepoetin Alfa Therapy for reduction of Cardiovascular Events
Table 2. Composite and Component End Points.*
End Point
Darbepoetin Alfa
(N = 2012)
1.95); among the 2570 white patients, the hazard
ratio was 1.05 (95% CI, 0.92 to 1.21), and among
the 815 black patients, the hazard ratio was 0.87
(95% CI, 0.68 to 1.10; P = 0.03 for the interaction,
uncorrected for multiple comparisons).
Patient-Reported Outcomes
The primary prespecified analysis for the patientreported
outcomes was the change from baseline
to 25 weeks in the FACT-Fatigue score. Among
patients with both baseline and week-25 scores,
from a baseline score of 30.2 in the group of 1762
patients assigned to darbepoetin alfa and a baseline
score of 30.4 in the 1769 patients assigned to
placebo, there was a greater degree of improvement
in the mean (±SD) score in the darbepoetin
alfa group than in the placebo group (an increase
of 4.2±10.5 points vs. 2.8±10.3 points, P

B Death from Any Cause
50
Hazard ratio, 1.05 (95% CI, 0.92–1.21)
40 P=0.48
A Cardiovascular Composite End Point
50
Hazard ratio, 1.05 (95% CI, 0.94–1.17)
40 P=0.41
Darbepoetin alfa
30
2028
Darbepoetin alfa
30
Patients with Event (%)
Placebo
Placebo
20
20
10
10
0
0
Patients with Event (%)
0 6 12 18 24 30 36 42 48
Months since Randomization
0 6 12 18 24 30 36 42 48
Months since Randomization
164
156
386
385
655
636
945
970
1337
1345
1659
1652
1847
1839
1947
1943
2012
2026
No. at Risk
Darbepoetin alfa
Placebo
130
122
318
319
551
529
817
834
1180
1178
1515
1487
1717
1687
1882
1836
2012
2026
No. at Risk
Darbepoetin alfa
Placebo
C Fatal or Nonfatal Congestive Heart Failure
50
Hazard ratio, 0.89 (95% CI, 0.74–1.08)
40 P=0.24
The new england journal of medicine
D Fatal or Nonfatal Myocardial Infarction and Myocardial Ischemia
50 Fatal or Nonfatal
Myocardial Infarction Myocardial Ischemia
40
Hazard ratio, 0.96 (95% CI, Hazard ratio, 0.84 (95% CI,
0.75–1.22)
0.55–1.27)
30
P=0.73
P=0.40
20
Darbepoetin alfa
Darbepoetin alfa
Placebo
Placebo
Patients with Event (%)
30
20
10
Darbepoetin alfa
Placebo
10
0
0
Patients with Event (%)
0 6 12 18 24 30 36 42 48
Months since Randomization
0 6 12 18 24 30 36 42 48
Months since Randomization
136
115
319
307
555
519
819
835
1191
1187
1525
1495
1742
1702
1890
1859
2012
2026
No. at Risk
Darbepoetin alfa
Placebo
137
123
325
324
577
539
851
863
1232
1235
1566
1550
146
132
347
338
597
556
869
880
1255
1251
1583
1561
No. at Risk
Fatal or Nonfatal Myocardial Infarction
Darbepoetin alfa 2012 1920 1785
Placebo 2026 1907 1765
Myocardial Ischemia
Darbepoetin alfa 2012 1924 1794
Placebo 2026 1906 1767
E Fatal or Nonfatal Stroke
50
Hazard ratio, 1.92 (95% CI, 1.38–2.68)
40 P

Darbepoetin Alfa Therapy for reduction of Cardiovascular Events
Figure 2 (facing page). Kaplan–Meier Estimates of the
Probability of the Primary and Secondary End Points.
Panel A shows the primary cardiovascular composite
end point. The secondary end points of deaths from
any cause (Panel B), fatal or nonfatal congestive heart
failure (Panel C), fatal or nonfatal myocardial infarction
and myocardial ischemia (Panel D), and fatal or nonfatal
stroke (Panel E) are also shown. P values are not
adjusted for multiple comparison.
Venous thromboembolic events were reported
in 41 patients in the darbepoetin alfa group
(2.0%), as compared with 23 patients in the placebo
group (1.1%) (P = 0.02). Arterial thromboembolic
events (some of which were adjudicated
as cardiovascular events) were also reported more
frequently in the darbepoetin alfa group (in 178
patients [8.9%] vs. 144 patients [7.1%], P = 0.04).
Cancer
There was no significant between-group difference
in the number of patients reporting a cancerrelated
adverse event: 139 in the darbepoetin alfa
group (6.9%) and 130 in the placebo group (6.4%)
(P = 0.53). Overall, 39 deaths were attributed to
cancer in the 2012 patients in the darbepoetin
alfa group and 25 deaths were attributed to cancer
in the 2026 patients in the placebo group (P = 0.08
by the log-rank test). Among patients with a history
of a malignant condition at baseline, there
were 60 deaths from any cause in the 188 patients
assigned to darbepoetin alfa and 37 deaths in the
160 patients assigned to placebo (P = 0.13 by the
log-rank test). In this subgroup, 14 of the 188 patients
assigned to darbepoetin alfa died from cancer,
as compared with 1 of the 160 patients assigned
to placebo (P = 0.002 by the log-rank test).
Discussion
The main purpose of the TREAT was to determine
whether treatment of a low hemoglobin level with
darbepoetin alfa would reduce the risk of death
and major cardiovascular and renal events among
patients with type 2 diabetes mellitus, chronic
kidney disease, and anemia. There was no significant
difference in the overall rates of either the
primary cardiovascular composite end point or
the primary renal composite end point with the
use of the therapeutic strategy of increasing the
hemoglobin level with darbepoetin alfa (to attempt
to reach a target hemoglobin level of 13.0 g
per deciliter) versus placebo (with darbepoetin
alfa as rescue therapy if hemoglobin values decreased
below 9.0 g per deciliter). In this study
involving 4038 patients randomly assigned to a
study group, with 9941 patient-years of follow-up
and with 1234 patients who had a composite outcome
of a cardiovascular event and 1270 patients
who had a composite outcome of a renal event,
we found no overall significant differences in the
hazard rates for these clinically important outcomes
or in the rates of death, heart failure, myocardial
infarction, admission for myocardial ischemia,
or end-stage renal disease.
Although there was no significant increase in
the overall cardiovascular composite outcome in
the group assigned to darbepoetin alfa, there was
an increased incidence of stroke (annualized event
rate, 2.1%, vs. 1.1% in the placebo group). Data
from a study of patients undergoing dialysis suggested
this risk of stroke 19 ; however, this risk
was not identified in either the Correction of
Hemoglobin and Outcomes in Renal Insufficiency
(CHOIR) study (ClinicalTrials.gov number,
NCT00211120) 17 or a meta-analysis. 24 This previously
undocumented and clinically important
finding of a heightened risk of stroke is not explained
by an increase in systolic blood pressure.
In a recent report of a randomized, placebo-controlled
trial of intravenous erythropoietin in the
early phase of an acute ischemic stroke, 25 there
were more deaths in the group assigned to the
ESA (42 of 256 patients [16.4%]) as compared
with the group assigned to placebo (24 of 266
patients [9.0%]). This study also provides support
for an adverse relationship between ESAs and
stroke. We also confirmed previous findings of
increased rates of venous thromboembolic events
among patients treated with ESAs.
Our findings appear to differ from those of
the CHOIR trial, the next largest trial involving
patients with chronic kidney disease who were
not undergoing dialysis, which used epoetin alfa
and was discontinued prematurely after a median
follow-up of 16 months and after 222 patients
had a primary event. 17 Both groups in the CHOIR
trial were randomly assigned to receive epoetin
alfa, and in that study, there was a higher risk of
cardiovascular events in the group assigned to a
target hemoglobin level of 13.5 g per deciliter
than in the group assigned to a target level of
11.3 g per deciliter (hazard ratio, 1.34; 95% CI,
1.03 to 1.74; P = 0.03). This excess of cardiovas-
n engl j med 361;21 nejm.org november 19, 2009 2029
Downloaded from www.nejm.org at Amgen, Inc. on November 20, 2009 .
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
Page 134

2030
A Renal Composite (ESRD or Death)
Patients with Event (%)
No. at Risk
Darbepoetin alfa
Placebo
B ESRD
Patients with Event (%)
No. at Risk
Darbepoetin alfa
Placebo
50
40
30
20
10
cular events was largely accounted for by more
deaths and heart-failure events; only 12 patients
in each group had a stroke. These findings led
to the suggestion that our trial should be discon-
tinued. 26 However, by that time, there were more
accumulated cardiovascular composite events in
our trial than in the CHOIR trial, and the independent
data and safety monitoring committee
recommended continuation. 27 The TREAT executive
committee updated the consent form to reflect
the results from the CHOIR trial, and our
patients again were asked to provide written informed
consent. The final results of our study
demonstrate the importance of completing the
The new england journal of medicine
Hazard ratio, 1.06 (95% CI, 0.95–1.19)
P=0.29
0
0 6 12 18 24 30 36 42 48
2012
2026
50
40
30
20
10
1910
1915
1762
1748
Months since Randomization
1544
1519
Hazard ratio, 1.02 (95% CI, 0.87–1.18)
P=0.83
1207
1193
820
842
n engl j med 361;21 nejm.org november 19, 2009
552
540
Darbepoetin alfa
309
312
Placebo
0
0 6 12 18 24 30 36 42 48
2012
2026
1908
1907
1755
1729
Months since Randomization
1527
1491
1187
1162
802
811
535
513
Darbepoetin alfa
Placebo
Figure 3. Kaplan–Meier Estimates of the Probability of Renal Outcomes.
AUTHOR: Pfeffer
RETAKE: 1st
Panel A shows the primary renal composite end point (end-stage renal disease [ESRD] 2nd or death), and Panel B shows
the end-point component of ESRD. FIGURE: P values 3 of 3have
not been adjusted for multiple comparisons.
3rd
Revised
ARTIST: MRL
SIZE
6 col
TYPE: Line Combo 4-C H/T 33p9
AUTHOR, PLEASE NOTE:
Figure has been redrawn and type has been reset.
Please check carefully.
JOB: 361xx ISSUE:
296
292
134
123
129
115
planned follow-up of trials and the potential to
draw misleading conclusions when premature discontinuation
results in an insufficient number
of events to allow for a reliable estimation of the
effect of treatment. 28,29
Before we conducted our study, there was a
suggestion that ESAs might increase mortality
among patients with cancer, so patients with an
active malignant condition were excluded from
our trial. 30 During the trial, mounting concern led
to new regulatory warnings about the use of ESAs
in patients with “anemia of cancer,” 31 xx-xx-09
prompting
a protocol amendment to discontinue the study
drug in patients in whom cancer developed. Al-
Downloaded from www.nejm.org at Amgen, Inc. on November 20, 2009 .
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
Page 135

Darbepoetin Alfa Therapy for reduction of Cardiovascular Events
though the number of these patients did not differ
significantly between the study groups, among
the 348 patients with a history of cancer at baseline,
mortality was higher among those assigned
to darbepoetin alfa than among those assigned
to placebo; these findings are consistent with
those in a recent meta-analysis. 32
Our trial provides long-term, placebo-controlled
data on the use of ESAs in patients with diabetes,
chronic kidney disease, and moderate anemia
who are not undergoing dialysis. No single trial
can address the multitude of pertinent issues
without limitations, and our data may not be
fully applicable to other patient populations. It is
possible that other dosing strategies could be
developed to mitigate the risk of stroke while
conserving the modest benefits of treatment. In
light of the baseline imbalances between the
study groups, additional sensitivity analyses were
conducted, and they did not alter our results.
More extensive analyses of the patient-reported
outcome measures are needed to assess the durability
of the rather modest improvement we observed
only in the FACT-Fatigue score.
The long-standing presumption of a benefit of
ESAs in this patient population led to non–placebo-controlled
trials, which, by design, cannot
provide a reliable assessment of risk and benefit.
21 The present trial supplies these missing data,
which may assist physicians and patients in making
more informed decisions about individual
care. It is our view that, in many patients with
diabetes, chronic kidney disease, and moderate
anemia who are not undergoing dialysis, the increased
risk of stroke and possibly death among
patients with a history of a malignant condition
will outweigh any potential benefit of an ESA.
Supported by Amgen, which provided independent statistical
support through a contract with the board of regents for the
University of Wisconsin System for data analysis for the TREAT
Data and Safety Monitoring Committee as well as for the study.
Dr. Pfeffer reports receiving consulting fees from Abbott,
Amgen, AstraZeneca, Biogen, Boehringer Ingelheim, Boston
Scientific, Bristol-Myers Squibb, Centocor, CVRx, Genentech,
Cytokinetics, Daiichi Sankyo, Genzyme, Medtronic, Novartis,
Roche, Sanofi-Aventis, Servier, and VIA Pharmaceutics and grant
support from Amgen, Baxter, Celladon, Novartis, and Sanofi-
Aventis, and being named coinventor on a patent for the use of
inhibitors of the renin–angiotensin system in selected survivors
of myocardial infarction; Dr. Burdmann, receiving consulting
fees from Amgen and Sigma Pharma and grant support from
Amgen and Roche; Drs. Chen and Kewalramani, being employees
of and owning stock in Amgen; Dr. Cooper, receiving consulting
fees from Amgen; Dr. de Zeeuw, receiving consulting
fees from Amgen and Novartis and lecture fees from Amgen; Dr.
Eckardt, receiving consulting fees from Amgen, Ortho Biotech,
Roche, Affymax, Stada, and Sandoz–Hexal and lecture fees from
Amgen, Ortho Biotech, and Roche; Dr. Ivanovich, receiving consulting
fees from Amgen, Baxter, Biogen, and Reata; Dr. Levey,
receiving grant support from Amgen; Dr. Lewis, receiving consulting
fees from Amgen and grant support from Amgen and
the Robert Wood Johnson Foundation; Dr. McGill, receiving consulting
fees from Amgen and Boehringer Ingelheim, lecture fees
from Novartis, and grant support from Novartis and Boehringer
Ingel heim; Dr. McMurray, receiving consulting fees from Menarini,
Bristol-Myers Squibb, Roche, Novocardia, Boehringer
Ingelheim, Novartis, BioMérieux, and Boston Scientific, lecture
fees from AstraZeneca, Solvay, Takeda, Novartis, BMS Sanofi,
and Vox Media, and grant support from BMS, Novartis, Amgen,
AstraZeneca, Cytokinetics, Hoffmann–La Roche, Pfizer, Scios,
and GlaxoSmithKline; Dr. Parfrey, receiving consulting and
lecture fees from Amgen and lecture fees from Ortho Biotech;
Dr. Parving, receiving consulting fees from Amgen and Novartis
and lecture fees from Novartis; Dr. Singh, receiving consulting
fees from Johnson & Johnson and Watson, lecture fees from Johnson
& Johnson, Amgen, and Watson, and grant support from
Johnson & Johnson, Amgen, Roche, AMAG Pharmaceuticals, and
Watson; Dr. Solomon, receiving grant support from Amgen;
and Dr. Toto, receiving consulting and lecture fees from Amgen
and grant support from Novartis, Reata, and Abbott. No other
potential conflict of interest relevant to this article was reported.
Appendix
From the Department of Medicine, Brig ham and Women’s Hospital, Harvard Medical School (M.A.P., E.F.L., A.K.S., S.D.S.); and the
Division of Nephrology, Tufts Medical Center (A.S.L.) — both in Boston; Hospital de Base, Faculdade de Medicina de São José do Rio
Preto, São José do Rio Preto, Brazil (E.A.B.); Global Biostatistics and Epidemiology and Global Clinical Development, Amgen, Thousand
Oaks, CA (C.-Y.C., R.K.); Baker Heart Research Institute, Melbourne, VIC, Australia (M.E.C.); the Division of Clinical Pharmacology,
University Medical Center, Groningen, the Netherlands (D.Z.); the Department of Nephrology and Hypertension, University of Erlangen-Nuremberg,
Erlangen, Germany (K.-U.E.); the Department of Biostatistics and Medical Informatics, University of Wisconsin,
Madison (J.M.F.); the Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago (P.I.); the Department
of Medicine, Washington University School of Medicine, St. Louis (J.B.M.); the Department of Cardiology, British Heart Foundation
Glasgow Cardiovascular Research Centre, Glasgow, United Kingdom (J.J.V.M.); the Division of Nephrology, Health Sciences Centre, St.
John’s, NF, Canada (P.P.); the Department of Medical Endocrinology, Rigshospitalet, University of Copenhagen, Copenhagen, and Faculty
of Health Science, Aarhus University, Aarhus (H.-H.P.) — both in Denmark; Mario Negri Institute for Pharmacological Research,
Bergamo, Italy (G.R.); and the Department of Medicine, University of Texas Southwestern Medical Center, Dallas (R.T.).
The TREAT committees and teams are as follows: protocol development committee — R. Brenner and M. Pfeffer (cochairs), N.
Abrouk, W. Carroll, D. Green, P. Klassen, J. Lubina, J. McMurray, L. Messer, R. Newmark, P. Parfrey, S. Solomon; executive committee
— M. Pfeffer (chair), E. Burdmann, M. Cooper, D. de Zeeuw, K.-U. Eckardt, P. Ivanovich, A. Levey, J. McGill, J. McMurray, P. Parfrey,
H.-H. Parving, B. Pereira, G. Remuzzi, A. Singh, S. Solomon, R. Toto; hemoglobin monitoring committee — P. Ivanovich (chair), S.
Fishbane, A. Nissenson; statistical data analysis center — R. Bechhofer, D. DeMets, J. Feyzi; data and safety monitoring committee — C.
Hennekens (chair), G. Chertow, E. Frohlich, P. O’Brien, J. Rouleau; clinical end-point center — S. Solomon (chair), E. Lewis (cochair),
A. Desai, C. Duong, P. Finn, L.H. Hartley, S. Kesari, J. Lin, J. Meadows, L. Mielniczuk, F. Shamshad, A. Singh, O. Vasylyeva, L. Weinrauch;
study team — A. Adee, A. Ahmed, P. Blaisdell, P. Brady, L. Bucker, J. Cap, W. Carroll, D. Chaparro, C.-Y. Chen, B. Condon, R.
n engl j med 361;21 nejm.org november 19, 2009 2031
Downloaded from www.nejm.org at Amgen, Inc. on November 20, 2009 .
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
Page 136

2032
Darbepoetin Alfa Therapy for reduction of Cardiovascular Events
Cravario, R. Damato, K. Dellacroce, K. DiRocco, D. Donovan, B. Doria, M. Dougherty, J. Droge, A. Grewal, L. Guimaraes, P. Haynes,
J. Hevy, T. Jambusaria, N. James, A. Kacprzak, M. Kasenda, S.R.K. Reddy, R. Kewalramani, H. Kirby, C. Klacker, T. Kocsis, A. Kondo,
M. Lauw, M. Liebelt, P. McElligott, B. Mesquita, C. Mix, A. Mueller, S. O’Callaghan, H. Ocampo, K. Olson, S. Perez, J. Pillai, A. Pollock,
F. Poloni, K. Polu, R. Poston, G. Ramirez, A. Rasmussen, L. Reeves, M. Rocha, A.P. Ross, J. Rossert, B. Sahl, T. See, S. Shahinfar, K.
Shart, C. Stehman-Breen, A. Teh, L. Tierra, M. Vaghela, C. White. Enrollment by country was as follows (numbers of patients are in
parentheses): Argentina (84), Australia (58), Austria (3), Brazil (190), Bulgaria (28), Canada (176), Chile (14), Czech Republic (96),
Denmark (10), Estonia (16), France (11), Germany (135), Hungary (70), Italy (87), Latvia (51), Mexico (150), Poland (213), Portugal
(17), Romania (47), Russia (117), Slovakia (65), Slovenia (12), United Kingdom (49), United States (2339).
References
1. Stamler J, Vaccaro O, Neaton JD, mortality in end-stage renal disease. Am J 21. Pfeffer MA. Critical missing data in
Wentworth D. Diabetes, other risk fac- Kidney Dis 1996;28:53-61.
erythropoiesis-stimulating agents in CKD:
tors, and 12-yr cardiovascular mortality 11. Keane WF, Brenner BM, de Zeeuw D, first beat placebo. Am J Kidney Dis 2008;
for men screened in the Multiple Risk Fac- et al. The risk of developing end-stage re- 51:366-9.
tor Intervention Trial. Diabetes Care 1993; nal disease in patients with type 2 diabe- 22. Pfeffer MA, Burdmann EA, Chen CY,
16:434-44.
tes and nephropathy: the RENAAL study. et al. Baseline characteristics in the Trial
2. Booth GL, Kapral MK, Fung K, Tu JV. Kidney Int 2003;63:1499-507.
to Reduce Cardiovascular Events With
Relation between age and cardiovascular 12. Vlagopoulos PT, Tighiouart H, Weiner Aranesp Therapy (TREAT). Am J Kidney
disease in men and women with diabetes DE, et al. Anemia as a risk factor for car- Dis 2009;54:59-69.
compared with non-diabetic people: a popdiovascular disease and all-cause mortal- 23. Mozzicato P. Standardised MedDRA
ulation-based retrospective cohort study. ity in diabetes: the impact of chronic kid- queries: their role in signal detection.
Lancet 2006;368:29-36.
ney disease. J Am Soc Nephrol 2005;16: Drug Saf 2007;30:617-9.
3. Turnbull F, Neal B, Algert C, et al. Ef- 3403-10.
24. Strippoli GF, Navaneethan SD, Craig
fects of different blood pressure-lowering 13. Tong PCY, Kong APS, So W-Y, et al. JC. Haemoglobin and haematocrit targets
regimens on major cardiovascular events Hematocrit, independent of chronic kid- for the anaemia of chronic kidney dis-
in individuals with and without diabetes ney disease, predicts adverse cardiovascuease. Cochrane Database Syst Rev 2006;4:
mellitus: results of prospectively designed lar outcomes in Chinese patients with CD003967.
overviews of randomized trials. Arch In- type 2 diabetes. Diabetes Care 2006;29: 25. Ehrenreich H, Weissenborn K, Prange
tern Med 2005;165:1410-9.
2439-44.
H, et al. Recombinant human erythro-
4. Baigent C, Keech A, Kearney PM, et al. 14. Pfeffer MA, Solomon SD, Singh AK, poietin in the treatment of acute ischemic
Efficacy and safety of cholesterol-lower- Ivanovich P, McMurray JJ. Uncertainty in stroke. Stroke 2009 October 15 (Epub
ing treatment: prospective meta-analysis the treatment of anemia in chronic kidney ahead of print).
of data from 90,056 participants in 14 ran- disease. Rev Cardiovasc Med 2005;Suppl 3: 26. Strippoli GFM, Tognoni G, Navaneedomised
trials of statins. Lancet 2005;366: S35-S41.
than SD, Nicolucci A, Craig JC. Hemoglo-
1267-78. [Errata, Lancet 2005;366:1358, 15. Eschbach JW, Abdulhadi MH, Browne bin targets: we were wrong, time to move
2008;371:2084.]
JK, et al. Recombinant human erythro- on. Lancet 2007;369:349-50.
5. Gaede P, Lund-Andersen H, Parving poietin in anemic patients with end-stage 27. Pfeffer MA. An ongoing study of ane-
H-H, Pedersen O. Effect of a multifacto- renal disease: results of a phase III multi- mia correction in chronic kidney disease.
rial intervention on mortality in type 2 center clinical trial. Ann Intern Med N Engl J Med 2007;356:959-61.
diabetes. N Engl J Med 2008;358:580-91. 1989;111:992-1000.
28. DeMets DL, Pocock SJ, Julian DG. The
6. Brenner BM, Cooper ME, de Zeeuw D, 16. Canadian Erythropoietin Study Group. agonising negative trend in monitoring of
et al. Effects of losartan on renal and car- Association between recombinant human clinical trials. Lancet 1999;354:1983-8.
diovascular outcomes in patients with erythropoietin and quality of life and exer- 29. Califf RM, DeMets DL. Principles from
type 2 diabetes and nephropathy. N Engl J cise capacity of patients receiving haemo- clinical trials relevant to clinical practice:
Med 2001;345:861-9.
dialysis. BMJ 1990;300:573-8.
Part I. Circulation 2002;106:1015-21.
7. Lewis EJ, Hunsicker LG, Clarke WR, 17. Singh AK, Szczech L, Tang KL, et al. 30. Leyland-Jones B, Semiglazov V, Pawlet
al. Renoprotective effect of the angio- Correction of anemia with epoetin alfa in icki M, et al. Maintaining normal hemotensin-receptor
antagonist irbesartan in chronic kidney disease. N Engl J Med globin levels with epoetin alfa in mainly
patients with nephropathy due to type 2 2006;355:2085-98.
nonanemic patients with metastatic breast
diabetes. N Engl J Med 2001;345:851-60. 18. Drueke TB, Locatelli F, Clyne N, et al. cancer receiving first-line chemotherapy:
8. Tonelli M, Isles C, Craven T, et al. Ef- Normalization of hemoglobin level in pa- a sur vival study. J Clin Oncol 2005;23:5960fect
of pravastatin on rate of kidney functients with chronic kidney disease and 72.
tion loss in people with or at risk for coro- anemia. N Engl J Med 2006;355:2071-84. 31. FDA black box warning for Aranesp.
nary disease. Circulation 2005;112:171-8. 19. Parfrey PS, Foley RN, Wittreich BH, Silver Spring, MD: Food and Drug Admin-
9. Locatelli F, Pisoni RL, Combe C, et al. Sullivan DJ, Zagari MJ, Frei D. Doubleistration. (Accessed October 23, 2009, at
Anaemia in haemodialysis patients of five blind comparison of full and partial ane- http://www.accessdata.fda.gov/drugsatfda
European countries: association with mormia correction in incident hemodialysis _docs/label/2008/103951s5195PI.pdf.)
bidity and mortality in the Dialysis Out- patients without symptomatic heart dis- 32. Bohlius J, Schmidlin K, Brillant C, et
comes and Practice Patterns Study (DOPPS). ease. J Am Soc Nephrol 2005;16:2180-9. al. Recombinant human erythropoiesis-
Nephrol Dial Transplant 2004;19:121-32. 20. Phrommintikul A, Haas SJ, Elsik M, stimulating agents and mortality in pa-
[Erratum, Nephrol Dial Transplant 2004; Krum H. Mortality and target haemoglotients with cancer: a meta-analysis of ran-
19:1666.]
bin concentrations in anemic patients with domised trials. Lancet 2009;373:1532-42.
10. Foley RN, Parfrey PS, Harnett JD, Kent chronic kidney disease treated with eryth- [Erratum, Lancet 2009;374:28.]
GM, Murray DC, Barre PE. The impact of ropoietin: a meta-analysis. Lancet 2007; Copyright © 2009 Massachusetts Medical Society.
anemia on cardiomyopathy, morbidity, and 369:381-8.
n engl j med 361;21 nejm.org november 19, 2009
Downloaded from www.nejm.org at Amgen, Inc. on November 20, 2009 .
Copyright © 2009 Massachusetts Medical Society. All rights reserved.
Page 137

The new england journal of medicine
original article
Correction of Anemia with Epoetin Alfa
in Chronic Kidney Disease
Ajay K. Singh, M.B., B.S., Lynda Szczech, M.D., Kezhen L. Tang, Ph.D.,
Huiman Barnhart, Ph.D., Shelly Sapp, M.S., Marsha Wolfson, M.D.,
and Donal Reddan, M.B., B.S., for the CHOIR Investigators*
ABSTRACT
Background
Anemia, a common complication of chronic kidney disease, usually develops as a consequence
of erythropoietin deficiency. Recombinant human erythropoietin (epoetin
alfa) is indicated for the correction of anemia associated with this condition. However,
the optimal level of hemoglobin correction is not defined.
Methods
In this open-label trial, we studied 1432 patients with chronic kidney disease, 715 of
whom were randomly assigned to receive a dose of epoetin alfa targeted to achieve
a hemoglobin level of 13.5 g per deciliter and 717 of whom were assigned to receive
a dose targeted to achieve a level of 11.3 g per deciliter. The median study duration
was 16 months. The primary end point was a composite of death, myocardial infarction,
hospitalization for congestive heart failure (without renal replacement therapy),
and stroke.
Results
A total of 222 composite events occurred: 125 events in the high-hemoglobin group,
as compared with 97 events in the low-hemoglobin group (hazard ratio, 1.34; 95%
confidence interval, 1.03 to 1.74; P = 0.03). There were 65 deaths (29.3%), 101 hospitalizations
for congestive heart failure (45.5%), 25 myocardial infarctions (11.3%),
and 23 strokes (10.4%). Seven patients (3.2%) were hospitalized for congestive heart
failure and myocardial infarction combined, and one patient (0.5%) died after having a
stroke. Improvements in the quality of life were similar in the two groups. More patients
in the high-hemoglobin group had at least one serious adverse event.
Conclusions
The use of a target hemoglobin level of 13.5 g per deciliter (as compared with 11.3 g
per deciliter) was associated with increased risk and no incremental improvement in
the quality of life. (ClinicalTrials.gov number, NCT00211120.)
Page 138
From the Renal Division, Brigham and
Women’s Hospital and Harvard Medical
School, Boston (A.K.S.); the Renal Division,
Duke University Medical Center (L.S., D.R.),
Duke Clinical Research Institute (L.S., S.S.),
and the Department of Biostatistics and
Bioinformatics, Duke University (H.B.) —
all in Durham, NC; Ortho Biotech Clinical
Affairs, Bridgewater, NJ (K.L.T., M.W.);
and the Department of Medicine, University
College Galway, Galway, Ireland (D.R.).
Address reprint requests to Dr. Singh at
the Renal Division, Brigham and Women’s
Hospital, 75 Francis St., Boston, MA 02115,
or at asingh@partners.org.
*Investigators in the Correction of Hemogloblin
and Outcomes in Renal Insufficiency
(CHOIR) trial are listed in the Appendix.
N Engl J Med 2006;355:2085-98.
Copyright © 2006 Massachusetts Medical Society.
n engl j med 355;20 www.nejm.org november 16, 2006 2085

2086
A
nemia is common among patients
with chronic kidney disease. 1 In such patients,
treatment with erythropoietin has
been shown to enhance the quality of life. 2-5 However,
evidence suggesting that the correction of
anemia improves cardiovascular outcomes has
largely been derived from observational studies and
small interventional trials associating a high level
of hemoglobin (>12.0 g per deciliter) with a lower
rate of complications and death from cardiovascular
causes. 6-8 Other evidence has also indicated that
cardiovascular complications, such as left ventricular
hypertrophy, might be improved through the
use of a high hemoglobin level as a target. 4 However,
in a randomized, controlled study comparing
a hematocrit target of 42% with that of 30%
among patients with heart disease who were undergoing
hemodialysis, the former group had
higher rates of nonfatal myocardial infarction and
death, but not significantly so. 5
In 2000, a panel of the Kidney Disease Outcomes
Quality Initiative of the National Kidney
Foundation recommended that the target level of
hemoglobin should be 11.0 to 12.0 g per deciliter
in patients with chronic kidney disease, whether
or not they were receiving dialysis. 9 A recent update
of guidelines regarding anemia in such patients
expanded the target range to 11.0 to 13.0 g
per deciliter, 10 with the increase in the upper limit
of the target range justified on the basis of a potential
improvement in the patients’ quality of life.
In the Correction of Hemogloblin and Outcomes
in Renal Insufficiency (CHOIR) trial, we hypothesized
that in patients with chronic kidney disease,
the use of recombinant human erythropoietin
(epoetin alfa) to achieve a high hemoglobin
level (13.5 g per deciliter) would decrease the
risk of complications from cardiovascular causes
and death, as compared with a lower hemoglobin
level (11.3 g per deciliter).
Methods
Study Subjects
We conducted an open-label, randomized trial to
study the risks and benefits of the correction of
anemia in patients with chronic kidney disease
who were not receiving dialysis. We enrolled 1432
patients at 130 sites in the United States. At enrollment,
patients had to be at least 18 years of age,
have a hemoglobin level of less than 11.0 g per
deciliter, and have chronic kidney disease, defined
The new england journal of medicine
n engl j med 355;20 www.nejm.org november 16, 2006
Page 139
by an estimated glomerular filtration rate (GFR)
of 15 to 50 ml per minute per 1.73 m 2 of body-surface
area, with the use of the Modification of Diet
in Renal Disease (MDRD) formula. 11 Key exclusion
criteria included the presence of uncontrolled
hypertension, active gastrointestinal bleeding, an
iron-overload state, a history of frequent transfusions
in the previous 6 months, refractory irondeficiency
anemia, active cancer, previous therapy
with epoetin alfa, or angina pectoris that was
unstable or present at rest.
Intervention
Patients were assigned by computer-generated permuted-block
randomization to one of two groups:
a high-hemoglobin group (with an initial hemoglobin
target of 13.0 to 13.5 g per deciliter) or a
low-hemoglobin group (with an initial target of
10.5 to 11.0 g per deciliter). A protocol amendment
on February 25, 2003, changed the original hemoglobin
targets to 13.5 g per deciliter and 11.3 g per
deciliter, respectively. At the time of the protocol
amendment, 347 of the 1432 patients (24.2%) had
been enrolled, and only 132 of the total of 1939
patient-years had accrued. Both groups of patients
initially received epoetin alfa subcutaneously weekly;
administration was subsequently permitted every
other week if the hemoglobin level was stable.
(For details about the epoetin alfa regimen, see the
Supplementary Appendix, available with the full
text of this article at www.nejm.org.) The institutional
review board at each center approved the
protocol, and all the patients gave written informed
consent.
Laboratory Tests and Clinical Outcomes
A central laboratory (Covance) performed all biochemical
and hematologic analyses. We assessed
the patients’ quality of life using the Linear Analogue
Self-Assessment (LASA) (scores range from
0 to 100, with higher scores indicating better function),
12 the Kidney Disease Questionnaire (KDQ)
(total scores range from 4 to 35, with higher scores
indicating better health), 13 and the Medical Outcomes
Study 36-item Short-Form Health Survey
(SF-36) (scores for each subscale range from 0 to
100, with higher scores indicating better health). 14
Investigator-reported events were independently
adjudicated by the clinical committee reviewing
end points at the Duke Clinical Research Institute
(DCRI), whose members were unaware of patients’
study-group assignments. The primary end point

Correction of Anemia and Chronic Kidney Disease
was the time to the composite of death, myocardial
infarction, hospitalization for congestive heart
failure (excluding renal replacement therapy), or
stroke. Myocardial infarction was defined on the
basis of any two of the following: chest pain that
lasted for 15 minutes, abnormal cardiac enzyme
levels, or new findings on electrocardiography suggestive
of myocardial infarction. Hospitalization
for congestive heart failure was defined as an unplanned
presentation requiring admission, during
which the patient received intravenous therapy with
inotropes, diuretics, or vasodilators. If hospitalization
involved renal replacement therapy, the event
was not included in the primary composite end
point. Stroke was defined as a new neurologic deficit
of sudden onset that was not reversible within
24 hours and that was not due to a readily identifiable
nonvascular cause (e.g., a brain tumor or
trauma). Other secondary outcomes included the
time to renal replacement therapy, hospitalization
for either a cardiovascular cause or any cause, and
quality of life.
Statistical Analysis
We calculated that 1352 patients would need to
be enrolled for the study to have a statistical power
of 80% to detect a 25% reduction in the composite
event rate in the high-hemoglobin group over
a period of 3 years, assuming a 30% event rate in
the low-hemoglobin group, the occurrence of at
least 295 composite events overall during the 3-year
period, a 30% rate of early withdrawal for reasons
other than the occurrence of the primary end point,
and a type I error of 0.05. Four interim analyses
were planned in which efficacy guidelines used the
O’Brien–Fleming alpha-spending boundary, guidelines
for futility in which the likelihood that the
study would be mistakenly stopped because of a
nonsignificant difference between groups was 2%
or less, and a conditional power calculation. 15-17
An independent data and safety monitoring board
reviewed the study.
We used the Kaplan–Meier method to analyze
the time to the first event for events that occurred
during the study period. We used the log-rank test
to compare the times to the first event between
the two groups. Data on patients who did not have
an event were censored at the time of study termination
(either completion or early withdrawal).
Repeated-measures analysis of variance was used
to evaluate hemoglobin levels over time. Hemoglobin
levels obtained within 28 days after a transfu-
sion were excluded. All patients who received at
least one dose of study medication were included
in the safety analysis. Serious adverse events were
defined as life-threatening, resulting in death,
hospitalization, or substantial disability, or leading
to a congenital anomaly or birth defect. The principal
investigators (Drs. Singh and Reddan) developed
the protocol and all amendments in collaboration
with the DCRI and the industry sponsor.
The DCRI acquired and queried all data. The
database was developed and locked at DCRI, and
a copy was provided to the sponsor. The investigators
had full access to the data. DCRI performed
all the primary analyses, and the sponsor performed
all secondary analyses, the results of
which were verified by the DCRI. All analyses
were performed with the use of SAS software,
version 8.2 or higher.
Results
early termination of the study
The data and safety monitoring board recommended
that the study be terminated in May 2005 at the
time of the second interim analysis, even though
715 Assigned to high-hemoglobin
group (target level, 13.5 g/dl)
312 Completed 36 mo or withdrew
at study termination
without having primary event
125 Had a primary event
278 Withdrew before early
termination of study
131 Required RRT
147 Withdrew for other
reasons
1432 Patients enrolled
Page 140
717 Assigned to low-hemoglobin
group (target level, 11.3 g/dl)
349 Completed 36 mo or withdrew
at study termination
without having primary event
97 Had a primary event
271 Withdrew before early
termination of study
111 Required RRT
160 Withdrew for other
reasons
Figure 1. Enrollment and Outcomes.
A total of 1432 patients were enrolled; 715 were assigned to the high-hemoglobin
group (with a target level of 13.5 g per deciliter), and 717 were assigned
to the low-hemoglobin group (with a target level of 11.3 g per deciliter).
In addition to the stated reasons for withdrawal from the study, other
reasons included a request from a patient, an investigator, or the study
sponsor; pregnancy; an adverse event; a protocol violation; or a loss to follow-up.
RRT denotes renal replacement therapy.
n engl j med 355;20 www.nejm.org november 16, 2006 2087

2088
Table 1. Baseline Characteristics of the Patients.*
neither the efficacy nor the futility boundaries had
been crossed, because the conditional power for
demonstrating a benefit for the high-hemoglobin
group by the scheduled end of the study was less
than 5% for all plausible values of the true effect
for the remaining data. Other factors that the board
considered included an examination of differences
between the treatment groups in adverse events,
biochemical data, and quality-of-life data.
On the basis of the intention-to-treat principle,
data from all 1432 patients were included in the
final analysis (Fig. 1), and the nominal P value at
final analysis is reported. Both the mean and the
The new england journal of medicine
High-Hemoglobin Low-Hemoglobin
Characteristic
Group (N = 715) Group (N = 717)
Age (yr) 66.0±14.3 66.3±13.5
Female sex (%)
Race (%)
56.2 54.1
White 62.3 61.1
Black 28.6 29.3
American Indian or Alaskan Native 0.1 0.4
Asian or Pacific Islander 3.4 3.2
Other 5.6 6.0
Hispanic ethnic background (%) 12.5 13.5
History of smoking tobacco (%)
Cause of chronic kidney disease (%)
47.5 44.6
Diabetes 46.8 50.8
Hypertension 29.9 27.5
Other
Cardiovascular history (%)
23.3 21.6
Hypertension 95.8 93.2†
Myocardial infarction 16.4 15.0
CABG 17.4 13.5‡
PCI 10.9 11.9
Congestive heart failure 24.4 22.9
Atrial fibrillation 9.4 8.6
Stroke 9.8 10.0
Peripheral vascular disease 16.4 16.4
Myocardial infarction, stroke, CABG, PCI, or amputation of a lower limb 36.3 34.5
Body-mass index
Blood pressure (mm Hg)
30.4±7.7 30.4±7.5
Systolic 136.7±19.7 135.6±20.0
Diastolic 71.6±11.6 70.9±11.2
Mean arterial 93.3±12.1 92.5±12.0
n engl j med 355;20 www.nejm.org november 16, 2006
Page 141
median duration of follow-up of all patients were
16 months; 661 patients (46.2%) completed 36
months of study or withdrew at study termination
without having had a composite event. A total of
549 patients (38.3%) withdrew before termination
of the study without having had a composite event.
Among these patients, 242 (16.9%) withdrew because
they began renal replacement therapy, and
307 patients (21.4% [147 from the high-hemoglobin
group and 160 from the low-hemoglobin
group]) withdrew for other reasons. However, the
low-hemoglobin group had more patient-years of
follow-up (980, as compared with 959 in the high-

Table 1. (Continued.)
Correction of Anemia and Chronic Kidney Disease
hemoglobin group). The reasons for early withdrawal
were similar in the two groups (data not
shown).
characteristics of the patients
The demographic and baseline characteristics of
the two groups were similar (Table 1), except for
a higher rate of a self-reported history of hypertension
(P = 0.03) and coronary-artery bypass grafting
in the high-hemoglobin group (P = 0.05). During
the course of the study, the overall use of iron
in both the high-hemoglobin group and the lowhemoglobin
group was similar (52.0% and 48.3%,
respectively; P = 0.18). The mean (±SD) systolic
blood pressure decreased modestly from baseline
to the end of study, with a decrease of 2.3±22.8
mm Hg in the high-hemoglobin group and a de-
High-Hemoglobin
Group (N = 715)
Low-Hemoglobin
Group (N = 717)
Hemoglobin (g/dl) 10.1±0.9 10.1±0.9
Hematocrit (%) 31.4±2.9 31.4±2.9
Transferrin saturation (%) 25.2±11.8 24.6 ±10.1
Ferritin (ng/ml) 167.8±157.2 179.2±171.5
Creatinine clearance (ml/min/1.73 m2 )§ 36.7±17.0 37.1±17.9
GFR (ml/min)¶ 27.0±8.7 27.3±9.1
Albumin (g/dl) 3.7±0.5 3.8±0.5
Ratio of total protein to creatinine in urine
Medications (%)
1.6±2.3 1.5±2.3
ACE inhibitor only 35.7 37.8
ARB only 29.7 26.8
Combination of ACE inhibitor and ARB 8.3 9.6
Beta-blocker (including labetalol) 46.9 47.7
Platelet aggregation inhibitor (excluding heparin) 42.8 45.0
HMG CoA reductase inhibitor
Iron
52.8 52.3
Intravenous 2.6 1.6
Oral 26.5 26.7
Unknown route 3.1 1.6
* Plus–minus values are means ±SD. The body-mass index is the weight in kilograms divided by the square of the height
in meters. Race and ethnic group were assigned by the investigators. Cardiovascular history was reported either by the
patient or by chart review. CABG denotes coronary-artery bypass grafting, PCI percutaneous coronary intervention, ACE
angiotensin-converting enzyme, ARB angiotensin II–receptor blocker, and HMG CoA 3-hydroxy-3-methylglutaryl coenzyme
A.
† P = 0.03 for the comparison with the high-hemoglobin group.
‡ P = 0.05 for the comparison with the high-hemoglobin group.
§ The rate was calculated with the use of the Cockcroft–Gault formula.
¶ The GFR was calculated according to the MDRD formula.
crease of 2.6±21.9 mm Hg in the low-hemoglobin
group. The difference was not statistically significant
between the two groups (P = 0.27). The mean
diastolic blood pressure increased by 0.2±12.9
mm Hg in the high-hemoglobin group and decreased
by 0.7±12.4 mm Hg in the low-hemoglobin
group by the end of the study, as compared
with baseline (P = 0.02).
The hemoglobin levels over time are shown in
Figure 2A. The mean change in the hemoglobin
level from baseline to the final measurement was
2.5 g per deciliter for the high-hemoglobin group
and 1.2 g per deciliter for the low-hemoglobin
group, a mean difference of 1.3 g per deciliter
(P

2090
level in the high-hemoglobin group was nearly
twice that required in the low-hemoglobin group
(11,215 U and 6276 U per week, respectively).
Primary Outcomes
In the primary analysis of the composite events,
a patient was counted only once (e.g., if a myocardial
infarction occurred before a stroke, then only
the time from randomization to the myocardial infarction
was included in the composite event for
the patient). A total of 222 composite events
(death, myocardial infarction, hospitalization for
congestive heart failure without renal replacement
therapy, or stroke) occurred: 125 among the
715 patients in the high-hemoglobin group, as
compared with 97 among the 717 patients in the
low-hemoglobin group (17.5% vs. 13.5%; hazard
ratio, 1.34; 95% confidence interval [CI], 1.03 to
1.74; P = 0.03) (Fig. 3A). Of the 222 composite
events, there were 65 deaths (29.3%), 101 hospitalizations
for congestive heart failure without
renal replacement therapy (45.5%), 25 myocardial
infarctions (11.3%), and 23 strokes (10.4%). Notably,
seven patients (3.2%) were hospitalized for
congestive heart failure and had a myocardial infarction
on the same day, and one patient (0.5%)
died after having a stroke on the same day. Death
and hospitalization for congestive heart failure
accounted for 74.8% of the composite events. Sensitivity
analyses yielded similar results. A perprotocol
analysis of primary outcome data for
1395 patients showed a hazard ratio of 1.34
(95% CI, 1.03 to 1.75; P = 0.03). The per-protocol
population excluded 37 patients from the intention-to-treat
population who did not meet the criteria
required by the protocol and either underwent
incorrect randomization, did not receive any
dose of study medication, had no measurements
of hemoglobin obtained after randomization, or
did not meet all criteria for inclusion.
On the basis of data from the intention-totreat
population, but including all events from
randomization to study termination or 30 days
after the last administration of study medication,
the hazard ratio for the primary outcome in the
high-hemoglobin group, as compared with the
low-hemoglobin group, was 1.30 (95% CI, 1.01 to
1.68; P = 0.04). When the analysis included all
events from randomization to 90 days after study
termination, the hazard ratio was also 1.30 (95%
CI, 1.01 to 1.66; P = 0.04).
The new england journal of medicine
n engl j med 355;20 www.nejm.org november 16, 2006
Page 143
Figure 2 (facing page). Mean Monthly Hemoglobin
Levels (Panel A) and Mean Weekly Doses of Epoetin
Alfa (Panel B).
In Panel A, a separation in hemoglobin values between
the two groups was observed just before the third
month. Mean hemoglobin values were close to the target
of 11.3 g per deciliter in the low-hemoglobin group
but were consistently below the target of 13.5 g per deciliter
in the high-hemoglobin group. A total of 93.9% of
patients in the low-hemoglobin group and 75.9% of patients
in the high-hemoglobin group had at least one hemoglobin
value that reached the target. The median
time for patients in the high-hemoglobin group to reach
the target of 13.5 g per deciliter was 126 days (95% confidence
interval [CI], 113 to 139), whereas it took a median
of 36 days (95% CI, 29 to 43) for patients in the lowhemoglobin
group to reach the target level of 11.3 g per
deciliter (P

A
B
Mean Hemoglobin (g/dl)
No. of Patients
High-hemoglobin
Low-hemoglobin
Mean Dose of Epoetin Alfa (U)
No. of Patients
High-hemoglobin
Low-hemoglobin
15.5
15.0
14.5
14.0
13.5
13.0
12.5
12.0
11.5
11.0
10.5
10.0
9.5
20,000
18,000
16,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
Correction of Anemia and Chronic Kidney Disease
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
710
707
667
672
632
625
600
603
558
549
507
528
485
510
was examined; the difference between the two
groups persisted (hazard ratio, 1.28; 95% CI, 1.07
to 1.54; P = 0.007). Results similar to those in the
primary analysis were observed when hospitalization
for congestive heart failure with renal replacement
therapy was included in the composite end
point (hazard ratio for the high-hemoglobin group
High-hemoglobin group
433
471
Low-hemoglobin group
367
384
Month
306
334
252
250
194
182
High-hemoglobin group
Low-hemoglobin group
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
709
707
693
691
659
655
623
621
578
577
530
549
500
526
452
479
370
393
Month
310
333
258
262
189
189
139
141
132
141
vs. the low-hemoglobin group, 1.37; P = 0.02). In
addition, the results comparing all hospitalizations
for congestive heart failure (including those
involving renal replacement therapy) in the two
groups were similar to the results of hospitalization
for congestive heart failure in the primary end
point (hazard ratio, 1.44; P = 0.04).
n engl j med 355;20 www.nejm.org november 16, 2006 2091
95
101
97
95
81
75
79
73
67
60
65
54
49
45
52
43
31
30
27
27
13
13
11
12
Page 144

2092
The patients’ quality of life (as assessed by LASA,
KDQ, and SF-36 scores) showed similar levels of
improvement from baseline values in both groups,
except for the score for the emotional role subscale
of the SF-36, which was significantly higher
in the low-hemoglobin group (Table 2).
Of the patients who reported adverse events,
a total of 376 of 686 patients (54.8%) in the highhemoglobin
group and 334 of 688 patients (48.5%)
in the low-hemoglobin group had at least one serious
adverse event between the time of randomization
and the end of the study (P = 0.02) (Table 3).
The types of serious adverse events were similar
in the two groups, with the exception of congestive
heart failure, which occurred more frequently
in the high-hemoglobin group (11.2% vs. 7.4%,
P = 0.02). The types of serious adverse events that
occurred after the end of study were also similar
in the two groups (data not shown).
Discussion
We observed an increased risk of the primary composite
end point in the high-hemoglobin group, as
compared with the low-hemoglobin group. Death
and hospitalization for congestive heart failure
accounted for 74.8% of the composite events. On
the basis of findings of three validated instruments
(LASA, KDQ, and SF-36), the overall quality
of life improved when anemia was treated with
epoetin alfa, but aiming for a target value of 13.5 g
of hemoglobin per deciliter provided no additional
quality-of-life benefit. Since our study showed no
apparent additional benefit in quality of life, and
since the cost of epoetin alfa treatment increases
with higher doses, we believe that the use of a
high target hemoglobin level provides no cost
benefit for either patients or payers in this population,
even before considering risk.
Several studies have demonstrated that the
correction of anemia in patients with chronic
kidney disease improves the quality of life and
exercise tolerance while reducing the need for
transfusion. 5,18 However, as Strippoli et al. have
observed, 19 there remains much uncertainty about
the validity of various assessments of the quality
of life in published studies.
Data on the effects of the correction of anemia
on cardiovascular outcomes and survival have been
both discordant and controversial. Recent large,
controlled studies involving patients with pre–endstage
or end-stage renal disease have shown either
The new england journal of medicine
n engl j med 355;20 www.nejm.org november 16, 2006
Page 145
Figure 3 (facing page). Kaplan–Meier Estimates
of the Probability of the Primary Composite End Point
and Secondary End Points of Individual Components
— Hospitalization for Congestive Heart Failure (CHF)
without Renal Replacement Therapy (RRT), Myocardial
Infarction, Stroke, and Death.
Panel A shows that the largest separation between the
two groups in the primary composite end point occurred
at 15 months. At that time, the Kaplan–Meier
estimate of the difference in cumulative event rates between
the two groups reached 4.7 percentage points
(15.8% in the high-hemoglobin group vs. 11.1% in the
low-hemoglobin group). After 15 months, the difference
between the two groups remained constant, with
752 patients (52.5%) remaining in the study (355 in
the high-hemoglobin group and 397 in the low-hemoglobin
group). There were no significant differences
between the two groups in the four individual components
of the primary composite end point (Panels B,
C, D, and E). However, the hazard ratios for death and
hospitalization for CHF had strong trends toward a
higher risk in the high-hemoglobin group than in the
low-hemoglobin group.
an increase in adverse events or no benefit from
the normalization of hemoglobin levels. 20-26 Furthermore,
in several studies, complete correction
of anemia, as compared with partial correction,
did not improve left ventricular hypertrophy. 18,20-
23 Our results, coupled with the results of other
recent interventional trials involving patients with
chronic kidney disease, 27,28 reinforce the differences
between observational and clinical trial data,
which appear particularly notable in the setting
of anemia therapy. 29
The Anemia Guideline Committee of the Dialysis
Outcomes Quality Initiative has recently
updated its guidelines. 10 The lower limit of the
hemoglobin level was set at 11.0 g per deciliter
as an “evidence-based recommendation,” whereas
the upper limit was set at 13.0 g per deciliter as
a “clinical practice recommendation.” The committee
concluded that there was insufficient evidence
to recommend the routine maintenance of
a hemoglobin level of 13.0 g per deciliter or higher
in patients being treated with erythropoiesis-stimulating
agents. There was also concern that the
narrow range of 11 to 12 g per deciliter could not
be achieved because of hemoglobin cycling. The
panel emphasized that the use of a high target
hemoglobin level may be associated with an increased
risk. In the high-hemoglobin group in our
study, we used a level of 13.5 g per deciliter as a
target but achieved a mean level of just 12.6 g per

Primary Composite End Point
0.30
A
High-hemoglobin group
0.25
0.20
Low-hemoglobin group
0.15
0.10
Probability of
Composite Event
Correction of Anemia and Chronic Kidney Disease
0.05
0.00
0 3 6 9 12 15 18 21 24 27 30 33 36 39
Month
23
23
55
44
72
67
101
107
176
182
270
293
355
397
457
499
520
539
587
594
654
660
715
717
No. at Risk
High-hemoglobin
Low-hemoglobin
Hazard ratio: 0.91
P=0.78
Low-hemoglobin group
Myocardial Infarction
0.20
0.15
0.10
0.05
B Hospitalization for CHF (without RRT)
C
0.20
0.15 Hazard ratio: 1.41
High-hemoglobin group
P=0.07
0.10
0.05
Low-hemoglobin group
High-hemoglobin group
0.00
Probability of Event
0.00
Probability of Event
0 3 6 9 12 15 18 21 24 27 30 33 36 39
Month
0 3 6 9 12 15 18 21 24 27 30 33 36 39
Month
25
26
59
49
79
73
113
115
193
192
295
307
387
415
487
520
543
560
612
609
674
672
715
717
No. at Risk
High-hemoglobin
Low-hemoglobin
23
24
56
45
73
70
102
111
179
187
273
299
359
402
461
504
523
544
591
596
656
663
715
717
No. at Risk
High-hemoglobin
Low-hemoglobin
n engl j med 355;20 www.nejm.org november 16, 2006 2093
Death
0.20
D Stroke
E
0.20
High-hemoglobin group
Hazard ratio: 1.48
P=0.07
0.15
Probability of Event
Hazard ratio: 1.01
P=0.98
0.15
0.10
Low-hemoglobin group
0.10
Low-hemoglobin group
0.05
High-hemoglobin group
0.05
0.00
0.00
Probability of Event
0 3 6 9 12 15 18 21 24 27 30 33 36 39
Month
39
0 3 6 9 12 15 18 21 24 27 30 33 36
Month
25
26
60
49
80
74
114
117
196
195
297
310
389
418
490
523
545
564
614
610
675
676
715
717
No. at Risk
High-hemoglobin
Low-hemoglobin
25
25
59
48
79
72
113
115
195
193
295
306
386
414
487
518
543
559
611
608
672
675
715
717
No. at Risk
High-hemoglobin
Low-hemoglobin
Page 146

Table 2. Secondary End Points.*
End Point
2094
High-Hemoglobin
Group (N=715)
The new england journal of medicine
no. of patients (%)
Low-Hemoglobin
Group (N=717) Hazard Ratio (95% CI) P Value†
Clinical results
Components of the primary end point‡
Death 52 (7.3) 36 (5.0) 1.48 (0.97–2.27) 0.07
Hospitalization for congestive
heart failure (without renal
replacement therapy)
64 (9.0) 47 (6.6) 1.41 (0.97–2.05) 0.07
Myocardial infarction 18 (2.5) 20 (2.8) 0.91 (0.48–1.73) 0.78
Stroke
Renal replacement therapy
12 (1.7) 12 (1.7) 1.01 (0.45–2.25) 0.98
Any renal replacement therapy§ 155 (21.7) 134 (18.7) 1.19 (0.94–1.49) 0.15
Hospitalization for renal replacement
therapy
Hospitalization
99 (13.8) 81 (11.3) 1.25 (0.93–1.68) 0.13
Cardiovascular causes 233 (32.6) 197 (27.5) 1.23 (1.01–1.48) 0.03
Any cause 369 (51.6) 334 (46.6) 1.18 (1.02–1.37) 0.03
High-Hemoglobin Group Low-Hemoglobin Group P Value¶
Baseline Change from Baseline∥ Baseline Change from Baseline**
Quality of life††
LASA score
Energy 38.1±23.7 16.6±28.6 38.2±23.1 15.5±28.6 0.67
Activity 40.8±25.9 15.0±39.9 42.5±25.8 13.3±29.8 0.98
Overall quality of life 46.3±26.2 11.2±29.7 46.1±25.4 11.9±28.1 0.46
KDQ total score
SF-36 score
20.3±5.80 1.6±5.6 20.6±6.00 1.1±5.6 0.26
Physical function 41.9±28.2 3.2±24.0 42.4±27.3 2.1±23.3 0.49
Physical role 31.9±38.9 6.4±40.7 32.5±39.2 7.5±43.2 0.32
Pain 57.8±28.5 0.4±28.1 58.0±27.1 2.4±26.7 0.15
General health 41.3±20.1 3.0±19.2 42.6±20.1 1.8±17.8 0.87
Vitality 35.2±22.6 10.0±23.8 36.6±22.4 8.2±20.6 0.58
Social function 63.7±29.5 1.3±33.1 63.7±29.0 3.5±28.7 0.16
Emotional role 57.2±43.6 0.8±48.3 57.4±43.3 5.9±48.1 0.01
Mental health 69.6±19.5 1.7±18.5 70.2±20.1 2.4±18.2 0.31
* Plus–minus values are means ±SD. Hazard ratios are for the comparison of the high-hemoglobin group with the low-hemoglobin group.
† P values were calculated by the log-rank test.
‡ Components of the primary end point were analyzed separately. For example, if a patient had more than one type of event, each event was
counted the first time it occurred. Therefore, a patient could be included in more than one category of events. In the primary analysis of
the composite events, a patient was counted only once (e.g., if a myocardial infarction occurred before a stroke, then only the time from
randomization to the myocardial infarction was included in the composite event for the patient).
§ A total of 47 patients (24 in the high-hemoglobin group and 23 in the low-hemoglobin group) had a composite event.
¶ P values were calculated by analysis of covariance with the baseline score as a covariate.
∥ P values for comparisons of the change from baseline were between

Table 3. Adverse Events.*
Correction of Anemia and Chronic Kidney Disease
High-Hemoglobin Group
(N = 686)
deciliter, with an increase in risk with no qualityof-life
benefit. Furthermore, the number of patients
who had at least one serious adverse event
was higher in the high-hemoglobin group than
in the low-hemoglobin group. Thus, our study
does not provide support for the expanded target
range recently advocated by the National
Kidney Foundation.
Patients in the high-hemoglobin group had a
higher (but not significantly higher) rate of both
progression to renal replacement therapy and hospitalization
for renal replacement therapy. Smaller
interventional studies have suggested the contrary.
In a recent randomized, controlled study
involving 88 patients, Gouva et al. reported that
the early initiation of epoetin alfa treatment in
patients with chronic kidney disease in an effort
to achieve a hemoglobin level of 13.0 g per deciliter
reduced the rate of the composite end point
Low-Hemoglobin Group
(N = 688) P Value†
Adverse Event
no. of patients (%)
Any event
Thrombovascular event
607 (88.5) 589 (85.6) 0.11
Any event 126 (18.4) 120 (17.4) 0.65
Any clinically relevant event‡ 74 (10.8) 82 (11.9) 0.51
Any serious adverse event 376 (54.8) 334 (48.5) 0.02
Any serious adverse event associated with
epoetin alfa§
Serious adverse events**
10 (1.5)¶ 3 (0.4)∥ 0.05
Congestive heart failure 77 (11.2) 51 (7.4) 0.02
Myocardial infarction 10 (1.5) 19 (2.8) 0.09
Gastrointestinal hemorrhage 18 (2.6) 18 (2.6) 0.99
Chest pain 23 (3.4) 16 (2.3) 0.25
Cellulitis 16 (2.3) 11 (1.6) 0.33
Pneumonia 32 (4.7) 28 (4.1) 0.59
Renal failure 95 (13.8) 73 (10.6) 0.07
* The analysis includes 1374 patients in the two study groups who received at least one dose of epoetin alfa and for
whom data were collected regarding adverse events.
† P values were calculated by the chi-square test.
‡ The thrombovascular events that were considered to be clinically relevant included myocardial infarction, stroke, angina
pectoris, transient ischemic attack, deep-vein thrombosis, pulmonary embolism, and retinal-vein occlusion.
§ The event was deemed by the investigators to be related to drug administration.
¶ These events included two cases of deep-vein thrombosis and one case each of pulmonary embolism, retinal-vein occlusion,
transient ischemic attack, deep-vein thrombosis and pulmonary embolism, priapism, rash, allergic dermatitis,
and unstable angina.
∥ These events included one case each of hypertension, pulmonary embolism, and stroke.
** Serious adverse events that occurred in at least 2% of patients in either study group are listed.
of a doubling of creatinine levels, renal replacement,
or death, as compared with deferred initiation
of treatment (P = 0.008 by the log-rank
test). 30 However, elsewhere in this issue of the
Journal, the Cardiovascular Risk Reduction by Early
Anemia Treatment with Epoetin Beta (CREATE)
investigators report that more patients assigned
to complete correction of anemia than to partial
correction progressed to dialysis at the end of the
study (P = 0.03). 24 Thus, it appears that larger studies
either demonstrate no apparent benefit or
actually may show an increased risk of progression
to renal replacement therapy with the targeting
of a high hemoglobin value. Clearly, additional
studies will be required to address this issue.
Our study has several potential limitations.
Since we prespecified the censoring of data on
patients at the time of the initiation of renal replacement
therapy, no further data were collected;
Page 148
n engl j med 355;20 www.nejm.org november 16, 2006 2095

2096
there is a possibility of bias if the rates of renal
replacement therapy differed between the two
groups. Our analysis showed that there was no
significant difference between the two groups
with respect to the time to renal replacement
therapy (P = 0.15). Furthermore, when renal replacement
therapy was treated as an event and
added to the composite outcome, the difference
between the two groups was similar to that obtained
in the primary analysis. A potential limitation
of the reporting of the components of the
primary end point is the issue of death as a competing
risk. It could be argued that the results
with respect to myocardial infarction and stroke
should be interpreted with caution; however, our
primary results remain unchanged. We also censored
data from the time-to-event analysis for a
large number of patients because they required
renal replacement therapy (16.9%) or withdrew
from the study (21.4%) (Fig. 1). This factor would
be a limitation if the censoring that occurred was
not random in nature; however, such confounding
is unlikely because the numbers of patients
whose data were censored or who withdrew for
other reasons did not differ significantly between
the two groups. The differential withdrawal rate
could also have generated bias. The low-hemoglobin
group did have a higher number of early
withdrawals for other reasons than did the highhemoglobin
group. However, the low-hemoglobin
group had more patient-years of follow-up (980,
as compared with 959 in the high-hemoglobin
group). Furthermore, the two groups had similar
demographic characteristics at baseline.
Another potential limitation is the lack of a
double-blind design; this could have biased the
assessment of some end points, such as congestive
heart failure and the quality of life, which
have an element of subjectivity. To address this
issue, we used a tighter definition of congestive
heart failure (i.e., without renal replacement therapy).
In addition, the adjudication process by the
committee reviewing clinical end points should
have attenuated the possibility of bias because
The new england journal of medicine
n engl j med 355;20 www.nejm.org november 16, 2006
Page 149
committee members were unaware of patients’
study-group assignments.
In conclusion, our study showed that the use
of a target hemoglobin level of 13.5 g per deciliter
(as compared with a level of 11.3 g per deciliter) is
associated with an increased risk among patients
with anemia caused by chronic kidney disease.
Furthermore, no incremental improvement in the
quality of life was observed. Hence, we recommend
the use of a target hemoglobin level of 11.0
to 12.0 g per deciliter rather than a level of 11.0 to
13.0 g per deciliter to correct anemia in patients
with chronic kidney disease, because of increased
risk, a likely increased cost, and no quality-of-life
benefit. This study did not provide a mechanistic
explanation for the poorer outcome with the use
of a high target hemoglobin level. More studies
will be required to explore the role of the level of
hemoglobin and the dose of epoetin alfa to understand
these findings more completely.
Supported by Ortho Biotech Clinical Affairs and Johnson &
Johnson Pharmaceutical Research and Development, both subsidiaries
of Johnson & Johnson.
Dr. Singh reports receiving consulting fees from Ortho Biotech
Clinical Affairs, Amgen, Roche, Merck (Germany), Abbott,
Watson, and Horizon Blue Cross Blue Shield and lecture fees
from Ortho Biotech Clinical Affairs, Roche, Amgen, Abbott,
Watson, Scios, Pfizer, and Genzyme; serving on advisory boards
for Ortho Biotech Clinical Affairs, Roche, Acologix, Watson,
Advanced Magnetics, and Amgen; and receiving grant support
from Ortho Biotech Clinical Affairs, Dialysis Clinic, Roche,
Baxter, Johnson & Johnson, Amgen, Watson, and Aspreva. Dr.
Szczech reports receiving consulting fees from Ortho Biotech
Clinical Affairs, Nabi Pharmaceuticals, Gilead, Kuraha, Acologix,
and Roche; lecture fees from Nabi Biopharmaceuticals,
GlaxoSmithKline, Genzyme, Abbott, Amgen, and Ortho Biotech;
and grant support from Ortho Biotech Clinical Affairs. Dr.
Barnhart reports receiving consulting fees and grant support
from Ortho Biotech Clinical Affairs. Drs. Tang and Wolfson report
being employees of Ortho Biotech Clinical Affairs. Dr. Reddan
reports receiving consulting fees from Ortho Biotech Clinical
Affairs and Shire Pharmaceuticals; lecture fees from Amgen,
Novartis, Pfizer, AstraZeneca, and General Electric; and grant
support from Ortho Biotech Clinical Affairs, Amgen, and Novartis.
No other potential conflict of interest relevant to this article
was reported.
We thank Dr. Anil Londhe for biostatistical support; Drs. William
F. Owen, Jr., Mark Klausner, and Michael Corwin for help
in the development of the study protocol; and the many investigators,
study nurses, and coordinators who contributed to this
study.
Appendix
The following investigators participated in the CHOIR trial: Study Chairs: A.K. Singh and D. Reddan. Investigators: University of Tennessee,
Memphis — S. Acchiardo; Outcomes Research International, Hudson, FL — M.K. Acharya; University of Southern California, Los Angeles — M. Akmal;
Research Institute of Dallas, Dallas — S. Aronoff; North Shore University Hospital, Great Neck, NY — A. Ashfaq; Worcester Medical Center, Worcester, MA
— R. Black; Louisiana State University Medical Center, Shreveport — J. Blondin; Balboa Nephrology Medical Group, La Jolla, CA — M. Boiskin; University
of Virginia Health System, Charlottesville — W.K. Bolton; South Dakota Health Research Foundation, Sioux Falls — L. Burris; Clinica Las Americas,
San Juan, Puerto Rico — J.L. Cangiano; Emory Hypertension Research Center, Decatur, GA — A. Chapman; New York Hospital Medical Center of Queens,
Flushing — C. Charytan; Regional Kidney Disease Center/Associates in Nephrology, Erie, PA — E. Clark, F. Foti; Internal Medicine Specialists, Orlando,
FL — J. Cohen; Carolina Kidney Associates, Greensboro, NC — J.A. Coladonato; Renal Hypertension Physicians, Mount Laurel, NJ — M. Conrad;

Correction of Anemia and Chronic Kidney Disease
Southwest Kidney Institute, Tempe, AZ — R. Cooper; University of California, Los Angeles, Medical Center, Sylmar — D. Corry; Cleveland Clinic Foundation,
Cleveland — V. Dennis; Advanced Medical Research Institute, Fresno, CA — G. Dhillon; University of Miami Hospital and Clinics, Miami — J.
Diego; Virginia Commonwealth University, Richmond — S. DiGiovanni; St. Louis — D.T. Domoto; Research Center of Florida, Miami — F. Dumenigo;
Sislen and Associates, Washington, DC — G. Eisner; University of Southern California, Los Angeles — M. El Shahawy; Charles River Medical
Associates, Framingham, MA — L. Epstein; California Institute of Renal Research, San Diego — G. Fadda; Kidney Associates, Houston — S. Fadem;
Nephrology Associates, Rock Hill, SC — J. Fassler; Winthrop University Hospital, Mineola, NY — S. Fishbane; Talbert Medical Group, Huntington Beach,
CA — M. Fredrick; San Antonio Kidney Disease Center, San Antonio, TX — T. Fried; Wake Nephrology Associates, Raleigh, NC — L. Garrett; Western
New England Renal and Transplant Associates, Springfield, MA — M. Germain; South Florida Nephrology Associates, Lauderdale Lakes — R. Geronemus;
Georgetown University Medical Center, Washington, DC — J. Gonin; Renal Medical Group, Visalia, CA — R.J. Haley; Mayo Clinic, Jacksonville, FL
— W.E. Haley; Regional Kidney Disease Center, Erie, PA — R.D. Halligan; University of Oklahoma, Health Science Center, Oklahoma City — L. Haragsim;
Bronx Nephrology Hypertension, Bronx, NY — M. Henriquez; Nephrology Associates of Western New York, Amherst — T. Herman; North Shore
Diabetes and Endocrine Associates, New Hyde Park, NY — K. Hershon; Miami Kidney Group, Miami — D. Hoffman; University of California, Los Angeles,
Medical Center, Los Angeles — E. Jacobson; Indiana University Medical Center, Indianapolis — M. Jaradat; Empire Health Services, Spokane, WA
— S. Joshi; University Internal Medicine Associates, Cincinnati — S. Kant; Northwest Louisiana Nephrology, Shreveport — M. Kaskas; George Washington
University Medical Center, Washington, DC — P. Kimmel; Medical Group of North County, Vista, CA — S. Kipper; Stanford University Medical
Center, Stanford, CA — R. Lafayette; Apex Research of Riverside, Riverside, CA — J. Lee; Southbay Pharma Research, Buena Park, CA — S. Lee; University
of Arizona Health Sciences Center, Tucson — H.Y. Lien; Discovery Medical Research Group, Ocala, FL — H.R. Locay; Twin Cities Clinical Research,
Brooklyn Center, MN — N. Lunde; Bronx Westchester Medical Group, Bronx, NY — R. Lynn; Health Research Association, Los Angeles — H. Madkour;
Jefferson Nephrology, Charlottesville, VA — K. McConnell; Foundation Research, St. Petersburg, FL — M. McIvor; Arlington Nephrology, Arlington, TX
— B.R. Mehta; University of California, San Diego, Medical Center, San Diego — R. Mehta; Model Clinical Research, Baltimore — J.H. Mersey; Glendale
Internal Medicine and Cardiology Medical Group, Glendale, CA — R. Minasian; Hurley Medical Center, Flint, MI — A. Mohammed; University of Chicago
Hospitals, Chicago — P. Murray; Nephrology Associates of Orangeburg, Orangeburg, SC — M. Nassri; Genesis Clinical Research Corporation, Tampa,
FL — J. Navarro; University of California at Los Angeles, Los Angeles — A.R. Nissenson; State University of New York at Stony Brook, Stony Brook — E.
Nord; Empire Health Services, Spokane, WA — L.E. Obermiller; Morehouse School of Medicine, Atlanta — C. Obialo; Grand Street Medical Associates,
Kingston, NY — P. Pagnozzi; Odyssey Research Services, Bullhead City, AZ — S.M. Parekhehzad; InterMedix, Winter Haven, FL — E. Perez; Medical
College of Wisconsin, Milwaukee — W. Piering; Nephrology and Endocrine Associates Research, Las Vegas — N. Pokroy; Saint Clair Specialty Physicians,
Detroit — R. Provenzano; Columbia University Medical Center, New York — J. Radhakrishnan; Charlotte Nephrology Associates, Port Charlotte, FL — R.
Rajaram; InterMedix, Tampa, FL — G. Ramirez, G. Serrano; Nephrology Associates of Central Florida, Orlando — U. Ranjit; Citrus Nephrology,
Crystal River, FL — P. Reddy; East Bay Nephrology, San Pablo, CA — D. Ricker; Dallas Nephrology Associates, Dallas — J. Roman-Latorre; University
of Pennsylvania Health System–Presbyterian Medical Center, Philadelphia — M. Rudnick; State University of New York, Brooklyn — M. Salifu;
University of Kentucky Medical Center, Lexington — B.P. Sawaya; West Virginia University, Morgantown — R. Schmidt; Vanderbilt University Medical
Center, Nashville — G. Schulman; Drexel University College of Medicine, Philadelphia — A.B. Schwartz; Diabetes and Glandular Disease Research Associates,
San Antonio, TX — S.L. Schwartz; Springfield Gardens, NY — D. Scott; Clinical Research Solutions, Knoxville, TN — D.M. Sellers; Brookdale
University Hospital and Medical Center, Brooklyn, NY — W. Shapiro; Jefferson Renal Association, Philadelphia — K. Sharma; Rhode Island Hospital,
Providence — D. Shemin; Redrock Research Center, Las Vegas — L.K. Shete; Temple University Hospital, Philadelphia — P. Silva; Metabolism Associates,
New Haven, CT — D.B. Simon; Brigham and Women’s Hospital, Boston — A. Singh; FutureCare Studies, Springfield, MA — J. Slater; Nephrology
Associates, Augusta, GA — M. Smith; Fletcher Allen Health Care, Burlington, VT — R. Solomon; Ohio State University Hospitals, Columbus — D. Spetie;
University of Colorado Hospitals and Health Sciences Center, Denver — D. Spiegel; National Institute of Clinical Research, Los Angeles — M. Spira;
Evanston Northwestern Healthcare, Evanston, IL — S.M Sprague; Methodist Research Institute, Indianapolis — T. Taber; Salem Hospital, Salem, OR
— D. Tran; California Kidney Medical Group, Simi Valley — K. Tucker; Emory Clinic, Atlanta — J. Tumlin; Associates in Family Practice Clinical Research,
Davie, FL — M. Vacker; Associates in Nephrology, Fort Myers, FL — J. Van Sickler; Penn State Milton S. Hershey Medical Center, Hershey — N.
Verma; Long Island Jewish Medical Center, New Hyde Park, NY — J.D. Wagner; Baltimore — M. Waseem; Hypertension and Nephrology, Providence,
RI — M. Weinberg; West Palm Beach, FL — R. Weiss; Tennessee Cardiovascular Research Institute, Nashville — M. Wigger; Clinical Research Associates
of Tidewater, Norfolk, VA — D. Wombolt; Queens Medical Center, Honolulu — E. Wong; Western Nephrology and Metabolic Bone Disease, Lakewood,
CO — M. Yanover; Northwest Louisiana Nephrology, Shreveport — R.I. Zabaneh; and Southwest Nephrology, Evergreen Park, IL — D. Zikos; Clinical
Event Adjudication Committee: DCRI, Duke University School of Medicine, Durham, NC — K.W. Mahaffey (chair), D.T. Laskowitz, L.A. Szczech;
University of Texas Southwestern, Dallas — S. Hedayati; Hennepin County Medical Center, Minnepolis — C.A. Herzog; William Beaumont Hospital,
Royal Oak, MI — P. McCullough; Data and Safety Monitoring Board: Brigham and Women’s Hospital, Boston — M. Pfeffer (initial chair); Beth
Israel Deaconess Medical Center, Boston — T. Steinman (subsequent chair); Baylor College of Medicine, Houston — G. Eknoyan; Cleveland Clinic
Foundation, Cleveland — T. Greene; Washington University School of Medicine, St. Louis — S. Klahr; Dallas Transplant Institute, Dallas — T.F.
Parker; CHOIR Scientific Advisory Board: Rush University Medical Center, Chicago — G.L. Bakris; DCRI, Duke University School of Medicine, Durham,
NC — R.M. Califf; Minneapolis Medical Research Foundation, Minneapolis — R.N. Foley; State University of New York, Brooklyn — E.A. Friedman;
Stanford Medical Center, Stanford, CA — L.T. Goodnough; George Washington University, Washington, DC — P.L. Kimmel; University of
British Columbia, Vancouver, Canada — A. Levin; Georgia Medical Care Foundation, Atlanta — W. McClennan; Memorial University, St. John’s, NL, Canada
— P. Parfrey.
References
1. Astor BC, Muntner P, Levin A, Eustace 3. Ross SD, Fahrbach K, Frame D, Scheye
JA, Coresh J. Association of kidney function R, Connelly JE, Glaspy J. The effect of ane-
with anemia: the Third National Health and mia treatment on selected health-related
Nutrition Examination Survey (1988-1994). quality-of-life domains: a systematic re-
Arch Intern Med 2002;162:1401-8. view. Clin Ther 2003;25:1786-805.
2. Jones M, Ibels L, Schenkel B, Zagari 4. Levin A. Anemia and left ventricular
M. Impact of epoetin alfa on clinical end hypertrophy in chronic kidney disease pop-
points in patients with chronic renal failulations: a review of the current state of
ure: a meta-analysis. Kidney Int 2004;65: knowledge. Kidney Int Suppl 2002;80:35-8.
757-67.
5. Besarab A, Bolton WK, Browne JK, et
al. The effects of normal as compared with
low hematocrit values in patients with
cardiac disease who are receiving hemodialysis
and epoetin. N Engl J Med 1998;
339:584-90.
6. Strippoli GF, Manno C, Schena FP,
Craig JC. Haemoglobin and haematocrit
targets for the anaemia of chronic renal
disease. Cochrane Database Syst Rev 2003;
1:CD003967.
Page 150
n engl j med 355;20 www.nejm.org november 16, 2006 2097

2098
Correction of Anemia and Chronic Kidney Disease
7. Collins AJ, Li S, St Peter W, et al. mental health constructs. Med Care 1993;
Death, hospitalization, and economic as- 31:247-63.
sociations among incident hemodialysis 15. Lan KKG, DeMets DL. Discrete se-
patients with hematocrit values of 36 to quential boundaries for clinical trials.
39%. J Am Soc Nephrol 2001;12:2465-73. Biometrika 1983;70:659-63.
8. Xue JL, St Peter WL, Ebben JP, Everson 16. DeMets DL, Lan, KK. Interim analy-
SE, Collins AJ. Anemia treatment in the sis: the alpha spending function approach.
pre-ESRD period and associated mortality Stat Med 1994;13:1241-52.
in elderly patients. Am J Kidney Dis 2002; 17. Jennison C, Turnbull BW. Group se-
40:1153-61.
quential methods with applications to
9. National Kidney Foundation. K/DOQI clinical trials. Boca Raton, FL: Chapman
clinical practice guidelines for anemia of & Hall/CRC, 2000:29-31,
chronic kidney disease: update 2000. Am 18. Parfrey PS, Foley RN, Wittreich BH,
J Kidney Dis 2001;37:Suppl 1:S182-S238. Sullivan DJ, Zagari MJ, Frei D. Double-
10. Idem. K/DOQI clinical practice guideblind comparison of full and partial anelines
and clinical practice recommendamia correction in incident hemodialysis
tions for anemia in chronic kidney dis- patients without symptomatic heart disease.
Am J Kidney Dis 2006;47:Suppl 3: ease. J Am Soc Nephrol 2005;16:2180-9.
S11-S145.
19. Strippoli GF, Craig JC, Manno C,
11. Levey AS, Bosch JP, Lewis JB, Greene T, Schena FP. Hemoglobin targets for the
Rogers N, Roth D. A more accurate meth- anemia of chronic kidney disease: a metaod
to estimate glomerular filtration rate analysis of randomized controlled trials.
from serum creatinine: a new prediction J Am Soc Nephrol 2004;15:3154-65.
equation. Ann Intern Med 1999;130:461- 20. Foley RN, Parfrey PS, Morgan J, et al.
70.
Effect of hemoglobin levels in hemodialy-
12. Coates A, Dillenbeck CF, McNeil DR, sis patients with asymptomatic cardiomy-
et al. On the receiving end — II. Linear opathy. Kidney Int 2000;58:1325-35.
Analogue Self-Assessment (LASA) in eval- 21. Levin A, Djurdjev O, Thompson C, et al.
uation of aspects of the quality of life of Canadian randomized trial of hemoglo-
cancer patients receiving therapy. Eur J bin maintenance to prevent or delay left
Cancer Clin Oncol 1983;19:1633-7. ventricular mass growth in patients with
13. Hays RD, Kallich JD, Mapes DL, CKD. Am J Kidney Dis 2005;46:799-811.
Coons SJ, Carter WB. Development of the 22. McMahon LP, Roger SD, Levin A,
Kidney Disease Quality of Life (KDQOL) Slimheart Investigators Group. Develop-
instrument. Qual Life Res 1994;3:329-38. ment, prevention, and potential reversal
14. McHorney CA, Ware JE Jr, Raczek AE. of left ventricular hypertrophy in chronic
The MOS 36-Item Short-Form Health Sur- kidney disease. J Am Soc Nephrol 2004;15:
vey (SF-36): II. Psychometric and clinical 1640-7.
tests of validity in measuring physical and 23. Roger SD, McMahon LP, Clarkson A,
EARLY JOB ALERT SERVICE AVAILABLE AT THE NEJM CAREERCENTER
Register to receive weekly e-mail messages with the latest job openings
that match your specialty, as well as preferred geographic region,
practice setting, call schedule, and more. Visit the NEJM CareerCenter
at www.nejmjobs.org for more information.
n engl j med 355;20 www.nejm.org november 16, 2006
Page 151
et al. Effects of early and late intervention
with epoetin alfa on left ventricular mass
among patients with chronic kidney disease
(stage 3 or 4): results of a randomized
clinical trial. J Am Soc Nephrol 2004;
15:148-56.
24. Drüeke TB, Locatelli F, Clyne N, et al.
Normalization of hemoglobin level in patients
with chronic kidney disease and
anemia. N Engl J Med 2006;355:2071-84.
25. Rossert J, Levin A, Roger SD, et al. Effect
of early correction of anemia on the
progression of CKD. Am J Kidney Dis
2006;47:738-50.
26. Furuland H, Linde T, Ahlmen J, Christensson
A, Strombom U, Danielson BG. A
randomized controlled trial of haemoglobin
normalization with epoetin alfa in
pre-dialysis and dialysis patients. Nephrol
Dial Transplant 2003;18:353-61.
27. Wanner C, Krane V, Marz W, et al.
Atorvastatin in patients with type 2 diabetes
mellitus undergoing hemodialysis.
N Engl J Med 2005;353:238-48. [Erratum,
N Engl J Med 2005;353:1640.]
28. Eknoyan G, Beck GJ, Cheung AK, et
al. Effect of dialysis dose and membrane
flux in maintenance hemodialysis. N Engl
J Med 2002;347:2010-9.
29. Parfrey PS. Target hemoglobin level
for EPO therapy in CKD. Am J Kidney Dis
2006;47:171-3.
30. Gouva C, Nikolopoulos P, Ioannidis JP,
Siamopoulos KC. Treating anemia early in
renal failure patients slows the decline of
renal function: a randomized controlled
trial. Kidney Int 2004;66:753-60.
Copyright © 2006 Massachusetts Medical Society.

Estimated Effect of Epoetin Dosage on Survival among
Elderly Hemodialysis Patients in the United States
Yi Zhang,* Mae Thamer,* Dennis Cotter,* James Kaufman, † and Miguel A. Hernán ‡
*Medical Technology and Practice Patterns Institute, Bethesda, Maryland; and † Renal Section, Veterans Affairs Boston
Healthcare System and Boston University School of Medicine, and ‡ Department of Epidemiology, Harvard School of
Public Health, and Harvard-MIT Division of Health Sciences and Technology, Boston, Massachusetts
Background and objectives: The common finding that low achieved hemoglobin in observational studies and high target
hemoglobin in randomized trials each were associated with increased mortality and high epoetin dosage has suggested the
possibility that high epoetin dosage might explain the increased mortality risk.
Design, setting, participants, & measurements: We considered data from 18,454 patients who were >65 yr, were in the US
Renal Data System, started hemodialysis in 2003, and survived 3 mo on dialysis. We estimated the association between
cumulative average epoetin dosage and survival through the subsequent 9 mo by using inverse probability weighting to adjust
for time-dependent confounding by indication.
Results: Survival was similar throughout the entire follow-up period for the three hypothetical treatment regimens selected:
Low dosage 15,000 U/wk, medium dosage 30,000 U/wk, and high dosage 45,000 U/wk. Compared with a cumulative average
dosage of 20,000 to 30,000 U/wk, the estimated hazard ratio (HR; 95% confidence interval [CI]) was 0.90 (0.52 to 1.54) for

Institutional claims, including treatment information, were used as the
primary data set and merged with variables from patient, medical
evidence, facility, and physician/supplier files from the USRDS core
CD based on unique patient identifiers.
We identified incident hemodialysis patients who were �65 yr and
had their first ESRD service in 2003. We chose an elderly incident
cohort so that we would have the complete data on their epoetin
therapy (elderly patients are already enrolled in the Medicare program,
unlike their younger counterparts who initiate dialysis, and do not
have to wait for 3 mo before receiving Medicare dialysis benefits).
Beginning in 2003, patients from hospital-based facilities might receive
darbepoetin (Aranesp®; Amgen Inc., Thousand Oaks, CA) instead of
epoetin alfa (11), the therapy of interest in this study, so we restricted
our study population to patients who received dialysis care in freestanding
facilities. Patients who did not receive their first outpatient
dialysis claims in the first 90 d after being certified as having ESRD and
those identified as having Medicare as a secondary payer were also
excluded to ensure that we had the complete patient claims treatment
history. The analysis was further restricted to patients who, during the
first 3 mo on hemodialysis, did not have missing data, a history of HIV
or cancer (studies suggest that patients with ESRD and HIV might
respond differently to epoetin therapy compared with the ESRD population
at large) (12–14), a change of dialysis modality, a kidney transplant,
a change of dialysis provider, or a gap in billable dialysis services
(defined as missing dialysis treatment or hospitalization information
for �30 consecutive days).
Follow-up started after 3 complete months of outpatient dialysis
treatment and ended 9 mo later; at the time of death; at the time of
censoring by change of dialysis modality (3.2% censored), a kidney
transplant (0.5%), 60 d after change of dialysis provider (8.4%), or a gap
in outpatient dialysis services (11.7%); or as a result of data anomalies
(3%), whichever happened earliest. We could not use cause-specific
mortality (e.g., cardiovascular mortality) as an outcome because the
information on causes of death in the USRDS is not validated.
Statistical Analysis
We fit a pooled logistic model to estimate the probability of death at
each month conditional on cumulative average log epoetin dosage
(cubic splines with knots located at 5th, 25th, 75th, and 95th percentiles
corresponding to weekly dosages of 3000, 8330, 22,900 and 46,950 U,
respectively) through that month. For each patient and month, the
cumulative average dosage was calculated as the total epoetin dosage
received since the start of follow-up divided by the number of weeks
elapsed. The model included covariates for month of follow-up (linear
and quadratic terms) and their product (interaction) terms with the log
epoetin dosage variables (log of zero dose was treated as zero). The
model also included the non–time-varying covariates age (years) at
ESRD onset, race (black, white, or other), gender, underlying cause of
ESRD (diabetes, glomerulonephritis, hypertension, cystic kidney, or
other), US geographic region (Northeast, Southeast, Midwest, or West),
dialysis chain membership (five largest chains and small/nonchain
facilities); hematocrit level (�30, 30 to �33, 33 to �36, 36 to �39, or
�39%), patient weight (�61, 61 to �72, 72 to �84, or �84 kg), Charlson
index score (in four groups, �3 to�6,6to�8, or �8), and presence of
cardiovascular and noncardiovascular comorbidities (15) at the start of
follow-up (cardiovascular comorbidities include cardiac arrest, congestive
heart failure, cerebrovascular disease, cardiac dysrhythmia, pericarditis,
peripheral vascular disease, ischemic heart disease, and myocardial
infarction; noncardiovascular comorbidities include alcohol
dependence, cancer, drug dependence, HIV, AIDS, inability to ambulate,
inability to transfer, chronic obstructive pulmonary disease, tobacco
use, diabetes, currently on insulin, diabetes, primary or contrib-
Page 153
Clin J Am Soc Nephrol 4: 638–644, 2009 Epoetin Dosage and Survival 639
uting), and average monthly intravenous iron administered (obtained
using Healthcare Common Procedure Coding System codes and categorized
in quartiles as �450, 450 to �800, 800 to �1200, or �1200 mg),
cumulative hospital days (0, 0 to �5,5to�10, or �10 d), and average
epoetin dosage and number of administrations during the first 3 mo on
dialysis.
We then plotted the predicted survival under three hypothetical
scenarios: All patients received a cumulative average dosage of (1)
15,000 U/wk, (2) 30,000 U/wk, and (3) 45,000 U/wk. These levels were
selected to represent approximately low, average, and high dosages. To
estimate average HRs during the entire follow-up, we also fitted the
model without interaction terms and with continuous epoetin dosage
replaced by a categorical variable with five levels: 0 to �10,000; 10,000
to �20,000; 20,000 to �30,000; 30,000 to �40,000; and �40,000 U/wk.
We computed conservative 95% CIs by using a robust variance.
Because high epoetin dosages are more likely to be prescribed to
patients with higher mortality risk (e.g., with lower hematocrit) (6), the
estimates from the regression model need to be adjusted for this timedependent
confounding by indication effect. To do so, one could add
the time-varying confounders (e.g., hematocrit) as covariates in the
logistic regression model; however, this standard approach might introduce
bias because hematocrit is affected by previous epoetin treatment
(and might be on the causal pathway between epoetin treatment
and mortality) (16). We therefore used another method to adjust for
time-dependent confounding by indication: Inverse probability weighting.
Unlike the standard approach, inverse probability weighting appropriately
adjusts for measured time-dependent confounders that are
affected by previous epoetin therapy (e.g., previous hematocrit). Formally,
under the assumption that all time-varying predictors of both
epoetin therapy and hematocrit were included in the analysis (as described
in the next paragraph), our weighted model estimates the
parameters of a marginal structural Cox model (17–19) and can be used
to mimic a sequentially randomized trial in which patients were assigned
to different epoetin dosing regimens.
Each patient in the logistic models received a time-varying weight
inversely proportional to the estimated probability of having his or her
own observed epoetin dosage history, as described elsewhere (20).
These weights were estimated by fitting two nested models: A logistic
regression model to estimate each patient’s probability of not receiving
epoetin at any given month (7.3% of the patient-months had zero dose)
and a linear regression model to estimate each patient’s density (assumed
to be normal) of log epoetin dosage among those with nonzero
dose in that month. Both models included the covariates listed already
for the mortality model plus the following time-varying covariates
measured during the previous month: Hematocrit (�24, 24 to �27, 27
to �30, 30 to �33, 33 to �36, 36 to �39, and �39%; most recent
hematocrit levels were carried forward when hematocrit value was
missing), inpatient days (0, 0 to �4,4to�7, �7), iron treatment (yes,
no), and log epoetin dosage (cubic splines with five knots located at 5.5,
9.5, 10.5, 11.5, and 13.5, which correspond to weekly epoetin dosages of
60, 3340, 9100, 24,580, and 182,350 U, respectively). The model also
included product terms between log epoetin dosage spline variables
and time-varying hematocrit and between chains and both baseline
epoetin dosage and time-varying epoetin dosage (because chain characteristics
are associated with epoetin dosing patterns [21]). Inverse
probability weights were also computed to adjust for potential selection
bias as a result of censoring. Both the epoetin and censoring weights
were stabilized and their product was used to fit the weighted regression
model.
Because Medicare does not reimburse separately for epoetin treatment
in the hospital, no data on inpatient use are available. Furthermore,
there are no empirical studies on inpatient epoetin use, and

anecdotal evidence varies
widely regarding how often
and how much epoetin is
prescribed to dialysis inpatients;
however, because patients
with longer hospital
stays might be more likely to
receive some epoetin, we selected
a primary analysis in
which patients with �4 inpatient
days were assumed to
receive the same epoetin dosage
from day 5 onward as
they received during their
most recent preceding outpatient
dialysis period (per
dialysis session). We also conducted
two secondary analyses:
Analysis A, which assumes
that epoetin was given throughout the duration of the hospital
stay, and analysis B, which assumes that epoetin was not given during
a hospital stay, regardless of the duration. All analyses were conducted
with SAS 9.1 (SAS Institute, Cary, NC).
Results
Figure 1 depicts the selection process for the patients who
were included in the analysis. The final cohort consisted of
18,454 patients. The demographics, clinical history, comorbidities,
and baseline anemia management of the study population
are shown in Table 1. The study cohort was 51% men, 69%
white, 47% with diabetes, and 63% with cardiovascular comorbidities.
During the first 3 mo on dialysis, the average weekly
epoetin dosage was �10,000 U in 11% of the patients, 10,000 to
20,000 U in 36%, 20,000 to 30,000 Us in 30%, and �30,000 U in
24%. At the start of follow-up, 83% had a hematocrit �33%,
60% of the patients had a hematocrit �36%, and 35% had a
hematocrit of �39%. During the follow-up, 4105 (22%) patients
died and 4906 (27%) were censored before the end of the study.
The mean of the estimated inverse probability weights was 1.05
(SD 5.2), and the 1st and 99th percentiles were 0.06 and 6.20,
respectively.
Figure 2 shows the estimated survival functions under
three hypothetical dosing levels: 15,000 U/wk, 30,000 U/wk,
and 45,000 U/wk throughout the entire follow-up. Our primary
analysis estimated similar survival at 9 mo for all three
dosage levels. Secondary analyses indicate that survival
probabilities vary depending on assumptions of epoetin use
during hospitalizations. Specifically, the lowest epoetin dosage
(15,000 U/wk) resulted in the highest survival when
assuming that epoetin was used throughout the hospital stay
(analysis A; see supplemental information available online at
http://www.cjasn.org) but in the lowest survival when assuming
that epoetin was not used in hospital at all (analysis
B; see supplemental information available online at http://
www.cjasn.org). The main features of the estimated epoetin
dosage-survival curves were not sensitive to the particular
functional form used for cumulative average dosage (results
did not change when we used cubic splines with knots at
different locations) or to alternative summaries of dosage
Page 154
640 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 4: 638–644, 2009
Figure 1. Selection of study population from US Renal Data System (USRDS) data.
history (i.e., total cumulative dosage from start of follow-up,
recent and past cumulative average dosage).
Table 2 presents the estimated average HRs during the entire
follow-up from our weighted model under different assumptions
regarding the use of epoetin during hospital stays. These
average HRs are consistent with the survival curves (Figure 2).
The HRs for the two lowest epoetin dosage groups are sensitive
to the assumptions about epoetin use during hospitalizations,
but the HRs for the two highest epoetin dosage group are less
sensitive to these assumptions. In the primary analysis, compared
with a cumulative average dosage of 20,000 to �30,000
U/wk, the HR estimates (95% CI) were 0.91 (0.67 to 1.22) for
�40,000 U/wk, 0.96 (0.76 to 1.21) for 30,000 to �40,000 U/wk,
0.84 (0.67 to 1.05) for 10,000 to �20,000 U/wk, and 0.90 (0.52 to
1.54) for �10,000 U/wk (P � 0.83 for trend).
Our estimates did not materially change (data not shown)
when we used alternative categorizations of hospital days and
hematocrit values, when we expanded our billable service gap
definition from 30 to 60 d, when we did not censor on change
of provider, when we used different knots’ locations for splines
of log epoetin dosage, when the estimated weights were truncated
to a maximum of 20, and when the weights were estimated
under a � distribution for the log of epoetin dosage. We
also tested the effect of using a log scale to compute the cumulative
epoetin dosage. In the primary analysis, compared with
a cumulative average dosage of 20,000 to �30,000 U/wk, use of
the logarithmic scale resulted in HR estimates (95% CI) of 1.15
(0.92 to 1.44) for �40,000 U/wk, 1.06 (0.86 to 1.31) for 30,000 to
�40,000 U/wk, 1.00 (0.83 to 1.20) for 10,000 to �20,000 U/wk,
and 0.81 (0.56 to 1.17) for �10,000 U/wk (P � 0.19 for trend).
Compared with the primary analysis in Figure 2, the 9-mo
survival estimated using the log-transformed dosage was similar
for the lowest dosage group (15,000 U/wk) and 4% lower
for both the middle and higher dosage groups (30,000 and
45,000 U/wk).
For comparison purposes, Table 3 shows the corresponding
estimated HRs from standard unweighted models that include
the measured time-varying confounders as covariates. In both
the primary and secondary analyses, these standard Cox re-

Table 1. Characteristics of study population (n � 18,454) a
Characteristic Value
Patient demographics
race (%)
white 69.4
black 25.3
other/unknown 5.3
age (yr; mean) 75.7
male gender (%) 51.2
Facility characteristics
region (%)
Northeast (networks 1 to 5) 24.0
Southeast (networks 6 to 8, 13, 14) 35.4
Midwest (networks 9 to 12) 23.9
West (networks 15 to 18) 16.6
chain membership (%)
chain 1 30.6
chain 2 14.1
chain 3 15.1
chain 4 8.8
chain 5 3.8
chain 6 2.2
small chain/nonchain 25.4
Patient clinical history (%)
cause of ESRD
diabetes 46.9
hypertension 36.1
glomerulonephritis 5.1
cystic kidney 1.0
other 10.9
body weight (kg)
�61 23.9
61 to �72 25.7
72 to 84 24.3
�84 26.2
Charlson index score b
�3 18.2
3to�6 29.4
6to�8 21.0
�8 31.4
baseline hospitalization (d) b
0 56.1
�5 11.4
5to�10 12.8
�10 19.7
cardiovascular comorbidities (% yes) 63.1
noncardiovascular comorbidities (% yes) 61.4
Anemia management practices
receipt of predialysis epoetin (% yes) 35.1
hematocrit level at end of month 3 (%)
�30 6.5
30 to �33 10.5
33 to �36 20.1
36 to �39 27.6
�39 35.3
Table 1. Continued
Page 155
Clin J Am Soc Nephrol 4: 638–644, 2009 Epoetin Dosage and Survival 641
Characteristic Value
epoetin dosage administered (U/wk) b
�10,000 9.8
10,000 to �20,000 34.5
20,000 to �30,000 30.5
30,000 to �40,000 14.5
�40,000 10.6
Average iron administered (mg/mo) b
�450 20.2
450 to �800 19.4
800 to �1200 24.6
�1200 35.8
a All variables in Table 1 were adjusted for in the survival
analyses. Unless otherwise noted, all variables were collected
at study entry.
b First 3 mo of dialysis therapy.
Figure 2. Survival probabilities for three hypothetical regimens:
Low dosage (15,000 U/wk), medium dosage (30,000 U/wk),
and high dosage (45,000 U/wk).
gression models suggest that mortality risk increases with epoetin
dosage.
Discussion
We used data from an elderly hemodialysis population to
answer the question, “What would be the survival if every
patient were to receive the same epoetin dosage during the
entire follow-up?” We found no indication of a harmful effect
of higher epoetin dosages on survival during months 4 through
12 after dialysis initiation among patients aged �65 yr, who
comprise 50% of the patients with newly diagnosed ESRD and
one third of the prevalent hemodialysis population.

Table 2. Cumulative average epoetin dosage and mortality HRs for weighted regression models a
U/wk
Patient-
Months
No. of
Deaths
%
Deaths
Our finding of no harmful effect of high dosages is unlikely
to be explained by unmeasured confounding by indication,
because better confounder measurement (e.g., more frequent
hematocrit measurements or information on other prognostic
factors such as nutritional status, BP, or inflammatory state)
would be expected to result in effect estimates that are further
from a harmful effect because patients who receive higher
dosages tend to be those with more severe disease.
The concern about the relationship between high epoetin
dosages and mortality has been raised by the seeming paradox
between the association of higher hematocrit levels with better
survival found in observational studies (3–7) and the adverse
effects of targeting higher hematocrit levels found in randomized
clinical trials (1,2). Because in observational studies higher
hematocrit levels are associated with lower epoetin dosages,
whereas in randomized trials the group assigned to higher
hematocrit targets receives higher epoetin dosages, some ob-
Primary Analysis Secondary Analysis A Secondary Analysis B
HR 95% CI HR 95% CI HR 95% CI
�10,000 41,113 776 1.9 0.90 0.52 to 1.54 0.56 0.38 to 0.82 1.21 0.79 to 1.83
10,000 to �20,000 45,251 1111 2.5 0.84 0.67 to 1.05 0.84 0.65 to 1.09 1.12 0.82 to 1.55
20,000 to �30,000
(reference
group)
20,682 722 3.5 Reference Reference Reference
30,000 to �40,000 9070 390 4.3 0.96 0.76 to 1.21 0.99 0.82 to 1.21 0.95 0.69 to 1.30
�40,000 9708 499 5.1 0.91 0.67 to 1.22 1.04 0.84 to 1.29 0.88 0.68 to 1.14
Total 125,824 3498 2.8
a The weighted models include baseline covariates (age, race, gender, underlying cause of ESRD, geographic region, Charlson
index score, cardiovascular and noncardiovascular comorbidities, baseline hematocrit, average intravenous iron administered,
length of inpatient stays in the baseline, and average epoetin dosage and number of administrations at baseline). Variables in
the models to estimate weights include baseline covariates as listed above plus time-varying covariates including hematocrit,
inpatient days, iron treatment, log epoetin dosage, and interaction terms between log epoetin dosage and time-varying
hematocrit and between chains and both baseline epoetin dosage and time-varying epoetin dosage. Primary analysis: Epoetin
was imputed for patients with a hospital stay �4 d on the basis of the previous outpatient epoetin prescription; secondary
analysis A: Epoetin was imputed throughout duration of hospital stay; secondary analysis B: Epoetin was assumed not to be
administered during hospital stay. Distribution of patient-months, number of deaths, and percentage death are based on
primary analysis.
Table 3. Cumulative average epoetin dosage and mortality HRs from standard (unweighted) regression models a
U/wk
Page 156
642 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 4: 638–644, 2009
Primary Analysis Secondary Analysis A Secondary Analysis B
HR 95% CI HR 95% CI HR 95% CI
�10,000 0.73 0.65 to 0.83 0.70 0.62 to 0.79 0.85 0.75 to 0.96
10,000 to �20,000 0.82 0.74 to 0.91 0.80 0.72 to 0.88 0.89 0.80 to 0.99
20,000 to �30,000
(reference group)
Reference Reference Reference
30,000 to �40,000 1.16 1.02 to 1.33 1.09 0.95 to 1.24 1.11 0.96 to 1.28
�40,000 1.26 1.10 to 1.44 1.26 1.10 to 1.43 1.27 1.10 to 1.46
a The standard regression models include baseline covariates (age, race, gender, underlying cause of ESRD, geographic
region, Charlson index score, cardiovascular and noncardiovascular comorbidities, baseline hematocrit, average intravenous
iron administered, length of inpatient stays in the baseline, and average epoetin dosage and number of administrations at
baseline) and time-varying covariates (previous month’s hematocrit, iron treatment, and length of hospital stay).
servers have suggested that mortality might be related to epoetin
dosage rather than achieved hematocrit level. In fact, several
extrahematopoietic effects of epoetin administration,
particularly worsening of hypertension, might account for an
increased mortality as a result of higher epoetin dosages (22);
however, all studies that suggested increased mortality with
higher epoetin dosages (3,4,5,7), including our own (6), and a
recent secondary analysis of a randomized trial (10) used
standard adjustment techniques. (Another study found that
the estimates were sensitive to modeling assumptions [23].)
These analyses might not have appropriately adjusted for
time-varying confounders that are affected by past epoetin
use and that might be on the causal pathway between epoetin
use and mortality. These confounders include hematocrit
level (the major determinant of epoetin dosage in clinical
practice) and previous hospitalization (a key prognostic factor
for death). In this study, we confirm the previous observations

that higher epoetin dosages are associated with increased mortality
when using standard adjustment techniques but were
unable to show a harmful association between high epoetin
dosage and mortality when we used inverse probability
weighting to adjust for measured time-varying confounders.
A limitation of our study is the lack of data on epoetin use
during hospitalizations. Our analysis suggests that the estimated
survival at lower epoetin dosages is quite sensitive to
assumptions regarding epoetin use in the hospital. Future studies
that provide more details regarding inpatient epoetin use
are needed to determine the mortality risk associated with low
epoetin dosages; however, regardless of the uncertainty for
mortality under low epoetin dosages, we did not find strong
support for an effect, either harmful or beneficial, of high
epoetin dosages on survival.
The results of our study may not be generalizable to patients
other than those included in the analysis, who were �65 yr,
incident to dialysis and survived the first 90 d of dialysis, and
are dialyzed in free-standing facilities. Furthermore, our analysis
included only limited center-specific attributes (profit and
chain status), and center effects may be associated with epoetin
dosing as well as mortality.
Our findings address but do not fully resolve the seeming
paradox between observational studies and RCTs regarding
the effect of hematocrit on mortality. The discrepancy might
be explained by a modification of the effect of high epoetin
dosages by patient characteristics, which differ for patients
who are included in observational studies compared with
those who are included in RCTs: One of the RCTs enrolled
patients who had chronic kidney disease and were not on
dialysis, and the other included prevalent hemodialysis patients
with underlying cardiovascular disease. Additional
appropriately adjusted analyses of observational databases,
as well as randomized clinical trials, are necessary to understand
the relationship among epoetin, hematocrit, and survival
and help design treatment algorithms that maximize
benefit.
Disclosures
All authors participated in the design, analysis, interpretation, writing,
and/or editing of this study and have seen and approved the final
version. Y.Z. had full access to all of the data in the study and had final
responsibility for the decision to submit for publication.
J.K. has served as a consultant and receives financial support from
Amgen and F. Hoffman-La Roche.
Acknowledgments
This project was funded in part by National Institutes of Health
grants R01-DK066011-01A2 and R01-HL080644-01. The National Institutes
of Health had no role in the study design; in the collection,
analysis, and interpretation of data; in the writing of the report; and in
the decision to submit the paper for publication. The data reported here
have been supplied by the US Renal Data System. The interpretation
and reporting of these data are the responsibility of the authors and in
no way should be seen as the official policy or interpretation of the US
government.
Page 157
Clin J Am Soc Nephrol 4: 638–644, 2009 Epoetin Dosage and Survival 643
References
1. Besarab A, Bolton WK, Browne JK, Egrie JC, Nissenson AR,
Okamoto DM, Schwab SJ, Goodkin DA: The effects of
normal as compared with low hematocrit values in patients
with cardiac disease who are receiving hemodialysis
and epoetin. N Engl J Med 339: 584–590, 1988
2. Singh AK, Szczech L, Tang KL, Barnhart H, Sapp S, Wolfson
M, Reddan D, CHOIR Investigators: Correction of
anemia with epoetin alfa in chronic kidney disease. N Engl
J Med 355: 2085–2098, 2006
3. Collins AJ, Li S, St Peter W, Ebben J, Roberts T, Ma JZ,
Manning W: Death, hospitalization, and economic associations
among incident hemodialysis patients with hematocrit
values of 36 to 39%. J Am Soc Nephrol 12: 2465–2473,
2001
4. Lowrie EG, Huang WH, Lew NL, Liu Y: The relative
contribution of measured variables to death risk among
hemodialysis patients. In: Death on Hemodialysis, edited
by Friedman EA, Dordrecht, Kluwer Academic Publishers,
1994, pp 121–141
5. Ma JZ, Ebben J, Xia H, Collins AJ: Hematocrit level and
associated mortality in hemodialysis patients. J Am Soc
Nephrol 10: 610–619, 1999
6. Zhang Y, Thamer M, Stefanik K, Kaufman J, Cotter DJ:
Epoetin requirements predict mortality in hemodialysis
patients. Am J Kidney Dis 44: 866–876, 2004
7. Madore F, Lowrie EG, Brugnara C, Lew NL, Lazarus JM,
Bridges K, Owen WF: Anemia in hemodialysis patients:
Variables affecting this outcome predictor. J Am Soc Nephrol
8: 1921–1929, 1997
8. Strippoli GF, Tognoni G, Navaneethan SD, Nicolucci A,
Craig JC: Haemoglobin targets: We were wrong, time to
move on. Lancet 369: 346–350, 2007
9. Busko M: Higher haemoglobin targets linked to increased
mortality risk in CKD patients with anemia, February 5,
2007, Available at: http://www.medscape.com/viewarticle/
551687. Accessed May 22, 2008
10. Szczech LA, Barnhart HX, Inrig JK, Reddan DN, Sapp S,
Califf RM, Patel UD, Singh AK: Secondary analysis of the
CHOIR trial epoetin-alpha dose and achieved hemoglobin
outcomes. Kidney Int 74: 791–798, 2008
11. EPO & DPO Use by Provider, Figure 5.39. In: US Renal Data
System, USRDS 2006 Annual Data Report: Atlas of End-Stage
Renal Disease in the United States, Bethesda, National Institutes
of Health, National Institute of Diabetes and Digestive
and Kidney Diseases, 2006. Available at: http://www.
usrds.org/2006/pdf/05_clin_ind_prev_hlth_06.pdf. Accessed
January 31, 2008
12. Henke M, Laszig R, Rube C, Schäfer U, Haase KD,
Schilcher B, Mose S, Beer KT, Burger U, Dougherty C,
Frommhold H: Erythropoietin to treat head and neck cancer
patients with anemia undergoing radiotherapy. Lancet
362: 1255–1260, 2003
13. Leyland-Jones B; BEST investigators: Breast cancer trial
with erythropoietin terminated unexpectedly. Lancet 4:
459–460, 2003
14. Ifudu O, Mathew JJ, Mayers JD, Macey LJ, Brezsnyak W,
Reydel C, McClendon E, Surgrue T, Rao S, Friedman EA:
Severity of AIDS and response to EPO in uremia. Am J
Kidney Dis 30: 28–35, 1997
15. Frankenfield DL, Roman SH, Rocco MV, Bedinger MR,

644 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 4: 638–644, 2009
McClellan WM: Disparity in outcomes for adult Native
American hemodialysis patients? Findings from the ESRD
clinical performance measures project, 1996 to 1999. Kidney
Int 65: 1426–1434, 2004
16. Hernán MA, Hernandez-Diaz S, Robins JM: A structural
approach to selection bias. Epidemiology 15: 615–625, 2004
17. Robins JM, Hernán MA, Brumback B: Marginal structural
models and causal inference in epidemiology. Epidemiology
11: 550–560, 2000
18. Hernán MA, Brumback B, Robins JM: Marginal structural
models to estimate the causal effect of zidovudine on the
survival of HIV-positive men. Epidemiology 11: 561–570,
2000
19. Hernán MA, Brumback B, Robins JM: Estimating the
causal effect of zidovudine on CD4 count with a marginal
structural model for repeated measures. Stat Med 21: 1689–
1709, 2002
20. Cotter D, Zhang Y, Thamer M, Kaufman J, Hernán M: The
effect of epoetin dose on hematocrit. Kidney Int 73: 347–353,
2008
21. Thamer M, Zhang Y, Kaufman J, Cotter D, Hernán MA:
Dialysis facility ownership and epoetin dosing in patients
receiving hemodialysis. JAMA 297: 1667–1674, 2007
22. Vaziri ND: Cardiovascular effects of erythropoietin and
anemia correction. Curr Opin Nephrol Hypertens 10: 633–
637, 2001
23. Bradbury BD, Wang O, Critchlow CW, Rothman KJ, Heagerty
P, Keen M, Acquavella JF: Exploring relative mortality
and epoetin alfa dose among hemodialysis patients.
Am J Kidney Dis 51: 62–70, 2008
Access to UpToDate on-line is available for additional clinical information
at http://www.cjasn.org/
Page 158

Technical Appendix 3 – PI for Aranesp and EPOGEN
Technical Appendix 3 – PI for Aranesp and EPOGEN
Page 159

Aranesp ®
(darbepoetin alfa)
For Injection
WARNINGS: INCREASED MORTALITY, SERIOUS CARDIOVASCULAR EVENTS,
THROMBOEMBOLIC EVENTS, STROKE and INCREASED RISK OF TUMOR PROGRESSION OR
RECURRENCE
Chronic Renal Failure:
Page 160
• In clinical studies, patients experienced greater risks for death, serious cardiovascular events,
and stroke when administered erythropoiesis-stimulating agents (ESAs) to target hemoglobin
levels of 13 g/dL and above.
• Individualize dosing to achieve and maintain hemoglobin levels within the range of 10 to 12
g/dL.
Cancer:
• ESAs shortened overall survival and/or increased the risk of tumor progression or recurrence
in some clinical studies in patients with breast, non-small cell lung, head and neck, lymphoid,
and cervical cancers (see WARNINGS: Table 1).
• To decrease these risks, as well as the risk of serious cardio- and thrombovascular events,
use the lowest dose needed to avoid red blood cell transfusion.
• Because of these risks, prescribers and hospitals must enroll in and comply with the ESA
APPRISE Oncology Program to prescribe and/or dispense Aranesp ® to patients with cancer.
To enroll in the ESA APPRISE Oncology Program, visit www.esa-apprise.com or call 1-866-
284-8089 for further assistance.
• Use ESAs only for treatment of anemia due to concomitant myelosuppressive chemotherapy.
• ESAs are not indicated for patients receiving myelosuppressive therapy when the anticipated
outcome is cure.
• Discontinue following the completion of a chemotherapy course.
(See WARNINGS: Increased Mortality, Serious Cardiovascular Events, Thromboembolic Events,
and Stroke, WARNINGS: Increased Mortality and/or Increased Risk of Tumor Progression or
Recurrence, INDICATIONS AND USAGE, and DOSAGE AND ADMINISTRATION.)
DESCRIPTION
Aranesp ® is an erythropoiesis stimulating protein, closely related to erythropoietin, that is produced in
Chinese hamster ovary (CHO) cells by recombinant DNA technology. Aranesp ® is a 165-amino acid
protein that differs from recombinant human erythropoietin in containing 5 N-linked oligosaccharide
chains, whereas recombinant human erythropoietin contains 3 chains. 1 The two additional N-glycosylation
sites result from amino acid substitutions in the erythropoietin peptide backbone. The additional
carbohydrate chains increase the approximate molecular weight of the glycoprotein from 30,000 to
37,000 daltons. Aranesp ® is formulated as a sterile, colorless, preservative-free protein solution for
intravenous or subcutaneous administration.
Single-dose vials are available containing 25, 40, 60, 100, 150, 200, 300, or 500 mcg of Aranesp ® .
1

Page 161
Single-dose prefilled syringes and prefilled SureClick autoinjectors are available containing 25, 40,
60, 100, 150, 200, 300, or 500 mcg of Aranesp ® . Each prefilled syringe is equipped with a needle guard
that covers the needle during disposal.
Single-dose vials, prefilled syringes and autoinjectors are available in two formulations that contain
excipients as follows:
Polysorbate solution Each 1 mL contains 0.05 mg polysorbate 80, and is formulated at pH 6.2 ±
0.2 with 2.12 mg sodium phosphate monobasic monohydrate, 0.66 mg sodium phosphate dibasic
anhydrous, and 8.18 mg sodium chloride in Water for Injection, USP (to 1 mL).
Albumin solution Each 1 mL contains 2.5 mg albumin (human), and is formulated at pH 6.0 ±
0.3 with 2.23 mg sodium phosphate monobasic monohydrate, 0.53 mg sodium phosphate dibasic
anhydrous, and 8.18 mg sodium chloride in Water for Injection, USP (to 1 mL).
CLINICAL PHARMACOLOGY
Mechanism of Action
Aranesp ® stimulates erythropoiesis by the same mechanism as endogenous erythropoietin. A primary
growth factor for erythroid development, erythropoietin is produced in the kidney and released into the
bloodstream in response to hypoxia. In responding to hypoxia, erythropoietin interacts with progenitor
stem cells to increase red blood cell (RBC) production. Production of endogenous erythropoietin is
impaired in patients with chronic renal failure (CRF), and erythropoietin deficiency is the primary cause of
their anemia. Increased hemoglobin levels are not generally observed until 2 to 6 weeks after initiating
treatment with Aranesp ® (see DOSAGE AND ADMINISTRATION). In patients with cancer receiving
concomitant chemotherapy, the etiology of anemia is multifactorial.
Pharmacokinetics
Adult Patients
The pharmacokinetics of Aranesp ® were studied in patients with CRF receiving or not receiving dialysis
and cancer patients receiving chemotherapy.
Following intravenous administration in CRF patients receiving dialysis, Aranesp ® serum concentrationtime
profiles were biphasic, with a distribution half-life of approximately 1.4 hours and a mean terminal
half-life of 21 hours. The terminal half-life of Aranesp ® was approximately 3-fold longer than that of
Epoetin alfa when administered intravenously.
Following subcutaneous administration of Aranesp ® to CRF patients (receiving or not receiving dialysis),
absorption was slow and peak concentrations occurred at 48 hours (range: 12 to 72 hours). In CRF
patients receiving dialysis, the average half-life was 46 hours (range: 12 to 89 hours), and in CRF patients
not receiving dialysis, the average half-life was 70 hours (range: 35 to 139 hours). Aranesp ® apparent
clearance was approximately 1.4 times faster on average in patients receiving dialysis compared to
patients not receiving dialysis. The bioavailability of Aranesp ® in CRF patients receiving dialysis after
subcutaneous administration was 37% (range: 30% to 50%).
Following the first subcutaneous dose of 6.75 mcg/kg (equivalent to 500 mcg for a 74-kg patient) in
patients with cancer, the mean terminal half-life was 74 hours (range: 24 to 144 hours). Peak
concentrations were observed at 90 hours (range: 71 to 123 hours) after a dose of 2.25 mcg/kg, and 71
hours (range: 28 to 120 hours) after a dose of 6.75 mcg/kg. When administered on a once every 3 week
schedule, 48-hour post-dose Aranesp ® levels after the fourth dose were similar to those after the first
dose.
Over the dose range of 0.45 to 4.5 mcg/kg Aranesp ® administered intravenously or subcutaneously on a
once weekly schedule and 4.5 to 15 mcg/kg administered subcutaneously on a once every 3 week
schedule, systemic exposure was approximately proportional to dose. No evidence of accumulation was
observed beyond an expected < 2-fold increase in blood levels when compared to the initial dose.
2

Pediatric Patients
Aranesp ® pharmacokinetics were studied in 12 pediatric CRF patients (age 3-16 years) receiving or not
receiving dialysis. Following a single intravenous or subcutaneous Aranesp ® dose, Cmax and half-life
were similar to those obtained in adult CRF patients on dialysis. Following a single subcutaneous dose,
the average bioavailability was 54% (range: 32% to 70%), which was higher than that obtained in adult
CRF patients on dialysis.
CLINICAL STUDIES
Throughout this section of the package insert, the Aranesp ® study numbers associated with the
nephrology and cancer clinical programs are designated with the letters “N” and “C”, respectively.
Chronic Renal Failure Patients
The safety and effectiveness of Aranesp ® have been assessed in a number of multicenter studies. Two
studies evaluated the safety and efficacy of Aranesp ® for the correction of anemia in adult patients with
CRF, and three studies (2 in adults and 1 in pediatric patients) assessed the ability of Aranesp ® to
maintain hemoglobin concentrations in patients with CRF who had been receiving other recombinant
erythropoietins.
De Novo Use of Aranesp ®
Once Weekly Aranesp ® Starting Dose
Page 162
In two open-label studies, Aranesp ® or Epoetin alfa was administered for the correction of anemia in CRF
patients who had not been receiving prior treatment with exogenous erythropoietin. Study N1 evaluated
CRF patients receiving dialysis; Study N2 evaluated patients not requiring dialysis. In both studies, the
starting dose of Aranesp ® was 0.45 mcg/kg administered once weekly. The starting dose of Epoetin alfa
was 50 Units/kg 3 times weekly in Study N1 and 50 Units/kg twice weekly in Study N2. When necessary,
dosage adjustments were instituted to maintain hemoglobin in the study target range of 11 to 13 g/dL.
(Note: The recommended hemoglobin target is lower than the target range of these studies. See
DOSAGE AND ADMINISTRATION for recommended clinical hemoglobin target.) The primary efficacy
endpoint was the proportion of patients who experienced at least a 1 g/dL increase in hemoglobin
concentration to a level of at least 11 g/dL by 20 weeks (Study N1) or 24 weeks (Study N2). The studies
were designed to assess the safety and effectiveness of Aranesp ® but not to support conclusions
regarding comparisons between the two products.
In Study N1, the hemoglobin target was achieved by 72% (95% CI: 62%, 81%) of the 90 patients treated
with Aranesp ® and 84% (95% CI: 66%, 95%) of the 31 patients treated with Epoetin alfa. The mean
increase in hemoglobin over the initial 4 weeks of Aranesp ® treatment was 1.1 g/dL (95% CI: 0.82 g/dL,
1.37 g/dL).
In Study N2, the primary efficacy endpoint was achieved by 93% (95% CI: 87%, 97%) of the 129 patients
treated with Aranesp ® and 92% (95% CI: 78%, 98%) of the 37 patients treated with Epoetin alfa. The
mean increase in hemoglobin from baseline through the initial 4 weeks of Aranesp ® treatment was
1.38 g/dL (95% CI: 1.21 g/dL, 1.55 g/dL).
Once Every 2 Week Aranesp ® Starting Dose
In two single arm studies (N3 and N4), Aranesp ® was administered for the correction of anemia in CRF
patients not receiving dialysis. In both studies, the starting dose of Aranesp ® was 0.75 mcg/kg
administered once every 2 weeks.
In Study N3 (study duration of 18 weeks), the hemoglobin goal (hemoglobin concentration ≥ 11 g/dL) was
achieved by 92% (95% CI: 86%, 96%) of the 128 patients treated with Aranesp ® .
In Study N4 (study duration of 24 weeks), the hemoglobin goal (hemoglobin concentration of 11-13 g/dL)
was achieved by 85% (95% CI: 77%, 93%) of the 75 patients treated with Aranesp ® .
3

Page 163
Conversion From Other Recombinant Erythropoietins
Two adult studies (N5 and N6) and one pediatric study (N7) were conducted in patients with CRF who
had been receiving other recombinant erythropoietins. The studies compared the abilities of Aranesp ®
and other erythropoietins to maintain hemoglobin concentrations within a study target range of 9 to
13 g/dL in adults and 10 to 12.5 g/dL in pediatric patients. (Note: The recommended hemoglobin target
is lower than the target range of these studies. See DOSAGE AND ADMINISTRATION for
recommended clinical hemoglobin target.) CRF patients who had been receiving stable doses of other
recombinant erythropoietins were randomized to Aranesp ® , or to continue with their prior erythropoietin at
the previous dose and schedule. For patients randomized to Aranesp ® , the initial weekly dose was
determined on the basis of the previous total weekly dose of recombinant erythropoietin.
Adult Patients
Study N5 was a double-blind study conducted in North America, in which 169 hemodialysis patients were
randomized to treatment with Aranesp ® and 338 patients continued on Epoetin alfa. Study N6 was an
open-label study conducted in Europe and Australia in which 347 patients were randomized to treatment
with Aranesp ® and 175 patients were randomized to continue on Epoetin alfa or Epoetin beta. Of the
347 patients randomized to Aranesp ® , 92% were receiving hemodialysis and 8% were receiving
peritoneal dialysis.
In Study N5, a median weekly dose of 0.53 mcg/kg Aranesp ® (25th, 75th percentiles: 0.30, 0.93 mcg/kg)
was required to maintain hemoglobin in the study target range. In Study N6, a median weekly dose of
0.41 mcg/kg Aranesp ® (25th, 75th percentiles: 0.26, 0.65 mcg/kg) was required to maintain hemoglobin in
the study target range.
Pediatric Patients
Study N7 was an open-label, randomized study, conducted in the United States in pediatric patients from
1 to 18 years of age with CRF receiving or not receiving dialysis. Patients that were stable on Epoetin alfa
were randomized to receive either darbepoetin alfa (n = 82) administered once weekly (subcutaneously or
intravenously) or to continue receiving Epoetin alfa (n = 42) at the current dose, schedule, and route of
administration. A median weekly dose of 0.41 mcg/kg Aranesp ® (25th, 75th percentiles: 0.25, 0.82
mcg/kg) was required to maintain hemoglobin in the study target range.
Cancer Patients Receiving Chemotherapy
Efficacy in patients with anemia due to concomitant chemotherapy was demonstrated based on reduction
in the requirement for RBC transfusions.
Once Weekly Dosing
The safety and effectiveness of Aranesp ® in reducing the requirement for RBC transfusions in patients
undergoing chemotherapy was assessed in a randomized, placebo-controlled, double-blind, multinational
study (C1). This study was conducted in anemic (Hgb ≤ 11 g/dL) patients with advanced, small cell or
non-small cell lung cancer, who received a platinum-containing chemotherapy regimen. Patients were
randomized to receive Aranesp ® 2.25 mcg/kg (n = 156) or placebo (n = 158) administered as a single
weekly SC injection for up to 12 weeks. The dose was escalated to 4.5 mcg/kg/week at week 6, in
subjects with an inadequate response to treatment, defined as less than 1 g/dL hemoglobin increase.
There were 67 patients in the Aranesp ® arm who had their dose increased from 2.25 to 4.5 mcg/kg/week,
at any time during the treatment period.
Efficacy was determined by a reduction in the proportion of patients who were transfused over the 12week
treatment period. A significantly lower proportion of patients in the Aranesp ® arm, 26% (95% CI:
20%, 33%) required transfusion compared to 60% (95% CI: 52%, 68%) in the placebo arm (Kaplan-Meier
estimate of proportion; p < 0.001 by Cochran–Mantel-Haenszel test). Of the 67 patients who received a
dose increase, 28% had a 2 g/dL increase in hemoglobin over baseline, generally occurring between
weeks 8 to 13. Of the 89 patients who did not receive a dose increase, 69% had a 2 g/dL increase in
4

hemoglobin over baseline, generally occurring between weeks 6 to 13. On-study deaths occurred in 14%
(22/156) of patients treated with Aranesp ® and 12% (19/158) of the placebo-treated patients.
Once Every 3 Week Dosing
The safety and effectiveness of once every 3 week Aranesp ® therapy in reducing the requirement for red
blood cell (RBC) transfusions in patients undergoing chemotherapy was assessed in a randomized,
double-blind, multinational study (C2). This study was conducted in anemic (Hgb < 11 g/dL) patients with
non-myeloid malignancies receiving multicycle chemotherapy. Patients were randomized to receive
Aranesp ® at 500 mcg once every 3 weeks (n = 353) or 2.25 mcg/kg (n = 352) administered weekly as a
subcutaneous injection for up to 15 weeks. In both groups, the dose was reduced by 40% of the previous
dose (e.g., for first dose reduction, to 300 mcg in the once every 3 week group and 1.35 mcg/kg in the
once weekly group) if hemoglobin increased by more than 1 g/dL in a 14-day period. Study drug was
withheld if hemoglobin exceeded 13 g/dL. In the once every 3 week group, 254 patients (72%) required
dose reductions (median time to first reduction at 6 weeks). In the once weekly group, 263 patients (75%)
required dose reductions (median time to first reduction at 5 weeks).
Efficacy was determined by a comparison of the Kaplan-Meier estimates of the proportion of patients who
received at least one RBC transfusion between day 29 and the end of treatment. Three hundred thirtyfive
patients in the once every 3 week group and 337 patients in the once weekly group remained on
study through or beyond day 29 and were evaluated for efficacy. Twenty-seven percent (95% CI: 22%,
32%) of patients in the once every 3 week group and 34% (95% CI: 29%, 39%) in the weekly group
required a RBC transfusion. The observed difference in the transfusion rates (once every 3 week-once
weekly) was -6.7% (95% CI: -13.8%, 0.4%).
INDICATIONS AND USAGE
Anemia With Chronic Renal Failure
Aranesp ® is indicated for the treatment of anemia associated with chronic renal failure, including patients
on dialysis and patients not on dialysis.
Anemia With Non-Myeloid Malignancies Due to Chemotherapy
Aranesp ® is indicated for the treatment of anemia due to the effect of concomitantly administered
chemotherapy based on studies that have shown a reduction in the need for RBC transfusions in patients
with metastatic, non-myeloid malignancies. Studies to determine whether Aranesp ® increases mortality
or decreases progression-free/recurrence-free survival are ongoing.
• Aranesp ® is not indicated for use in patients receiving hormonal agents, therapeutic biologic
products, or radiotherapy unless receiving concomitant myelosuppressive chemotherapy.
• Aranesp ® is not indicated for patients receiving myelosuppressive therapy when the anticipated
outcome is cure due to the absence of studies that adequately characterize the impact of
Aranesp ® on progression-free and overall survival (see WARNINGS: Increased Mortality and/or
Increased Risk of Tumor Progression or Recurrence).
• Aranesp ® use has not been demonstrated in controlled clinical trials to improve symptoms of
anemia, quality of life, fatigue, or patient well-being.
CONTRAINDICATIONS
Aranesp ® is contraindicated in patients with:
• uncontrolled hypertension
• known hypersensitivity to the active substance or any of the excipients
Page 164
5

WARNINGS
Increased Mortality, Serious Cardiovascular Events, Thromboembolic Events, and Stroke
Page 165
Patients with chronic renal failure experienced greater risks for death, serious cardiovascular events, and
stroke when administered erythropoiesis-stimulating agents (ESAs) to target hemoglobin levels of 13 g/dL
and above in clinical studies. Patients with chronic renal failure and an insufficient hemoglobin response
to ESA therapy may be at even greater risk for cardiovascular events and mortality than other patients.
Aranesp ® and other ESAs increased the risks for death and serious cardiovascular events in controlled
clinical trials of patients with cancer. These events included myocardial infarction, stroke, congestive
heart failure, and hemodialysis vascular access thrombosis. A rate of hemoglobin rise of > 1 g/dL over 2
weeks may contribute to these risks.
In a randomized prospective trial, 1432 anemic chronic renal failure patients who were not undergoing
dialysis were assigned to Epoetin alfa (rHuEPO) treatment targeting a maintenance hemoglobin
concentration of 13.5 g/dL or 11.3 g/dL. A major cardiovascular event (death, myocardial infarction,
stroke, or hospitalization for congestive heart failure) occurred among 125 (18%) of the 715 patients in
the higher hemoglobin group compared to 97 (14%) among the 717 patients in the lower hemoglobin
group [Hazard Ratio (HR) 1.3, 95% CI: 1.0, 1.7, p = 0.03]. 2
In a randomized, double-blind, placebo-controlled study of 4038 patients, there was an increased risk of
stroke when Aranesp ® was administered to patients with anemia, type 2 diabetes, and CRF who were not
on dialysis. Patients were randomized to Aranesp ® treatment targeted to a hemoglobin level of 13 g/dL or
to placebo. Placebo patients received Aranesp ® only if their hemoglobin levels were less than 9 g/dL. A
total of 101 patients receiving Aranesp ® experienced stroke compared to 53 patients receiving placebo
(5% vs. 2.6%; HR 1.92, 95% CI: 1.38, 2.68; p < 0.001).
Increased risk for serious cardiovascular events was also reported from a randomized, prospective trial of
1265 hemodialysis patients with clinically evident cardiac disease (ischemic heart disease or congestive
heart failure). In this trial, patients were assigned to Epoetin alfa treatment targeted to a maintenance
hemoglobin of either 14 ± 1 g/dL or 10 ± 1 g/dL. 3 Higher mortality (35% vs. 29%) was observed in the
634 patients randomized to a target hemoglobin of 14 g/dL than in the 631 patients assigned a target
hemoglobin of 10 g/dL. The reason for the increased mortality observed in this study is unknown;
however, the incidence of nonfatal myocardial infarction, vascular access thrombosis, and other
thrombotic events was also higher in the group randomized to a target hemoglobin of 14 g/dL.
An increased incidence of thrombotic events has also been observed in patients with cancer treated with
erythropoietic agents. In patients with cancer who received Aranesp ® , pulmonary emboli,
thrombophlebitis, and thrombosis occurred more frequently than in placebo controls (see ADVERSE
REACTIONS: Cancer Patients Receiving Chemotherapy, Table 5).
In a randomized controlled study (referred to as Cancer Study 1 - the ‘BEST’ study) with another ESA in
939 women with metastatic breast cancer receiving chemotherapy, patients received either weekly
Epoetin alfa or placebo for up to a year. This study was designed to show that survival was superior
when an ESA was administered to prevent anemia (maintain hemoglobin levels between 12 and 14 g/dL
or hematocrit between 36% and 42%). The study was terminated prematurely when interim results
demonstrated that a higher mortality at 4 months (8.7% vs. 3.4%) and a higher rate of fatal thrombotic
events (1.1% vs. 0.2%) in the first 4 months of the study were observed among patients treated with
Epoetin alfa. Based on Kaplan-Meier estimates, at the time of study termination, the 12-month survival
was lower in the Epoetin alfa group than in the placebo group (70% vs. 76%; HR 1.37, 95% CI: 1.07,
1.75, p = 0.012). 4
A systematic review of 57 randomized controlled trials (including Cancer Studies 1 and 5 - the ‘BEST’ and
‘ENHANCE’ studies) evaluating 9353 patients with cancer compared ESAs plus RBC transfusion with
6

RBC transfusion alone for prophylaxis or treatment of anemia in cancer patients with or without
concurrent antineoplastic therapy. An increased relative risk (RR) of thromboembolic events (RR 1.67,
95% CI: 1.35, 2.06; 35 trials and 6769 patients) was observed in ESA-treated patients. An overall
survival hazard ratio of 1.08 (95% CI: 0.99, 1.18; 42 trials and 8167 patients) was observed in ESAtreated
patients. 5
An increased incidence of deep vein thrombosis (DVT) in patients receiving Epoetin alfa undergoing
surgical orthopedic procedures has been observed. In a randomized controlled study (referred to as the
‘SPINE’ study), 681 adult patients, not receiving prophylactic anticoagulation and undergoing spinal
surgery, received Epoetin alfa and standard of care (SOC) treatment, or SOC treatment alone.
Preliminary analysis showed a higher incidence of DVT, determined by either Color Flow Duplex Imaging
or by clinical symptoms, in the Epoetin alfa group [16 patients (4.7%)] compared to the SOC group [7
patients (2.1%)]. In addition, 12 patients in the Epoetin alfa group and 7 patients in the SOC group had
other thrombotic vascular events.
Increased mortality was observed in a randomized placebo-controlled study of Epoetin alfa in adult
patients who were undergoing coronary artery bypass surgery (7 deaths in 126 patients randomized to
Epoetin alfa versus no deaths among 56 patients receiving placebo). Four of these deaths occurred
during the period of study drug administration and all four deaths were associated with thrombotic events.
Aranesp ® is not approved for reduction in allogeneic RBC transfusions in patients scheduled for surgical
procedures.
Increased Mortality and/or Increased Risk of Tumor Progression or Recurrence
Erythropoiesis-stimulating agents resulted in decreased locoregional control/progression-free survival
and/or overall survival (see Table 1). These findings were observed in studies of patients with advanced
head and neck cancer receiving radiation therapy (Cancer Studies 5 and 6), in patients receiving
chemotherapy for metastatic breast cancer (Cancer Study 1) or lymphoid malignancy (Cancer Study 2),
and in patients with non-small cell lung cancer or various malignancies who were not receiving
chemotherapy or radiotherapy (Cancer Studies 7 and 8).
Table 1: Randomized, Controlled Trials with Decreased Survival and/or Decreased Locoregional
Control
Study / Tumor / (n)
Chemotherapy
Cancer Study 1
Metastatic breast
cancer
(n=939)
Cancer Study 2
Lymphoid
malignancy
(n=344)
Cancer Study 3
Early breast cancer
(n=733)
Hemoglobin
Target
12-14 g/dL
13-15 g/dL (M)
13-14 g/dL (F)
12.5-13 g/dL
Achieved
Hemoglobin
(Median
Q1,Q3)
12.9 g/dL
12.2, 13.3 g/dL
11.0 g/dL
9.8, 12.1 g/dL
13.1 g/dL
12.5, 13.7 g/dL
Primary
Endpoint
12-month overall
survival
Proportion of
patients
achieving a
hemoglobin
response
Relapse-free and
overall survival
Adverse Outcome for ESA-containing
Arm
Decreased 12-month survival
Decreased overall survival
Page 166
Decreased 3 yr. relapse-free and overall
survival
7

Cancer Study 4
Cervical Cancer
(n=114)
Radiotherapy Alone
Cancer Study 5
Head and neck
cancer
(n=351)
Cancer Study 6
Head and neck
cancer
(n=522)
12-14 g/dL
≥15 g/dL (M)
≥14 g/dL (F)
No Chemotherapy or Radiotherapy
Cancer Study 7
Non-small cell lung
cancer
(n=70)
Cancer Study 8
Non-myeloid
malignancy
(n=989)
12.7 g/dL
12.1, 13.3 g/dL
Not available
14-15.5 g/dL Not available
Progression-free
and overall
survival and
locoregional
control
Locoregional
progression-free
survival
Locoregional
disease control
Decreased 3 yr. progression-free and
overall survival and locoregional control
Decreased 5-year locoregional
progression-free survival
Decreased overall survival
Decreased locoregional disease control
12-14 g/dL Not available Quality of life Decreased overall survival
12-13 g/dL
Decreased overall survival:
10.6 g/dL
9.4, 11.8 g/dL
RBC transfusions Decreased overall survival
Cancer Study 1 (the ‘BEST’ study) was previously described (see WARNINGS: Increased Mortality,
Serious Cardiovascular Events, Thromboembolic Events, and Stroke). Mortality at 4 months (8.7%
vs. 3.4%) was significantly higher in the Epoetin alfa arm. The most common investigator-attributed cause
of death within the first 4 months was disease progression; 28 of 41 deaths in the Epoetin alfa arm and 13
of 16 deaths in the placebo arm were attributed to disease progression. Investigator assessed time to
tumor progression was not different between the two groups. Survival at 12 months was significantly
lower in the Epoetin alfa arm (70% vs. 76%, HR 1.37, 95% CI: 1.07, 1.75; p = 0.012). 4
Cancer Study 2 was a Phase 3, double-blind, randomized (Aranesp ® vs. placebo) study conducted in 344
anemic patients with lymphoid malignancy receiving chemotherapy. With a median follow-up of 29
months, overall mortality rates were significantly higher among patients randomized to Aranesp ® as
compared to placebo (HR 1.36, 95% CI: 1.02, 1.82).
Cancer Study 7 was a Phase 3, multicenter, randomized (Epoetin alfa vs. placebo), double-blind study, in
which patients with advanced non-small cell lung cancer receiving only palliative radiotherapy or no active
therapy were treated with Epoetin alfa to achieve and maintain hemoglobin levels between 12 and 14
g/dL. Following an interim analysis of 70 of 300 patients planned, a significant difference in survival in
favor of the patients on the placebo arm of the trial was observed (median survival 63 vs. 129 days; HR
1.84; p = 0.04).
Cancer Study 8 was a Phase 3, double-blind, randomized (Aranesp ® vs. placebo), 16-week study in 989
anemic patients with active malignant disease, neither receiving nor planning to receive chemotherapy or
radiation therapy. There was no evidence of a statistically significant reduction in proportion of patients
receiving RBC transfusions. The median survival was shorter in the Aranesp ® treatment group (8
months) compared with the placebo group (10.8 months); HR 1.30, 95% CI: 1.07, 1.57.
Decreased progression-free survival and overall survival:
Page 167
8

Cancer Study 3 (the ‘PREPARE’ study) was a randomized controlled study in which Aranesp ® was
administered to prevent anemia conducted in 733 women receiving neo-adjuvant breast cancer
treatment. After a median follow-up of approximately 3 years the survival rate (86% vs. 90%, HR 1.42,
95% CI: 0.93, 2.18) and relapse-free survival rate were lower (72% vs. 78%, HR 1.33, 95% CI: 0.99,
1.79) in the Aranesp ® -treated arm compared to the control arm.
Cancer Study 4 (protocol GOG 191) was a randomized controlled study that enrolled 114 of a planned
460 cervical cancer patients receiving chemotherapy and radiotherapy. Patients were randomized to
receive Epoetin alfa to maintain hemoglobin between 12 and 14 g/dL or to transfusion support as needed.
The study was terminated prematurely due to an increase in thromboembolic events in Epoetin alfatreated
patients compared to control (19% vs. 9%). Both local recurrence (21% vs. 20%) and distant
recurrence (12% vs. 7%) were more frequent in Epoetin alfa-treated patients compared to control.
Progression-free survival at 3 years was lower in the Epoetin alfa-treated group compared to control (59%
vs. 62%, HR 1.06, 95% CI: 0.58, 1.91). Overall survival at 3 years was lower in the Epoetin alfa-treated
group compared to control (61% vs. 71%, HR 1.28, 95% CI: 0.68, 2.42).
Cancer Study 5 (the ‘ENHANCE’ study) was a randomized controlled study in 351 head and neck cancer
patients where Epoetin beta or placebo was administered to achieve target hemoglobins of 14 and 15
g/dL for women and men, respectively. Locoregional progression-free survival was significantly shorter in
patients receiving Epoetin beta (HR 1.62, 95% CI: 1.22, 2.14, p = 0.0008) with a median of 406 days
Epoetin beta vs. 745 days placebo. Overall survival was significantly shorter in patients receiving Epoetin
beta (HR 1.39, 95% CI: 1.05, 1.84; p = 0.02).
Decreased locoregional control:
Cancer Study 6 (DAHANCA 10) was conducted in 522 patients with primary squamous cell carcinoma of
the head and neck receiving radiation therapy randomized to Aranesp ® with radiotherapy or radiotherapy
alone. An interim analysis on 484 patients demonstrated that locoregional control at 5 years was
significantly shorter in patients receiving Aranesp ® (RR 1.44, 95% CI: 1.06, 1.96; p = 0.02). Overall
survival was shorter in patients receiving Aranesp ® (RR 1.28, 95% CI: 0.98, 1.68; p = 0.08).
ESA APPRISE Oncology Program
Page 168
Prescribers and hospitals must enroll in and comply with the ESA APPRISE Oncology Program to
prescribe and/or dispense Aranesp ® to patients with cancer. To enroll, visit www.esa-apprise.com or call
1-866-284-8089 for further assistance. Additionally, prescribers and patients must provide written
acknowledgment of a discussion of the risks associated with Aranesp ® .
Hypertension
Patients with uncontrolled hypertension should not be treated with Aranesp ® ; blood pressure should be
controlled adequately before initiation of therapy. Blood pressure may rise during treatment of anemia
with Aranesp ® or Epoetin alfa. In Aranesp ® clinical trials, approximately 40% of patients with CRF
required initiation or intensification of antihypertensive therapy during the early phase of treatment when
the hemoglobin was increasing. Hypertensive encephalopathy and seizures have been observed in
patients with CRF treated with Aranesp ® or Epoetin alfa.
Special care should be taken to closely monitor and control blood pressure in patients treated with
Aranesp ® . During Aranesp ® therapy, patients should be advised of the importance of compliance with
antihypertensive therapy and dietary restrictions. If blood pressure is difficult to control by pharmacologic
or dietary measures, the dose of Aranesp ® should be reduced or withheld (see DOSAGE AND
ADMINISTRATION). A clinically significant decrease in hemoglobin may not be observed for several
weeks.
9

Page 169
Seizures
Seizures have occurred in patients with CRF participating in clinical trials of Aranesp ® and Epoetin alfa.
During the first several months of therapy, blood pressure and the presence of premonitory neurologic
symptoms should be monitored closely. While the relationship between seizures and the rate of rise of
hemoglobin is uncertain, it is recommended that the dose of Aranesp ® be decreased if the hemoglobin
increase exceeds 1 g/dL in any 2-week period.
Pure Red Cell Aplasia
Cases of pure red cell aplasia (PRCA) and of severe anemia, with or without other cytopenias, associated
with neutralizing antibodies to erythropoietin have been reported in patients treated with Aranesp ® . This
has been reported predominantly in patients with CRF receiving ESAs by subcutaneous administration.
PRCA has also been reported in patients receiving ESAs while undergoing treatment for hepatitis C with
interferon and ribavirin. Any patient who develops a sudden loss of response to Aranesp ® , accompanied
by severe anemia and low reticulocyte count, should be evaluated for the etiology of loss of effect,
including the presence of neutralizing antibodies to erythropoietin (see PRECAUTIONS: Lack or Loss of
Response to Aranesp ® ). If anti-erythropoietin antibody-associated anemia is suspected, withhold
Aranesp ® and other ESAs. Contact Amgen (1-800-77AMGEN) to perform assays for binding and
neutralizing antibodies. Aranesp ® should be permanently discontinued in patients with antibody-mediated
anemia. Patients should not be switched to other ESAs as antibodies may cross-react (see ADVERSE
REACTIONS: Immunogenicity).
Albumin (Human)
Aranesp ® is supplied in two formulations with different excipients, one containing polysorbate 80 and
another containing albumin (human), a derivative of human blood (see DESCRIPTION). Based on
effective donor screening and product manufacturing processes, Aranesp ® formulated with albumin
carries an extremely remote risk for transmission of viral diseases. A theoretical risk for transmission of
Creutzfeldt-Jakob disease (CJD) also is considered extremely remote. No cases of transmission of viral
diseases or CJD have ever been identified for albumin.
PRECAUTIONS
General
The safety and efficacy of Aranesp ® therapy have not been established in patients with underlying
hematologic diseases (e.g., hemolytic anemia, sickle cell anemia, thalassemia, porphyria).
The needle cover of the prefilled syringe contains dry natural rubber (a derivative of latex), which may
cause allergic reactions in individuals sensitive to latex.
Lack or Loss of Response to Aranesp ®
A lack of response or failure to maintain a hemoglobin response with Aranesp ® doses within the
recommended dosing range should prompt a search for causative factors. Deficiencies of folic acid, iron,
or vitamin B12 should be excluded or corrected. Depending on the clinical setting, intercurrent infections,
inflammatory or malignant processes, osteofibrosis cystica, occult blood loss, hemolysis, severe
aluminum toxicity, and bone marrow fibrosis may compromise an erythropoietic response. In the absence
of another etiology, the patient should be evaluated for evidence of PRCA and sera should be tested for
the presence of antibodies to erythropoietin (see WARNINGS: Pure Red Cell Aplasia). See DOSAGE
AND ADMINISTRATION: Chronic Renal Failure Patients, Dose Adjustment for management of
patients with an insufficient hemoglobin response to Aranesp ® therapy.
Hematology
Sufficient time should be allowed to determine a patient’s responsiveness to a dosage of Aranesp ® before
adjusting the dose. Because of the time required for erythropoiesis and the RBC half-life, an interval of 2
to 6 weeks may occur between the time of a dose adjustment (initiation, increase, decrease, or
discontinuation) and a significant change in hemoglobin.
10

Page 170
In order to prevent the hemoglobin from exceeding the recommended target range (10 to 12 g/dL) or
rising too rapidly (greater than 1 g/dL in 2 weeks), the guidelines for dose and frequency of dose
adjustments should be followed (see WARNINGS and DOSAGE AND ADMINISTRATION).
Allergic Reactions
There have been rare reports of potentially serious allergic reactions, including skin rash and urticaria,
associated with Aranesp ® . Symptoms have recurred with rechallenge, suggesting a causal relationship
exists in some instances. If a serious allergic or anaphylactic reaction occurs, Aranesp ® should be
immediately and permanently discontinued and appropriate therapy should be administered.
Patients with CRF Not Requiring Dialysis
Patients with CRF not yet requiring dialysis may require lower maintenance doses of Aranesp ® than
patients receiving dialysis. Though CRF patients not on dialysis generally receive less frequent
monitoring of blood pressure and laboratory parameters than dialysis patients, CRF patients not on
dialysis may be more responsive to the effects of Aranesp ® , and require judicious monitoring of blood
pressure and hemoglobin. Renal function and fluid and electrolyte balance should also be closely
monitored.
Patients Transitioning to Dialysis
During the transition period onto dialysis, hemoglobin and blood pressure should be monitored carefully
and patients may need to have their maintenance doses adjusted to maintain hemoglobin levels within
the range of 10 to 12 g/dL (see DOSAGE AND ADMINISTRATION: Maintenance Dose).
Dialysis Management
Therapy with Aranesp ® results in an increase in RBCs and a decrease in plasma volume, which could
reduce dialysis efficiency; patients who are marginally dialyzed may require adjustments in their dialysis
prescription.
Laboratory Tests
After initiation of Aranesp ® therapy, the hemoglobin should be determined weekly until it has stabilized
and the maintenance dose has been established (see DOSAGE AND ADMINISTRATION). After a dose
adjustment, the hemoglobin should be determined weekly for at least 4 weeks, until it has been
determined that the hemoglobin has stabilized in response to the dose change. The hemoglobin should
then be monitored at regular intervals.
In order to ensure effective erythropoiesis, iron status should be evaluated for all patients before and
during treatment, as the majority of patients will eventually require supplemental iron therapy.
Supplemental iron therapy is recommended for all patients whose serum ferritin is below 100 mcg/L or
whose serum transferrin saturation is below 20%.
Information for Patients
Patients should be informed of the increased risks of mortality, serious cardiovascular events,
thromboembolic events, and increased risk of tumor progression or recurrence (see WARNINGS).
Patients should be informed of the possible side effects of Aranesp ® and be instructed to report them to
the prescribing physician. Patients should be informed of the signs and symptoms of allergic drug
reactions and be advised of appropriate actions. Patients should be counseled on the importance of
compliance with their Aranesp ® treatment, dietary and dialysis prescriptions, and the importance of
judicious monitoring of blood pressure and hemoglobin concentration should be stressed.
In those rare cases where it is determined that a patient can safely and effectively administer Aranesp ® at
home, appropriate instruction on the proper use of Aranesp ® should be provided for patients and their
caregivers. Patients should be instructed to read the Aranesp ® Medication Guide and Patient Instructions
for Use and should be informed that the Medication Guide is not a disclosure of all possible side effects.
Patients and caregivers should also be cautioned against the reuse of needles, syringes, prefilled
SureClick autoinjectors, or drug product, and be thoroughly instructed in their proper disposal. A
puncture-resistant container for the disposal of used syringes, autoinjectors, and needles should be made
available to the patient. Patients should be informed that the needle cover on the prefilled syringe
11

Page 171
contains dry natural rubber (a derivative of latex), which should not be handled by persons sensitive to
latex.
Drug Interactions
No formal drug interaction studies of Aranesp ® have been performed.
Carcinogenesis, Mutagenesis, and Impairment of Fertility
Carcinogenicity: The carcinogenic potential of Aranesp ® has not been evaluated in long-term animal
studies. Aranesp ® did not alter the proliferative response of non-hematological cells in vitro or in vivo. In
toxicity studies of approximately 6 months duration in rats and dogs, no tumorigenic or unexpected
mitogenic responses were observed in any tissue type. Using a panel of human tissues, the in vitro
tissue binding profile of Aranesp ® was identical to Epoetin alfa. Neither molecule bound to human tissues
other than those expressing the erythropoietin receptor.
Mutagenicity: Aranesp ® was negative in the in vitro bacterial and CHO cell assays to detect
mutagenicity and in the in vivo mouse micronucleus assay to detect clastogenicity.
Impairment of Fertility: When administered intravenously to male and female rats prior to and during
mating, reproductive performance, fertility, and sperm assessment parameters were not affected at any
doses evaluated (up to 10 mcg/kg/dose, administered 3 times weekly). An increase in post implantation
fetal loss was seen at doses equal to or greater than 0.5 mcg/kg/dose, administered 3 times weekly.
Pregnancy Category C
When Aranesp ® was administered intravenously to rats and rabbits during gestation, no evidence of a
direct embryotoxic, fetotoxic, or teratogenic outcome was observed at doses up to 20 mcg/kg/day. The
only adverse effect observed was a slight reduction in fetal weight, which occurred at doses causing
exaggerated pharmacological effects in the dams (1 mcg/kg/day and higher). No deleterious effects on
uterine implantation were seen in either species. No significant placental transfer of Aranesp ® was
observed in rats. An increase in post implantation fetal loss was observed in studies assessing fertility
(see PRECAUTIONS: Carcinogenesis, Mutagenesis, and Impairment of Fertility: Impairment of
Fertility).
Intravenous injection of Aranesp ® to female rats every other day from day 6 of gestation through day 23 of
lactation at doses of 2.5 mcg/kg/dose and higher resulted in offspring (F1 generation) with decreased
body weights, which correlated with a low incidence of deaths, as well as delayed eye opening and
delayed preputial separation. No adverse effects were seen in the F2 offspring.
There are no adequate and well-controlled studies in pregnant women. Aranesp ® should be used during
pregnancy only if the potential benefit justifies the potential risk to the fetus.
Nursing Mothers
It is not known whether Aranesp ® is excreted in human milk. Because many drugs are excreted in human
milk, caution should be exercised when Aranesp ® is administered to a nursing woman.
Pediatric Use
Pediatric CRF Patients
A study of the conversion from Epoetin alfa to Aranesp ® among pediatric CRF patients over 1 year of age
showed similar safety and efficacy to the findings from adult conversion studies (see CLINICAL
PHARMACOLOGY and CLINICAL STUDIES). Safety and efficacy in the initial treatment of anemic
pediatric CRF patients or in the conversion from another erythropoietin to Aranesp ® in pediatric CRF
patients less than 1 year of age have not been established.
Pediatric Cancer Patients
The safety and efficacy of Aranesp ® in pediatric cancer patients have not been established.
12

Page 172
Geriatric Use
Of the 1801 CRF patients in clinical studies of Aranesp ® , 44% were age 65 and over, while 17% were age
75 and over. Of the 873 cancer patients in clinical studies receiving Aranesp ® and concomitant
chemotherapy, 45% were age 65 and over, while 14% were age 75 and over. No overall differences in
safety or efficacy were observed between older and younger patients.
ADVERSE REACTIONS
General
Because clinical trials are conducted under widely varying conditions, adverse reaction rates observed in
the clinical trials of Aranesp ® cannot be directly compared to rates in the clinical trials of other drugs and
may not reflect the rates observed in practice.
Immunogenicity
As with all therapeutic proteins, there is a potential for immunogenicity. Neutralizing antibodies to
erythropoietin, in association with PRCA or severe anemia (with or without other cytopenias), have been
reported in patients receiving Aranesp ® (see WARNINGS: Pure Red Cell Aplasia) during post-marketing
experience.
In clinical studies, the percentage of patients with antibodies to Aranesp ® was examined using the
BIAcore assay. Sera from 1501 CRF patients and 1159 cancer patients were tested. At baseline, prior to
Aranesp ® treatment, binding antibodies were detected in 59 (4%) of CRF patients and 36 (3%) of cancer
patients. While receiving Aranesp ® therapy (range 22-177 weeks), a follow-up sample was taken. One
additional CRF patient and eight additional cancer patients developed antibodies capable of binding
Aranesp ® . None of the patients had antibodies capable of neutralizing the activity of Aranesp ® or
endogenous erythropoietin at baseline or at end of study. No clinical sequelae consistent with PRCA
were associated with the presence of these antibodies.
The incidence of antibody formation is highly dependent on the sensitivity and specificity of the assay.
Additionally, the observed incidence of antibody (including neutralizing antibody) positivity in an assay
may be influenced by several factors including assay methodology, sample handling, timing of sample
collection, concomitant medications, and underlying disease. For these reasons, comparison of the
incidence of antibodies across products within this class (erythropoietic proteins) may be misleading.
Chronic Renal Failure Patients
Adult Patients
In all studies, the most frequently reported serious adverse events with Aranesp ® were infection,
congestive heart failure, angina pectoris/cardiac chest pain, thrombosis vascular access, and cardiac
arrhythmia/cardiac arrest. The most frequently reported adverse events resulting in clinical intervention
(e.g., discontinuation of Aranesp ® , adjustment in dosage, or the need for concomitant medication to treat
an adverse reaction symptom) were infection, hypertension, hypotension, and muscle spasm. See
WARNINGS: Increased Mortality, Serious Cardiovascular Events, Thromboembolic Events, and
Stroke and Hypertension.
The data described below reflect exposure to Aranesp ® in 1801 CRF patients, including 675 exposed for
at least 6 months, of whom 185 were exposed for greater than 1 year. Aranesp ® was evaluated in activecontrolled
(n = 823) and uncontrolled studies (n = 978). These data include a pooled analysis of CRF
patients not on dialysis and dialysis patients who were studied for the correction of anemia and
maintenance of hemoglobin.
The population encompassed an age range from 18 to 94 years. Fifty-five percent of the patients were
male. The percentages of Caucasian, Black, Asian, and Hispanic patients were 80%, 13%, 3%, and 2%,
respectively. The median weekly dose of Aranesp ® for patients who received either once weekly or once
every 2 week administration was 0.44 mcg/kg (25th, 75th percentiles: 0.30, 0.64 mcg/kg).
Some of the adverse events reported are typically associated with CRF, or recognized complications of
dialysis, and may not necessarily be attributable to Aranesp ® therapy. No important differences in
13

adverse event rates between treatment groups were observed in controlled studies in which patients
received Aranesp ® or other recombinant erythropoietins.
The data in Table 2 reflect those adverse events occurring in at least 5% of patients treated with
Aranesp ® .
Table 2. Adverse Events Occurring in ≥ 5% of CRF Patients
(Continued)
Event
Patients Treated with
Aranesp ® APPLICATION SITE
(n = 1801)
Injection Site Pain
BODY AS A WHOLE
6%
Peripheral Edema 10%
Fatigue 9%
Fever 7%
Death 6%
Chest Pain, Unspecified 7%
Fluid Overload 6%
Access Infection 6%
Influenza-like Symptoms 6%
Access Hemorrhage 7%
Asthenia
CARDIOVASCULAR
5%
Hypertension 20%
Hypotension 20%
Cardiac Arrhythmias/Cardiac Arrest 8%
Angina Pectoris/Cardiac Chest Pain 8%
Thrombosis Vascular Access 6%
Congestive Heart Failure
CNS/PNS
5%
Headache 15%
Dizziness
GASTROINTESTINAL
7%
Diarrhea 14%
Vomiting 14%
Nausea 11%
Abdominal Pain 10%
Constipation
MUSCULO-SKELETAL
5%
Muscle Spasm 17%
Arthralgia 9%
Limb Pain 8%
Back Pain 7%
Page 173
14

Table 2. Adverse Events Occurring in ≥ 5% of CRF Patients (Continued)
Event
Patients Treated with Aranesp ®
RESISTANCE MECHANISM
(n = 1801)
Infection a 24%
RESPIRATORY
Upper Respiratory Infection 15%
Dyspnea 10%
Cough 9%
Bronchitis 5%
SKIN AND APPENDAGES
Pruritus 6%
a Infection includes sepsis, bacteremia, pneumonia, peritonitis, and abscess.
The incidence rates for other clinically significant events are shown in Table 3.
Table 3. Percent Incidence of Other Clinically Significant Events in CRF Patients
Event
Patients Treated with Aranesp ®
(n = 1801)
Acute Myocardial Infarction 2%
Stroke 2%
Seizure 1%
Transient Ischemic Attack ≤1%
Page 174
Pediatric Patients
In Study N7, Aranesp ® was administered to 81 pediatric CRF patients who had stable hemoglobin
concentrations while previously receiving Epoetin alfa (see CLINICAL STUDIES). In this study, the most
frequently reported serious adverse events with Aranesp ® were catheter sepsis, fever, catheter related
infection, chronic renal failure, and vascular access complication. The most commonly reported adverse
events were fever, headache, nasopharyngitis, hypertension, hypotension, injection site pain, cough,
peritonitis, and vomiting. Aranesp ® administration was discontinued because of injection site pain in two
patients and moderate hypertension in a third patient.
Studies have not evaluated the effects of Aranesp ® when administered to pediatric patients as the initial
treatment for the anemia associated with CRF.
Thrombotic Events
Vascular access thrombosis in hemodialysis patients occurred in clinical trials at an annualized rate of
0.22 events per patient year of Aranesp ® therapy. Rates of thrombotic events (e.g., vascular access
thrombosis, venous thrombosis, and pulmonary emboli) with Aranesp ® therapy were similar to those
observed with other recombinant erythropoietins in these trials; the median duration of exposure was 12
weeks.
Cancer Patients Receiving Chemotherapy
The incidence data described below reflect the exposure to Aranesp ® in 873 cancer patients including
patients exposed to Aranesp ® once weekly (547, 63%), once every 2 weeks (128, 16%), and once every
3 weeks (198, 23%). Aranesp ® was evaluated in seven studies that were active-controlled and/or
placebo-controlled studies of up to 6 months duration. The Aranesp ® -treated patient demographics were
as follows: median age of 63 years (range of 20 to 91 years); 40% male; 88% Caucasian, 5% Hispanic,
15

4% Black, and 3% Asian. Over 90% of patients had locally advanced or metastatic cancer, with the
remainder having early stage disease. Patients with solid tumors (e.g., lung, breast, colon, ovarian
cancers) and lymphoproliferative malignancies (e.g., lymphoma, multiple myeloma) were enrolled in the
clinical studies. All of the 873 Aranesp ® -treated subjects also received concomitant cyclic chemotherapy.
The most frequently reported serious adverse events included death (10%), fever (4%), pneumonia (3%),
dehydration (3%), vomiting (2%), and dyspnea (2%). The most commonly reported adverse events were
fatigue, edema, nausea, vomiting, diarrhea, fever, and dyspnea (see Table 4). Except for those events
listed in Tables 4 and 5, the incidence of adverse events in clinical studies occurred at a similar rate
compared with patients who received placebo and were generally consistent with the underlying disease
and its treatment with chemotherapy. The most frequently reported reasons for discontinuation of
Aranesp ® were progressive disease, death, discontinuation of the chemotherapy, asthenia, dyspnea,
pneumonia, and gastrointestinal hemorrhage. No important differences in adverse event rates between
treatment groups were observed in controlled studies in which patients received Aranesp ® or other
recombinant erythropoietins.
Table 4. Adverse Events Occurring in ≥ 5% of Patients Receiving Chemotherapy
Event
Aranesp ®
(n = 873)
Placebo
(n = 221)
BODY AS A WHOLE
Fatigue 33% 30%
Edema 21% 10%
Fever 19% 16%
CNS/PNS
Dizziness 14% 8%
Headache 12% 9%
GASTROINTESTINAL
Diarrhea 22% 12%
Constipation 18% 17%
METABOLIC/NUTRITION
Dehydration 5% 3%
MUSCULO-SKELETAL
Arthralgia 13% 6%
Myalgia 8% 5%
SKIN AND APPENDAGES
Rash 7% 3%
Page 175
16

Table 5. Incidence of Other Clinically Significant Adverse Events in Patients Receiving
Chemotherapy
Event
All Aranesp ®
(n = 873)
Placebo
(n = 221)
Hypertension 3.7% 3.2%
Seizures/Convulsions a 0.6% 0.5%
Thrombotic Events 6.2% 4.1%
Pulmonary Embolism 1.3% 0.0%
Thrombosis b 5.6% 4.1%
a
Seizures/Convulsions include the preferred terms: Convulsions, Convulsions
Grand Mal, and Convulsions Local.
b
Thrombosis includes: Thrombophlebitis, Thrombophlebitis Deep,
Thrombosis Venous, Thrombosis Venous Deep, Thromboembolism, and
Thrombosis.
Page 176
In a randomized controlled trial of Aranesp ® 500 mcg once every 3 weeks (n = 353) and Aranesp ® 2.25
mcg/kg once weekly (n = 352), the incidences of all adverse events and of serious adverse events were
similar between the two groups.
Thrombotic and Cardiovascular Events
Overall, the incidence of thrombotic events was 6.2% for Aranesp ® and 4.1% for placebo. However, the
following events were reported more frequently in Aranesp ® -treated patients than in placebo controls:
pulmonary embolism, thromboembolism, thrombosis, and thrombophlebitis (deep and/or superficial). In
addition, edema of any type was more frequently reported in Aranesp ® -treated patients (21%) than in
patients who received placebo (10%).
OVERDOSAGE
The expected manifestations of Aranesp ® overdosage include signs and symptoms associated with an
excessive and/or rapid increase in hemoglobin concentration, including any of the cardiovascular events
described in WARNINGS and listed in ADVERSE REACTIONS. Patients receiving an overdosage of
Aranesp ® should be monitored closely for cardiovascular events and hematologic abnormalities.
Polycythemia should be managed acutely with phlebotomy, as clinically indicated. Following resolution of
the effects due to Aranesp ® overdosage, reintroduction of Aranesp ® therapy should be accompanied by
close monitoring for evidence of rapid increases in hemoglobin concentration (> 1 g/dL in any 2-week
period). In patients with an excessive hematopoietic response, reduce the Aranesp ® dose in accordance
with the recommendations described in DOSAGE AND ADMINISTRATION.
DOSAGE AND ADMINISTRATION
IMPORTANT: See BOXED WARNINGS and WARNINGS: Increased Mortality, Serious
Cardiovascular Events, Thromboembolic Events, and Stroke.
Aranesp ® is supplied in vials or in prefilled syringes with UltraSafe ® Needle Guards * . Following
administration of Aranesp ® from the prefilled syringe, the UltraSafe ® Needle Guard should be activated to
prevent accidental needlesticks.
Aranesp ® is also supplied in prefilled SureClick autoinjectors containing the same dosage strengths as
the prefilled syringes. Because the autoinjectors are designed to deliver the full content, autoinjectors
should only be used for patients who need the full dose. If the required dose is not available in an
autoinjector, prefilled syringes, or vials should be used to administer the required dose. Autoinjectors are
for subcutaneous administration only.
17

Page 177
Chronic Renal Failure Patients
Aranesp ® may be administered either intravenously or subcutaneously as a single weekly injection. In
patients on hemodialysis, the intravenous route is recommended. The dose should be started and
slowly adjusted as described below based on hemoglobin levels. If a patient fails to respond or maintain
a response, this should be evaluated (see WARNINGS: Pure Red Cell Aplasia, PRECAUTIONS: Lack
or Loss of Response to Aranesp ® and PRECAUTIONS: Laboratory Tests). When Aranesp ® therapy
is initiated or adjusted, the hemoglobin should be followed weekly until stabilized and monitored at least
monthly thereafter. During therapy, hematological parameters should be monitored regularly. Doses
must be individualized to ensure that hemoglobin is maintained at an appropriate level for each patient.
For patients who respond to Aranesp ® with a rapid increase in hemoglobin (e.g., more than 1 g/dL in any
2-week period), the dose of Aranesp ® should be reduced.
Individualize dosing to achieve and maintain hemoglobin levels within the range of 10 to 12 g/dL.
Starting Dose
Correction of Anemia
The initial dose by subcutaneous or intravenous administration is 0.45 mcg/kg body weight, as a
single injection once weekly. Alternatively, in patients not receiving dialysis, an initial dose of 0.75 mcg/kg
may be administered subcutaneously as a single injection once every 2 weeks. If hemoglobin excursions
outside the recommended range occur, the Aranesp ® dose should be adjusted as described below.
The use of Aranesp ® in pediatric CRF patients as the initial treatment to correct anemia has not been
studied.
Maintenance Dose
The dose should be individualized to maintain hemoglobin levels within the range of 10 to 12 g/dL (see
Dose Adjustment). If hemoglobin excursions outside the recommended range occur, the Aranesp ® dose
should be adjusted as described below. For many patients, the appropriate maintenance dose will be
lower than the starting dose. CRF patients not on dialysis, in particular, may require lower maintenance
doses. In the maintenance phase, Aranesp ® may continue to be administered as a single injection once
weekly or once every 2 weeks.
Dose Adjustment
The dose should be adjusted for each patient to achieve and maintain hemoglobin levels within the range
of 10 to 12 g/dL. If hemoglobin excursions outside the recommended range occur, the Aranesp ® dose
should be adjusted as described below. Increases in dose should not be made more frequently than
once a month.
If the hemoglobin is increasing and approaching 12 g/dL, the dose should be reduced by approximately
25%. If the hemoglobin continues to increase, doses should be temporarily withheld until the hemoglobin
begins to decrease, at which point therapy should be reinitiated at a dose approximately 25% below the
previous dose. If the hemoglobin increases by more than 1 g/dL in a 2-week period, the dose should be
decreased by approximately 25%.
If the increase in hemoglobin is less than 1 g/dL over 4 weeks and iron stores are adequate (see
PRECAUTIONS: Laboratory Tests), the dose of Aranesp ® may be increased by approximately 25% of
the previous dose. Further increases may be made at 4-week intervals until the specified hemoglobin is
obtained.
For patients whose hemoglobin does not attain a level within the range of 10 to 12 g/dL despite the use of
appropriate Aranesp ® dose titrations over a 12-week period:
• do not administer higher Aranesp ® doses and use the lowest dose that will maintain a
hemoglobin level sufficient to avoid the need for recurrent RBC transfusions,
• evaluate and treat for other causes of anemia (see PRECAUTIONS: Lack or Loss of Response
to Aranesp ® ), and
18

• thereafter, hemoglobin should continue to be monitored and if responsiveness improves,
Aranesp ® dose adjustments should be made as described above; discontinue Aranesp ® if
responsiveness does not improve and the patient needs recurrent RBC transfusions.
Conversion From Epoetin alfa to Aranesp ®
The starting weekly dose of Aranesp ® for adults and pediatric patients should be estimated on the basis
of the weekly Epoetin alfa dose at the time of substitution (see Table 6). For pediatric patients receiving
a weekly Epoetin alfa dose of < 1,500 units/week, the available data are insufficient to determine an
Aranesp ® conversion dose. Because of variability, doses should be titrated to achieve and maintain
hemoglobin levels within the range of 10 to 12 g/dL. Due to the longer serum half-life, Aranesp ® should
be administered less frequently than Epoetin alfa. Aranesp ® should be administered once a week if a
patient was receiving Epoetin alfa 2 to 3 times weekly. Aranesp ® should be administered once every
2 weeks if a patient was receiving Epoetin alfa once per week. The route of administration (intravenous
or subcutaneous) should be maintained.
Table 6. Estimated Aranesp ® Starting Doses (mcg/week) for Patients
Based on Previous Epoetin alfa Dose (Units/week)
Weekly Aranesp ® Dose (mcg/week)
Previous Weekly Epoetin alfa
Dose (Units/week)
Adult Pediatric
< 1,500 6.25 See text*
1,500 to 2,499 6.25 6.25
2,500 to 4,999 12.5 10
5,000 to 10,999 25 20
11,000 to 17,999 40 40
18,000 to 33,999 60 60
34,000 to 89,999 100 100
≥ 90,000 200 200
*For pediatric patients receiving a weekly Epoetin alfa dose of < 1,500 units/week, the available data are insufficient
to determine an Aranesp ® conversion dose.
Cancer Patients Receiving Chemotherapy
Only prescribers enrolled in the ESA APPRISE Oncology Program may prescribe and/or dispense
Aranesp ® (see WARNINGS: ESA APPRISE Oncology Program).
For pediatric patients, see PRECAUTIONS: Pediatric Use.
Page 178
The recommended starting dose for Aranesp ® administered weekly is 2.25 mcg/kg as a subcutaneous
injection.
The recommended starting dose for Aranesp ® administered once every 3 weeks is 500 mcg as a
subcutaneous injection.
Therapy should not be initiated at hemoglobin levels ≥ 10 g/dL. For both dosing schedules, the dose
should be adjusted for each patient to maintain the lowest hemoglobin level sufficient to avoid RBC
19

transfusion. If the rate of hemoglobin increase is more than 1 g/dL per 2-week period or when the
hemoglobin reaches a level needed to avoid transfusion, the dose should be reduced by 40% of the
previous dose. If the hemoglobin exceeds a level needed to avoid transfusion, Aranesp ® should be
temporarily withheld until the hemoglobin approaches a level where transfusions may be required. At this
point, therapy should be reinitiated at a dose 40% below the previous dose.
For patients receiving weekly administration, if there is less than a 1 g/dL increase in hemoglobin after 6
weeks of therapy, the dose of Aranesp ® should be increased up to 4.5 mcg/kg.
Discontinue Aranesp ® if after 8 weeks of therapy there is no response as measured by hemoglobin levels
or if transfusions are still required.
Discontinue Aranesp ® following the completion of a chemotherapy course (see BOXED WARNINGS:
Cancer).
Preparation and Administration of Aranesp ®
Page 179
Do not shake Aranesp ® or leave vials, syringes, or prefilled SureClick autoinjectors exposed to light.
After removing the vials, prefilled syringes, or autoinjectors from the refrigerator, protect from room light
until administration. Vigorous shaking or exposure to light may denature Aranesp ® , causing it to become
biologically inactive. Always store vials, prefilled syringes, or autoinjectors of Aranesp ® in their carton
until use.
Parenteral drug products should be inspected visually for particulate matter and discoloration prior to
administration. Do not use any vials, prefilled syringes, or autoinjectors exhibiting particulate matter or
discoloration.
Do not dilute Aranesp ® .
Do not administer Aranesp ® in conjunction with other drug solutions.
Aranesp ® contains no preservatives. Discard any unused portion. Do not pool unused portions from
the vials or prefilled syringes. Do not use the vial, prefilled syringe, or autoinjector more than one
time.
Following administration of Aranesp ® from the prefilled syringe, activate the UltraSafe ® Needle Guard.
Place your hands behind the needle, grasp the guard with one hand, and slide the guard forward until the
needle is completely covered and the guard clicks into place. NOTE: If an audible click is not heard, the
needle guard may not be completely activated.
The prefilled SureClick autoinjector is designed to deliver the full dose. The completion of the injection
is signaled by an audible click. Removal of the autoinjector from the injection site automatically extends a
needle cover.
The autoinjectors, the syringes used with vials, and the entire prefilled syringe with activated needle
guard should be disposed of in a puncture-proof container.
See the accompanying “Patient Instructions for Use” insert for complete instructions on the preparation
and administration of Aranesp ® for patients, including injection site selection.
HOW SUPPLIED
Aranesp ® is available in single-dose vials in two solutions, an albumin solution and a polysorbate solution.
The words “Albumin Free” appear on the polysorbate container labels and the package main panels as
well as other panels as space permits. Aranesp ® single-dose prefilled syringes and prefilled SureClick
autoinjectors are available in albumin and polysorbate solutions. Both prefilled syringes and autoinjectors
are supplied with a 27-gauge, ½-inch needle.
Each prefilled syringe is equipped with an UltraSafe ® Needle Guard that is manually activated to cover
the needle during disposal. The needle cover of the prefilled syringe contains dry natural rubber (a
derivative of latex). The autoinjector has a needle cover that automatically extends as the autoinjector is
removed from the injection site after completion of the injection.
20

Aranesp ® is available in the following packages:
Single-dose Vial, Polysorbate Solution
1 Vial/Pack,
4 Packs/Case
200 mcg/1 mL
(NDC 55513-006-01)
300 mcg/1 mL
(NDC 55513-110-01)
500 mcg/1 mL
(NDC 55513-008-01)
Single-dose Vial, Albumin Solution
1 Vial/Pack,
4 Packs/Case
200 mcg/1 mL
(NDC 55513-014-01)
300 mcg/1 mL
(NDC 55513-015-01)
500 mcg/1 mL
(NDC 55513-016-01)
4 Vials/Pack, 4 Packs/Case 4 Vials/Pack, 10 Packs/Case
200 mcg/1 mL
(NDC 55513-006-04)
300 mcg/1 mL
(NDC 55513-110-04)
25 mcg/1 mL
(NDC 55513-002-04)
40 mcg/1 mL
(NDC 55513-003-04)
60 mcg/1 mL
(NDC 55513-004-04)
100 mcg/1 mL
(NDC 55513-005-04)
150 mcg/0.75 mL
(NDC 55513-053-04)
4 Vials/Pack, 4 Packs/Case 4 Vials/Pack, 10 Packs/Case
200 mcg/1 mL
(NDC 55513-014-04)
300 mcg/1 mL
(NDC 55513-015-04)
25 mcg/1 mL
(NDC 55513-010-04)
40 mcg/1 mL
(NDC 55513-011-04)
60 mcg/1 mL
(NDC 55513-012-04)
100 mcg/1 mL
(NDC 55513-013-04)
150 mcg/0.75 mL
(NDC 55513-054-04)
Page 180
21

Single-dose Prefilled Syringe (SingleJect ® ) with a 27-gauge, ½-inch needle with an UltraSafe ®
Needle Guard, Polysorbate Solution
1 Syringe/Pack,
4 Packs/Case
200 mcg/0.4 mL
(NDC 55513-028-01)
300 mcg/0.6 mL
(NDC 55513-111-01)
500 mcg/1 mL
(NDC 55513-032-01)
4 Syringes/Pack, 4 Packs/Case 4 Syringes/Pack, 10 Packs/Case
200 mcg/0.4 mL
(NDC 55513-028-04)
300 mcg/0.6 mL
(NDC 55513-111-04)
25 mcg/0.42 mL
(NDC 55513-057-04)
40 mcg/0.4 mL
(NDC 55513-021-04)
60 mcg/0.3 mL
(NDC 55513-023-04)
100 mcg/0.5 mL
(NDC 55513-025-04)
150 mcg/0.3 mL
(NDC 55513-027-04)
Single-dose Prefilled Syringe (SingleJect ® ) with a 27-gauge, ½-inch needle with an UltraSafe ®
Needle Guard, Albumin Solution
1 Syringe/Pack,
4 Packs/Case
200 mcg/0.4 mL
(NDC 55513-044-01)
300 mcg/0.6 mL
(NDC 55513-046-01)
500 mcg/1 mL
(NDC 55513-048-01)
4 Syringes/Pack, 4 Packs/Case 4 Syringes/Pack, 10 Packs/Case
200 mcg/0.4 mL
(NDC 55513-044-04)
300 mcg/0.6 mL
(NDC 55513-046-04)
25 mcg/0.42 mL
(NDC 55513-058-04)
40 mcg/0.4 mL
(NDC 55513-037-04)
60 mcg/0.3 mL
(NDC 55513-039-04)
100 mcg/0.5 mL
(NDC 55513-041-04)
150 mcg/0.3 mL
(NDC 55513-043-04)
Page 181
22

Single-dose Prefilled SureClick Autoinjector with a 27-gauge, ½-inch needle, Polysorbate Solution
1 Autoinjector/Pack
25 mcg/0.42 mL
(NDC 55513-090-01)
40 mcg/0.4 mL
(NDC 55513-091-01)
60 mcg/0.3 mL
(NDC 55513-092-01)
100 mcg/0.5 mL
(NDC 55513-093-01)
150 mcg/0.3 mL
(NDC 55513-094-01)
200 mcg/0.4 mL
(NDC 55513-095-01)
300 mcg/0.6 mL
(NDC 55513-096-01)
500 mcg/1 mL
(NDC 55513-097-01)
Single-dose Prefilled SureClick Autoinjector with a 27-gauge, ½-inch needle, Albumin Solution
1 Autoinjector/Pack
25 mcg/0.42 mL
(NDC 55513-080-01)
40 mcg/0.4 mL
(NDC 55513-081-01)
60 mcg/0.3 mL
(NDC 55513-082-01)
100 mcg/0.5 mL
(NDC 55513-083-01)
150 mcg/0.3 mL
(NDC 55513-084-01)
200 mcg/0.4 mL
(NDC 55513-085-01)
300 mcg/0.6 mL
(NDC 55513-086-01)
500 mcg/1 mL
(NDC 55513-087-01)
Page 182
23

Storage
Store at 2° to 8°C (36° to 46°F). Do not freeze or shake. Protect from light.
REFERENCES
1. Egrie JC, Browne JK. Development and characterization of novel erythropoiesis stimulating protein
(NESP). Br J Cancer. 2001; 84 (suppl 1): 3-10.
2. Singh AK, Szczech L, Tang KL, et al. Correction of Anemia with Epoetin Alfa in Chronic Kidney
Disease. N Engl J Med. 2006; 355: 2085-98.
3. Besarab A, Bolton WK, Browne JK, et al. The effects of normal as compared with low hematocrit
values in patients with cardiac disease who are receiving hemodialysis and epoetin. N Engl J Med.
1998; 339: 584-590.
4. Leyland-Jones B, Semiglazov V, Pawlicki M, et al. Maintaining Normal Hemoglobin Levels With
Epoetin Alfa in Mainly Nonanemic Patients With Metastatic Breast Cancer Receiving First-Line
Chemotherapy: A Survival Study. JCO. 2005; 23(25): 1-13.
5. Bohlius J, Wilson J, Seidenfeld J, et al. Recombinant Human Erythropoietins and Cancer Patients:
Updated Meta-Analysis of 57 Studies Including 9353 Patients. J Natl Cancer Inst. 2006; 98: 708-14.
Rx only
This product, or its use, may be covered by one or more US Patents, including US Patent No. 5,618,698,
in addition to others including patents pending.
Manufactured by:
Amgen Manufacturing, Limited, a subsidiary of Amgen Inc.
One Amgen Center Drive
Thousand Oaks, CA 91320-1799
©2001-2010 Amgen Inc. All rights reserved.
* UltraSafe ® is a registered trademark of Safety Syringes, Inc.
1xxxxxx – v23
Revised: 02/2010
Page 183
24

EPOGEN ®
(Epoetin alfa)
FOR INJECTION
Page 184
WARNINGS: INCREASED MORTALITY, SERIOUS CARDIOVASCULAR EVENTS,
THROMBOEMBOLIC EVENTS, STROKE and INCREASED RISK OF TUMOR PROGRESSION OR
RECURRENCE
Chronic Renal Failure:
• In clinical studies, patients experienced greater risks for death, serious cardiovascular
events, and stroke when administered erythropoiesis-stimulating agents (ESAs) to target
hemoglobin levels of 13 g/dL and above.
• Individualize dosing to achieve and maintain hemoglobin levels within the range of 10 to
12 g/dL.
Cancer:
• ESAs shortened overall survival and/or increased the risk of tumor progression or
recurrence in some clinical studies in patients with breast, non-small cell lung, head and
neck, lymphoid, and cervical cancers (see WARNINGS: Table 1).
• To decrease these risks, as well as the risk of serious cardio- and thrombovascular
events, use the lowest dose needed to avoid red blood cell transfusion.
• Because of these risks, prescribers and hospitals must enroll in and comply with the
ESA APPRISE Oncology Program to prescribe and/or dispense EPOGEN ® to patients
with cancer. To enroll in the ESA APPRISE Oncology Program, visit www.esaapprise.com
or call 1-866-284-8089 for further assistance.
• Use ESAs only for treatment of anemia due to concomitant myelosuppressive
chemotherapy.
• ESAs are not indicated for patients receiving myelosuppressive therapy when the
anticipated outcome is cure.
• Discontinue following the completion of a chemotherapy course.
Perisurgery: EPOGEN ® increased the rate of deep venous thromboses in patients not receiving
prophylactic anticoagulation. Consider deep venous thrombosis prophylaxis.
(See WARNINGS: Increased Mortality, Serious Cardiovascular Events, Thromboembolic Events,
and Stroke, WARNINGS: Increased Mortality and/or Increased Risk of Tumor Progression or
Recurrence, INDICATIONS AND USAGE, and DOSAGE AND ADMINISTRATION.)
DESCRIPTION
Erythropoietin is a glycoprotein which stimulates red blood cell production. It is produced in the kidney
and stimulates the division and differentiation of committed erythroid progenitors in the bone marrow.
EPOGEN ® (Epoetin alfa), a 165 amino acid glycoprotein manufactured by recombinant DNA
technology, has the same biological effects as endogenous erythropoietin. 1 It has a molecular weight
of 30,400 daltons and is produced by mammalian cells into which the human erythropoietin gene has
been introduced. The product contains the identical amino acid sequence of isolated natural
erythropoietin.
1

Page 185
EPOGEN ® is formulated as a sterile, colorless liquid in an isotonic sodium chloride/sodium citrate
buffered solution or a sodium chloride/sodium phosphate buffered solution for intravenous (IV) or
subcutaneous (SC) administration.
Single-dose, Preservative-free Vial: Each 1 mL of solution contains 2000, 3000, 4000 or 10,000
Units of Epoetin alfa, 2.5 mg Albumin (Human), 5.8 mg sodium citrate, 5.8 mg sodium chloride, and
0.06 mg citric acid in Water for Injection, USP (pH 6.9 ± 0.3). This formulation contains no
preservative.
Single-dose, Preservative-free Vial: 1 mL (40,000 Units/mL). Each 1 mL of solution contains 40,000
Units of Epoetin alfa, 2.5 mg Albumin (Human), 1.2 mg sodium phosphate monobasic monohydrate,
1.8 mg sodium phosphate dibasic anhydrate, 0.7 mg sodium citrate, 5.8 mg sodium chloride, and 6.8
mcg citric acid in Water for Injection, USP (pH 6.9 ± 0.3). This formulation contains no preservative.
Multidose, Preserved Vial: 2 mL (20,000 Units, 10,000 Units/mL). Each 1 mL of solution contains
10,000 Units of Epoetin alfa, 2.5 mg Albumin (Human), 1.3 mg sodium citrate, 8.2 mg sodium chloride,
0.11 mg citric acid, and 1% benzyl alcohol as preservative in Water for Injection, USP (pH 6.1 ± 0.3).
Multidose, Preserved Vial: 1 mL (20,000 Units/mL). Each 1 mL of solution contains 20,000 Units of
Epoetin alfa, 2.5 mg Albumin (Human), 1.3 mg sodium citrate, 8.2 mg sodium chloride, 0.11 mg citric
acid, and 1% benzyl alcohol as preservative in Water for Injection, USP (pH 6.1 ± 0.3).
CLINICAL PHARMACOLOGY
Chronic Renal Failure Patients
Endogenous production of erythropoietin is normally regulated by the level of tissue oxygenation.
Hypoxia and anemia generally increase the production of erythropoietin, which in turn stimulates
erythropoiesis. 2 In normal subjects, plasma erythropoietin levels range from 0.01 to 0.03 Units/mL and
increase up to 100- to 1000-fold during hypoxia or anemia. 2 In contrast, in patients with chronic renal
failure (CRF), production of erythropoietin is impaired, and this erythropoietin deficiency is the primary
cause of their anemia. 3,4
Chronic renal failure is the clinical situation in which there is a progressive and usually irreversible
decline in kidney function. Such patients may manifest the sequelae of renal dysfunction, including
anemia, but do not necessarily require regular dialysis. Patients with end-stage renal disease (ESRD)
are those patients with CRF who require regular dialysis or kidney transplantation for survival.
EPOGEN ® has been shown to stimulate erythropoiesis in anemic patients with CRF, including both
patients on dialysis and those who do not require regular dialysis. 4-12 The first evidence of a response
to the three times weekly (TIW) administration of EPOGEN ® is an increase in the reticulocyte count
within 10 days, followed by increases in the red cell count, hemoglobin, and hematocrit, usually within 2
to 6 weeks. 4,5 Because of the length of time required for erythropoiesis — several days for erythroid
progenitors to mature and be released into the circulation — a clinically significant increase in
hematocrit is usually not observed in less than 2 weeks and may require up to 6 weeks in some
patients. Once the hematocrit reaches the suggested target range (30% to 36%), that level can be
sustained by EPOGEN ® therapy in the absence of iron deficiency and concurrent illnesses.
The rate of hematocrit increase varies between patients and is dependent upon the dose of EPOGEN ® ,
within a therapeutic range of approximately 50 to 300 Units/kg TIW. 4 A greater biologic response is not
observed at doses exceeding 300 Units/kg TIW. 6 Other factors affecting the rate and extent of
response include availability of iron stores, the baseline hematocrit, and the presence of concurrent
medical problems.
2

Zidovudine-treated HIV-infected Patients
Responsiveness to EPOGEN ® in HIV-infected patients is dependent upon the endogenous serum
erythropoietin level prior to treatment. Patients with endogenous serum erythropoietin levels ≤ 500
mUnits/mL, and who are receiving a dose of zidovudine ≤ 4200 mg/week, may respond to EPOGEN ®
therapy. Patients with endogenous serum erythropoietin levels > 500 mUnits/mL do not appear to
respond to EPOGEN ® therapy. In a series of four clinical trials involving 255 patients, 60% to 80% of
HIV-infected patients treated with zidovudine had endogenous serum erythropoietin levels ≤ 500
mUnits/mL.
Response to EPOGEN ® in zidovudine-treated HIV-infected patients is manifested by reduced
transfusion requirements and increased hematocrit.
Cancer Patients on Chemotherapy
A series of clinical trials enrolled 131 anemic cancer patients who received EPOGEN ® TIW and who
were receiving cyclic cisplatin- or non cisplatin-containing chemotherapy. Endogenous baseline serum
erythropoietin levels varied among patients in these trials with approximately 75% (n = 83/110) having
endogenous serum erythropoietin levels ≤ 132 mUnits/mL, and approximately 4% (n = 4/110) of
patients having endogenous serum erythropoietin levels > 500 mUnits/mL. In general, patients with
lower baseline serum erythropoietin levels responded more vigorously to EPOGEN ® than patients with
higher baseline erythropoietin levels. Although no specific serum erythropoietin level can be stipulated
above which patients would be unlikely to respond to EPOGEN ® therapy, treatment of patients with
grossly elevated serum erythropoietin levels (eg, > 200 mUnits/mL) is not recommended.
Pharmacokinetics
In adult and pediatric patients with CRF, the elimination half-life of plasma erythropoietin after
intravenously administered EPOGEN ® ranges from 4 to 13 hours. 13-15 The half-life is approximately
20% longer in CRF patients than that in healthy subjects. After SC administration, peak plasma levels
are achieved within 5 to 24 hours. The half-life is similar between adult patients with serum creatinine
level greater than 3 and not on dialysis and those maintained on dialysis. The pharmacokinetic data
indicate no apparent difference in EPOGEN ® half-life among adult patients above or below 65 years of
age.
The pharmacokinetic profile of EPOGEN ® in children and adolescents appears to be similar to that of
adults. Limited data are available in neonates. 16 A study of 7 preterm very low birth weight neonates
and 10 healthy adults given IV erythropoietin suggested that distribution volume was approximately 1.5
to 2 times higher in the preterm neonates than in the healthy adults, and clearance was approximately
3 times higher in the preterm neonates than in the healthy adults. 39
The pharmacokinetics of EPOGEN ® have not been studied in HIV-infected patients.
Page 186
A pharmacokinetic study comparing 150 Units/kg SC TIW to 40,000 Units SC weekly dosing regimen
was conducted for 4 weeks in healthy subjects (n = 12) and for 6 weeks in anemic cancer patients (n =
32) receiving cyclic chemotherapy. There was no accumulation of serum erythropoietin after the 2
dosing regimens during the study period. The 40,000 Units weekly regimen had a higher Cmax (3- to 7fold),
longer Tmax (2- to 3-fold), higher AUC0-168h (2- to 3-fold) of erythropoietin and lower clearance
(50%) than the 150 Units/kg TIW regimen. In anemic cancer patients, the average t1/2 was similar (40
hours with range of 16 to 67 hours) after both dosing regimens. After the 150 Units/kg TIW dosing, the
values of Tmax and clearance are similar (13.3 ± 12.4 vs. 14.2 ± 6.7 hours, and 20.2 ± 15.9 vs. 23.6 ±
9.5 mL/h/kg) between Week 1 when patients were receiving chemotherapy (n = 14) and Week 3 when
patients were not receiving chemotherapy (n = 4). Differences were observed after the 40,000 Units
weekly dosing with longer Tmax (38 ± 18 hours) and lower clearance (9.2 ± 4.7 mL/h/kg) during Week 1
3

Page 187
when patients were receiving chemotherapy (n = 18) compared with those (22 ± 4.5 hours, 13.9 ± 7.6
mL/h/kg) during Week 3 when patients were not receiving chemotherapy (n = 7).
The bioequivalence between the 10,000 Units/mL citrate-buffered Epoetin alfa formulation and the
40,000 Units/mL phosphate-buffered Epoetin alfa formulation has been demonstrated after SC
administration of single 750 Units/kg doses to healthy subjects.
INDICATIONS AND USAGE
Treatment of Anemia of Chronic Renal Failure Patients
EPOGEN ® is indicated for the treatment of anemia associated with CRF, including patients on dialysis
and patients not on dialysis. EPOGEN ® is indicated to elevate or maintain the red blood cell level (as
manifested by the hematocrit or hemoglobin determinations) and to decrease the need for transfusions
in these patients.
Non-dialysis patients with symptomatic anemia considered for therapy should have a hemoglobin less
than 10 g/dL.
EPOGEN ® is not intended for patients who require immediate correction of severe anemia. EPOGEN ®
may obviate the need for maintenance transfusions but is not a substitute for emergency transfusion.
Prior to initiation of therapy, the patient’s iron stores should be evaluated. Transferrin saturation should
be at least 20% and ferritin at least 100 ng/mL. Blood pressure should be adequately controlled prior to
initiation of EPOGEN ® therapy, and must be closely monitored and controlled during therapy.
Treatment of Anemia in Zidovudine-treated HIV-infected Patients
EPOGEN ® is indicated for the treatment of anemia related to therapy with zidovudine in HIV-infected
patients. EPOGEN ® is indicated to elevate or maintain the red blood cell level (as manifested by the
hematocrit or hemoglobin determinations) and to decrease the need for transfusions in these patients.
EPOGEN ® is not indicated for the treatment of anemia in HIV-infected patients due to other factors
such as iron or folate deficiencies, hemolysis, or gastrointestinal bleeding, which should be managed
appropriately. EPOGEN ® use has not been demonstrated in controlled clinical trials to improve
symptoms of anemia, quality of life, fatigue, or patient well-being.
EPOGEN ® , at a dose of 100 Units/kg TIW, is effective in decreasing the transfusion requirement and
increasing the red blood cell level of anemic, HIV-infected patients treated with zidovudine, when the
endogenous serum erythropoietin level is ≤ 500 mUnits/mL and when patients are receiving a dose of
zidovudine ≤ 4200 mg/week.
Treatment of Anemia in Cancer Patients on Chemotherapy
EPOGEN ® is indicated for the treatment of anemia due to the effect of concomitantly administered
chemotherapy based on studies that have shown a reduction in the need for RBC transfusions in
patients with metastatic, non-myeloid malignancies receiving chemotherapy for a minimum of 2 months.
Studies to determine whether EPOGEN ® increases mortality or decreases progression-free/recurrencefree
survival are ongoing.
• EPOGEN ® is not indicated for use in patients receiving hormonal agents, therapeutic biologic
products, or radiotherapy unless receiving concomitant myelosuppressive chemotherapy.
• EPOGEN ® is not indicated for patients receiving myelosuppressive therapy when the
anticipated outcome is cure due to the absence of studies that adequately characterize the
impact of EPOGEN ® on progression-free and overall survival (see WARNINGS: Increased
Mortality and/or Increased Risk of Tumor Progression or Recurrence).
4

• EPOGEN ® is not indicated for the treatment of anemia in cancer patients due to other factors
such as iron or folate deficiencies, hemolysis, or gastrointestinal bleeding (see PRECAUTIONS:
Lack or Loss of Response).
• EPOGEN ® use has not been demonstrated in controlled clinical trials to improve symptoms of
anemia, quality of life, fatigue, or patient well-being.
Reduction of Allogeneic Blood Transfusion in Surgery Patients
EPOGEN ® is indicated for the treatment of anemic patients (hemoglobin > 10 to ≤ 13 g/dL) who are at
high risk for perioperative blood loss from elective, noncardiac, nonvascular surgery to reduce the need
for allogeneic blood transfusions. 17-19 EPOGEN ® is not indicated for anemic patients who are willing to
donate autologous blood (see BOXED WARNINGS and DOSAGE AND ADMINISTRATION).
CLINICAL EXPERIENCE: RESPONSE TO EPOGEN ®
Chronic Renal Failure Patients
When dosed with EPOGEN ® , patients responded with an increase in hematocrit. 5 After 3 months on
study, more than 95% of patients were transfusion-independent.
In the presence of adequate iron stores (see IRON EVALUATION), the time to reach the target
hematocrit is a function of the baseline hematocrit and the rate of hematocrit rise.
The rate of increase in hematocrit is dependent upon the dose of EPOGEN ® administered and
individual patient variation. In clinical trials at starting doses of 50 to 150 Units/kg TIW, adult patients
responded with an average rate of hematocrit rise of:
Starting Dose Hematocrit Increase
(TIW IV) Points/Day Points/2 Weeks
50 Units/kg 0.11 1.5
100 Units/kg 0.18 2.5
150 Units/kg 0.25 3.5
Page 188
In a 26 week, double-blind, placebo-controlled trial, 118 anemic dialysis patients with an average
hemoglobin of approximately 7 g/dL were randomized to either EPOGEN ® or placebo. By the end of
the study, average hemoglobin increased to approximately 11 g/dL in the EPOGEN ® -treated patients
and remained unchanged in patients receiving placebo. EPOGEN ® -treated patients experienced
improvements in exercise tolerance and patient-reported physical functioning at month 2 that was
maintained throughout the study.
Adult Patients on Dialysis: Thirteen clinical studies were conducted, involving IV administration to a
total of 1010 anemic patients on dialysis for 986 patient-years of EPOGEN ® therapy. In the three
largest of these clinical trials, the median maintenance dose necessary to maintain the hematocrit
between 30% to 36% was approximately 75 Units/kg TIW. In the US multicenter phase 3 study,
approximately 65% of the patients required doses of 100 Units/kg TIW, or less, to maintain their
hematocrit at approximately 35%. Almost 10% of patients required a dose of 25 Units/kg, or less, and
approximately 10% required a dose of more than 200 Units/kg TIW to maintain their hematocrit at this
level.
A multicenter unit dose study was also conducted in 119 patients receiving peritoneal dialysis who selfadministered
EPOGEN ® subcutaneously for approximately 109 patient-years of experience. Patients
responded to EPOGEN ® administered SC in a manner similar to patients receiving IV administration. 20
5

Page 189
Pediatric Patients on Dialysis: One hundred twenty-eight children from 2 months to 19 years of age with
CRF requiring dialysis were enrolled in 4 clinical studies of EPOGEN ® . The largest study was a
placebo-controlled, randomized trial in 113 children with anemia (hematocrit ≤ 27%) undergoing
peritoneal dialysis or hemodialysis. The initial dose of EPOGEN ® was 50 Units/kg IV or SC TIW. The
dose of study drug was titrated to achieve either a hematocrit of 30% to 36% or an absolute increase in
hematocrit of 6 percentage points over baseline.
At the end of the initial 12 weeks, a statistically significant rise in mean hematocrit (9.4% vs 0.9%) was
observed only in the EPOGEN ® arm. The proportion of children achieving a hematocrit of 30%, or an
increase in hematocrit of 6 percentage points over baseline, at any time during the first 12 weeks was
higher in the EPOGEN ® arm (96% vs 58%). Within 12 weeks of initiating EPOGEN ® therapy, 92.3% of
the pediatric patients were transfusion-independent as compared to 65.4% who received placebo.
Among patients who received 36 weeks of EPOGEN ® , hemodialysis patients required a higher median
maintenance dose (167 Units/kg/week [n = 28] vs 76 Units/kg/week [n = 36]) and took longer to achieve
a hematocrit of 30% to 36% (median time to response 69 days vs 32 days) than patients undergoing
peritoneal dialysis.
Patients With CRF Not Requiring Dialysis
Four clinical trials were conducted in patients with CRF not on dialysis involving 181 patients treated
with EPOGEN ® for approximately 67 patient-years of experience. These patients responded to
EPOGEN ® therapy in a manner similar to that observed in patients on dialysis. Patients with CRF not
on dialysis demonstrated a dose-dependent and sustained increase in hematocrit when EPOGEN ® was
administered by either an IV or SC route, with similar rates of rise of hematocrit when EPOGEN ® was
administered by either route. Moreover, EPOGEN ® doses of 75 to 150 Units/kg per week have been
shown to maintain hematocrits of 36% to 38% for up to 6 months. 21-22
Zidovudine-treated HIV-infected Patients
Efficacy in HIV-infected patients with anemia related to therapy with zidovudine was demonstrated
based on reduction in the requirement for RBC transfusions.
EPOGEN ® has been studied in four placebo-controlled trials enrolling 297 anemic (hematocrit < 30%)
HIV-infected (AIDS) patients receiving concomitant therapy with zidovudine (all patients were treated
with Epoetin alfa manufactured by Amgen Inc). In the subgroup of patients (89/125 EPOGEN ® and
88/130 placebo) with prestudy endogenous serum erythropoietin levels ≤ 500 mUnits/mL, EPOGEN ®
reduced the mean cumulative number of units of blood transfused per patient by approximately 40% as
compared to the placebo group. 24 Among those patients who required transfusions at baseline, 43% of
patients treated with EPOGEN ® versus 18% of placebo-treated patients were transfusion-independent
during the second and third months of therapy. EPOGEN ® therapy also resulted in significant
increases in hematocrit in comparison to placebo. When examining the results according to the weekly
dose of zidovudine received during month 3 of therapy, there was a statistically significant (p < 0.003)
reduction in transfusion requirements in patients treated with EPOGEN ® (n = 51) compared to placebo
treated patients (n = 54) whose mean weekly zidovudine dose was ≤ 4200 mg/week. 23
Approximately 17% of the patients with endogenous serum erythropoietin levels ≤ 500 mUnits/mL
receiving EPOGEN ® in doses from 100 to 200 Units/kg TIW achieved a hematocrit of 38% without
administration of transfusions or significant reduction in zidovudine dose. In the subgroup of patients
whose prestudy endogenous serum erythropoietin levels were > 500 mUnits/mL, EPOGEN ® therapy
did not reduce transfusion requirements or increase hematocrit, compared to the corresponding
responses in placebo-treated patients.
6

In a 6 month open-label EPOGEN ® study, patients responded with decreased transfusion requirements
and sustained increases in hematocrit and hemoglobin with doses of EPOGEN ® up to 300 Units/kg
TIW. 23-25
Responsiveness to EPOGEN ® therapy may be blunted by intercurrent infectious/inflammatory episodes
and by an increase in zidovudine dosage. Consequently, the dose of EPOGEN ® must be titrated based
on these factors to maintain the desired erythropoietic response.
Cancer Patients on Chemotherapy
Adult Patients
Efficacy in patients with anemia due to concomitant chemotherapy was demonstrated based on
reduction in the requirement for RBC transfusions.
Three-Times Weekly (TIW) Dosing
EPOGEN ® administered TIW has been studied in a series of six placebo-controlled, double-blind trials
that enrolled 131 anemic cancer patients receiving EPOGEN ® or matching placebo. Across all studies,
72 patients were treated with concomitant non cisplatin-containing chemotherapy regimens and 59
patients were treated with concomitant cisplatin-containing chemotherapy regimens. Patients were
randomized to EPOGEN ® 150 Units/kg or placebo subcutaneously TIW for 12 weeks in each study.
The results of the pooled data from these six studies are shown in the table below. Because of the
length of time required for erythropoiesis and red cell maturation, the efficacy of EPOGEN ® (reduction
in proportion of patients requiring transfusions) is not manifested until 2 to 6 weeks after initiation of
EPOGEN ® .
Chemotherapy
Regimen
Proportion of Patients Transfused During Chemotherapy
(Efficacy Population a )
On Study b During Months 2 and 3 c
Page 190
EPOGEN ® Placebo EPOGEN ® Placebo
Regimens without
cisplatin
44% (15/34) 44% (16/36) 21% (6/29) 33% (11/33)
Regimens
containing cisplatin
50% (14/28) 63% (19/30) 23% (5/22) d
56% (14/25)
Combined 47% (29/62) 53% (35/66) 22% (11/51) d
43% (25/58)
a ®
Limited to patients remaining on study at least 15 days (1 patient excluded from EPOGEN , 2 patients excluded from
placebo).
b
c
d
Includes all transfusions from day 1 through the end of study.
Limited to patients remaining on study beyond week 6 and includes only transfusions during weeks 5-12.
Unadjusted 2-sided p < 0.05
Intensity of chemotherapy in the above trials was not directly assessed, however the degree and timing
of neutropenia was comparable across all trials. Available evidence suggests that patients with
lymphoid and solid cancers respond similarly to EPOGEN ® therapy, and that patients with or without
tumor infiltration of the bone marrow respond similarly to EPOGEN ® therapy.
Weekly (QW) Dosing
EPOGEN ® was also studied in a placebo-controlled, double-blind trial utilizing weekly dosing in a total
of 344 anemic cancer patients. In this trial, 61 (35 placebo arm and 26 in the EPOGEN ® arm) patients
were treated with concomitant cisplatin containing regimens and 283 patients received concomitant
chemotherapy regimens that did not contain cisplatinum. Patients were randomized to EPOGEN ®
7

40,000 Units weekly (n = 174) or placebo (n = 170) SC for a planned treatment period of 16 weeks. If
hemoglobin had not increased by > 1 g/dL after 4 weeks of therapy or the patient received RBC
transfusion during the first 4 weeks of therapy, study drug was increased to 60,000 Units weekly.
Forty-three percent of patients in the Epoetin alfa group required an increase in EPOGEN ® dose to
60,000 Units weekly. 23
Results demonstrated that EPOGEN ® therapy reduced the proportion of patients transfused in day 29
through week 16 of the study as compared to placebo. Twenty-five patients (14%) in the EPOGEN ®
group received transfusions compared to 48 patients (28%) in the placebo group (p = 0.0010) between
day 29 and week 16 or the last day on study.
Comparable intensity of chemotherapy for patients enrolled in the two study arms was suggested by
similarities in mean dose and frequency of administration for the 10 most commonly administered
chemotherapy agents, and similarity in the incidence of changes in chemotherapy during the trial in the
two arms.
Pediatric Patients
The safety and effectiveness of EPOGEN ® were evaluated in a randomized, double-blind, placebocontrolled,
multicenter study in anemic patients ages 5 to 18 receiving chemotherapy for the treatment
of various childhood malignancies. Two hundred twenty-two patients were randomized (1:1) to
EPOGEN ® or placebo. EPOGEN ® was administered at 600 Units/kg (maximum 40,000 Units)
intravenously once per week for 16 weeks. If hemoglobin had not increased by 1g/dL after the first 4-5
weeks of therapy, EPOGEN ® was increased to 900 Units/kg (maximum 60,000 Units). Among the
EPOGEN ® -treated patients 60% required dose escalation to 900 Units/kg/week.
The effect of EPOGEN ® on transfusion requirements is shown in the table below:
Percentage of Patients Transfused:
On Study a
After 28 Days
Post-Randomization
EPOGEN ®
Placebo
EPOGEN
(n=111)
(n=111)
®
Placebo
(n= 111)
(n=111)
65% (72) 77% (86) 51%(57) b
69% (77)
a
Includes all transfusions from day 1 through the end of study
b
Adjusted 2 sided p 10 to ≤ 13 (n = 96), and > 13 to ≤ 15 g/dL (n = 218)] and then randomly
assigned to receive 300 Units/kg EPOGEN ® , 100 Units/kg EPOGEN ® or placebo by SC injection for 10
days before surgery, on the day of surgery, and for 4 days after surgery. 17 All patients received oral
iron and a low-dose post-operative warfarin regimen. 17
8

Treatment with EPOGEN ® 300 Units/kg significantly (p = 0.024) reduced the risk of allogeneic
transfusion in patients with a pretreatment hemoglobin of > 10 to ≤ 13; 5/31 (16%) of EPOGEN ® 300
Units/kg, 6/26 (23%) of EPOGEN ® 100 Units/kg, and 13/29 (45%) of placebo-treated patients were
transfused. 17 There was no significant difference in the number of patients transfused between
EPOGEN ® (9% 300 Units/kg, 6% 100 Units/kg) and placebo (13%) in the > 13 to ≤ 15 g/dL hemoglobin
stratum. There were too few patients in the ≤ 10 g/dL group to determine if EPOGEN ® is useful in this
hemoglobin strata. In the > 10 to ≤ 13 g/dL pretreatment stratum, the mean number of units transfused
per EPOGEN ® -treated patient (0.45 units blood for 300 Units/kg, 0.42 units blood for 100 Units/kg) was
less than the mean transfused per placebo-treated patient (1.14 units) (overall p = 0.028). In addition,
mean hemoglobin, hematocrit, and reticulocyte counts increased significantly during the presurgery
period in patients treated with EPOGEN ® . 17
EPOGEN ® was also studied in an open-label, parallel-group trial enrolling 145 subjects with a
pretreatment hemoglobin level of ≥ 10 to ≤ 13 g/dL who were scheduled for major orthopedic hip or
knee surgery and who were not participating in an autologous program. 18 Subjects were randomly
assigned to receive one of two SC dosing regimens of EPOGEN ® (600 Units/kg once weekly for 3
weeks prior to surgery and on the day of surgery, or 300 Units/kg once daily for 10 days prior to
surgery, on the day of surgery and for 4 days after surgery). All subjects received oral iron and
appropriate pharmacologic anticoagulation therapy.
From pretreatment to presurgery, the mean increase in hemoglobin in the 600 Units/kg weekly group
(1.44 g/dL) was greater than observed in the 300 Units/kg daily group. 18 The mean increase in
absolute reticulocyte count was smaller in the weekly group (0.11 x 10 6 /mm 3 ) compared to the daily
group (0.17 x 10 6 /mm 3 ). Mean hemoglobin levels were similar for the two treatment groups throughout
the postsurgical period.
The erythropoietic response observed in both treatment groups resulted in similar transfusion rates
[11/69 (16%) in the 600 Units/kg weekly group and 14/71 (20%) in the 300 Units/kg daily group]. 18 The
mean number of units transfused per subject was approximately 0.3 units in both treatment groups.
CONTRAINDICATIONS
EPOGEN ® is contraindicated in patients with:
1. Uncontrolled hypertension.
2. Known hypersensitivity to mammalian cell-derived products.
3. Known hypersensitivity to Albumin (Human).
WARNINGS
Pediatrics
Risk in Premature Infants
The multidose preserved formulation contains benzyl alcohol. Benzyl alcohol has been reported to be
associated with an increased incidence of neurological and other complications in premature infants
which are sometimes fatal.
Adults
Page 192
Increased Mortality, Serious Cardiovascular Events, Thromboembolic Events, and Stroke
Patients with chronic renal failure experienced greater risks for death, serious cardiovascular events,
and stroke when administered erythropoiesis-stimulating agents (ESAs) to target hemoglobin levels of
9

Page 193
13 g/dL and above in clinical studies. Patients with chronic renal failure and an insufficient hemoglobin
response to ESA therapy may be at even greater risk for cardiovascular events and mortality than other
patients. EPOGEN ® and other ESAs increased the risks for death and serious cardiovascular events in
controlled clinical trials of patients with cancer. These events included myocardial infarction, stroke,
congestive heart failure, and hemodialysis vascular access thrombosis. A rate of hemoglobin rise of >
1 g/dL over 2 weeks may contribute to these risks.
In a randomized prospective trial, 1432 anemic chronic renal failure patients who were not undergoing
dialysis were assigned to Epoetin alfa (rHuEPO) treatment targeting a maintenance hemoglobin
concentration of 13.5 g/dL or 11.3 g/dL. A major cardiovascular event (death, myocardial infarction,
stroke, or hospitalization for congestive heart failure) occurred among 125 (18%) of the 715 patients in
the higher hemoglobin group compared to 97 (14%) among the 717 patients in the lower hemoglobin
group (HR 1.3, 95% CI: 1.0, 1.7, p = 0.03). 40
In a randomized, double-blind, placebo-controlled study of 4038 patients, there was an increased risk of
stroke when darbepoetin alfa was administered to patients with anemia, type 2 diabetes, and CRF who
were not on dialysis. Patients were randomized to darbepoetin alfa treatment targeted to a hemoglobin
level of 13 g/dL or to placebo. Placebo patients received darbepoetin alfa only if their hemoglobin
levels were less than 9 g/dL. A total of 101 patients receiving darbepoetin alfa experienced stroke
compared to 53 patients receiving placebo (5% vs. 2.6%; HR 1.92, 95% CI: 1.38, 2.68; p < 0.001).
Increased risk for serious cardiovascular events was also reported from a randomized, prospective trial
of 1265 hemodialysis patients with clinically evident cardiac disease (ischemic heart disease or
congestive heart failure). In this trial, patients were assigned to EPOGEN ® treatment targeted to a
maintenance hematocrit of either 42 ± 3% or 30 ± 3%. 37 Increased mortality was observed in 634
patients randomized to a target hematocrit of 42% [221 deaths (35% mortality)] compared to 631
patients targeted to remain at a hematocrit of 30% [185 deaths (29% mortality)]. The reason for the
increased mortality observed in this study is unknown, however, the incidence of non-fatal myocardial
infarctions (3.1% vs. 2.3%), vascular access thromboses (39% vs. 29%), and all other thrombotic
events (22% vs. 18%) were also higher in the group randomized to achieve a hematocrit of 42%. An
increased incidence of thrombotic events has also been observed in patients with cancer treated with
erythropoietic agents.
In a randomized controlled study (referred to as Cancer Study 1 - the ‘BEST’ study) with another ESA
in 939 women with metastatic breast cancer receiving chemotherapy, patients received either weekly
Epoetin alfa or placebo for up to a year. This study was designed to show that survival was superior
when an ESA was administered to prevent anemia (maintain hemoglobin levels between 12 and 14
g/dL or hematocrit between 36% and 42%). The study was terminated prematurely when interim
results demonstrated that a higher mortality at 4 months (8.7% vs. 3.4%) and a higher rate of fatal
thrombotic events (1.1% vs. 0.2%) in the first 4 months of the study were observed among patients
treated with Epoetin alfa. Based on Kaplan-Meier estimates, at the time of study termination, the 12month
survival was lower in the Epoetin alfa group than in the placebo group (70% vs. 76%; HR 1.37,
95% CI: 1.07, 1.75; p = 0.012). 43
A systematic review of 57 randomized controlled trials (including Cancer Studies 1 and 5 - the ‘BEST’
and ‘ENHANCE’ studies) evaluating 9353 patients with cancer compared ESAs plus red blood cell
transfusion with red blood cell transfusion alone for prophylaxis or treatment of anemia in cancer
patients with or without concurrent antineoplastic therapy. An increased relative risk of thromboembolic
events (RR 1.67, 95% CI: 1.35, 2.06; 35 trials and 6769 patients) was observed in ESA-treated
patients. An overall survival hazard ratio of 1.08 (95% CI: 0.99, 1.18; 42 trials and 8167 patients) was
observed in ESA-treated patients. 41
10

An increased incidence of deep vein thrombosis (DVT) in patients receiving Epoetin alfa undergoing
surgical orthopedic procedures has been observed (see ADVERSE REACTIONS, Surgery Patients:
Thrombotic/Vascular Events). In a randomized controlled study (referred to as the ‘SPINE’ study), 681
adult patients, not receiving prophylactic anticoagulation and undergoing spinal surgery, received either
4 doses of 600 U/kg Epoetin alfa (7, 14, and 21 days before surgery, and the day of surgery) and
standard of care (SOC) treatment, or SOC treatment alone. Preliminary analysis showed a higher
incidence of DVT, determined by either Color Flow Duplex Imaging or by clinical symptoms, in the
Epoetin alfa group [16 patients (4.7%)] compared to the SOC group [7 patients (2.1%)]. In addition,
12 patients in the Epoetin alfa group and 7 patients in the SOC group had other thrombotic vascular
events. Deep venous thrombosis prophylaxis should be strongly considered when ESAs are used for
the reduction of allogeneic RBC transfusions in surgical patients (see BOXED WARNINGS and
DOSAGE AND ADMINISTRATION).
Increased mortality was also observed in a randomized placebo-controlled study of EPOGEN ® in adult
patients who were undergoing coronary artery bypass surgery (7 deaths in 126 patients randomized to
EPOGEN ® versus no deaths among 56 patients receiving placebo). Four of these deaths occurred
during the period of study drug administration and all four deaths were associated with thrombotic
events. 42 ESAs are not approved for reduction of allogeneic red blood cell transfusions in patients
scheduled for cardiac surgery.
Increased Mortality and/or Increased Risk of Tumor Progression or Recurrence
Erythropoiesis-stimulating agents resulted in decreased locoregional control/progression-free survival
and/or overall survival (see Table 1). These findings were observed in studies of patients with
advanced head and neck cancer receiving radiation therapy (Cancer Studies 5 and 6), in patients
receiving chemotherapy for metastatic breast cancer (Cancer Study 1) or lymphoid malignancy (Cancer
Study 2), and in patients with non-small cell lung cancer or various malignancies who were not
receiving chemotherapy or radiotherapy (Cancer Studies 7 and 8).
Table 1: Randomized, Controlled Trials with Decreased Survival and/or Decreased Locoregional
Control
Study / Tumor / (n)
Chemotherapy
Cancer Study 1
Metastatic breast
cancer
(n=939)
Cancer Study 2
Lymphoid
malignancy
(n=344)
Cancer Study 3
Early breast cancer
(n=733)
Hemoglobin
Target
12-14 g/dL
13-15 g/dL (M)
13-14 g/dL (F)
12.5-13 g/dL
Achieved
Hemoglobin
(Median
Q1,Q3)
12.9 g/dL
12.2, 13.3 g/dL
11.0 g/dL
9.8, 12.1 g/dL
13.1 g/dL
12.5, 13.7 g/dL
Primary Endpoint
12-month overall
survival
Proportion of
patients achieving
a hemoglobin
response
Relapse-free and
overall survival
Page 194
Adverse Outcome for ESA-containing
Arm
Decreased 12-month survival
Decreased overall survival
Decreased 3 yr. relapse-free and overall
survival
11

Cancer Study 4
Cervical Cancer
(n=114)
Radiotherapy Alone
Cancer Study 5
Head and neck
cancer
(n=351)
Cancer Study 6
Head and neck
cancer
(n=522)
12-14 g/dL
≥15 g/dL (M)
≥14 g/dL (F)
No Chemotherapy or Radiotherapy
Cancer Study 7
Non-small cell lung
cancer
(n=70)
Cancer Study 8
Non-myeloid
malignancy
(n=989)
12.7 g/dL
12.1, 13.3 g/dL
Not available
14-15.5 g/dL Not available
Progression-free
and overall survival
and locoregional
control
Locoregional
progression-free
survival
Locoregional
disease control
Decreased 3 yr. progression-free and
overall survival and locoregional control
Decreased 5-year locoregional
progression-free survival
Decreased overall survival
Decreased locoregional disease control
12-14 g/dL Not available Quality of life Decreased overall survival
12-13 g/dL
Decreased overall survival:
10.6 g/dL
9.4, 11.8 g/dL
Page 195
RBC transfusions Decreased overall survival
Cancer Study 1 (the ‘BEST’ study) was previously described (see WARNINGS: Increased Mortality,
Serious Cardiovascular Events, Thromboembolic Events, and Stroke). Mortality at 4 months (8.7% vs.
3.4%) was significantly higher in the Epoetin alfa arm. The most common investigator-attributed cause
of death within the first 4 months was disease progression; 28 of 41 deaths in the Epoetin alfa arm and
13 of 16 deaths in the placebo arm were attributed to disease progression. Investigator assessed time
to tumor progression was not different between the two groups. Survival at 12 months was significantly
lower in the Epoetin alfa arm (70% vs. 76%, HR 1.37, 95% CI: 1.07, 1.75; p = 0.012). 43
Cancer Study 2 was a Phase 3, double-blind, randomized (darbepoetin alfa vs. placebo) study
conducted in 344 anemic patients with lymphoid malignancy receiving chemotherapy. With a median
follow-up of 29 months, overall mortality rates were significantly higher among patients randomized to
darbepoetin alfa as compared to placebo (HR 1.36, 95% CI: 1.02, 1.82).
Cancer Study 7 was a Phase 3, multicenter, randomized (Epoetin alfa vs. placebo), double-blind study,
in which patients with advanced non-small cell lung cancer receiving only palliative radiotherapy or no
active therapy were treated with Epoetin alfa to achieve and maintain hemoglobin levels between 12
and 14 g/dL. Following an interim analysis of 70 of 300 patients planned, a significant difference in
survival in favor of the patients on the placebo arm of the trial was observed (median survival 63 vs.
129 days; HR 1.84; p = 0.04).
Cancer Study 8 was a Phase 3, double-blind, randomized (darbepoetin alfa vs. placebo), 16-week
study in 989 anemic patients with active malignant disease, neither receiving nor planning to receive
chemotherapy or radiation therapy. There was no evidence of a statistically significant reduction in
proportion of patients receiving RBC transfusions. The median survival was shorter in the darbepoetin
alfa treatment group (8 months) compared with the placebo group (10.8 months); HR 1.30, 95% CI:
1.07, 1.57.
12

Decreased progression-free survival and overall survival:
Cancer Study 3 (the ‘PREPARE’ study) was a randomized controlled study in which darbepoetin alfa
was administered to prevent anemia conducted in 733 women receiving neo-adjuvant breast cancer
treatment. After a median follow-up of approximately 3 years, the survival rate (86% vs. 90%, HR 1.42,
95% CI: 0.93, 2.18) and relapse-free survival rate (72% vs. 78%, HR 1.33, 95% CI: 0.99, 1.79) were
lower in the darbepoetin alfa-treated arm compared to the control arm.
Cancer Study 4 (protocol GOG 191) was a randomized controlled study that enrolled 114 of a planned
460 cervical cancer patients receiving chemotherapy and radiotherapy. Patients were randomized to
receive Epoetin alfa to maintain hemoglobin between 12 and 14 g/dL or to transfusion support as
needed. The study was terminated prematurely due to an increase in thromboembolic events in
Epoetin alfa-treated patients compared to control (19% vs. 9%). Both local recurrence (21% vs. 20%)
and distant recurrence (12% vs. 7%) were more frequent in Epoetin alfa-treated patients compared to
control. Progression-free survival at 3 years was lower in the Epoetin alfa-treated group compared to
control (59% vs. 62%, HR 1.06, 95% CI: 0.58, 1.91). Overall survival at 3 years was lower in the
Epoetin alfa-treated group compared to control (61% vs. 71%, HR 1.28, 95% CI: 0.68, 2.42).
Cancer Study 5 (the ‘ENHANCE’ study) was a randomized controlled study in 351 head and neck
cancer patients where Epoetin beta or placebo was administered to achieve target hemoglobin of 14
and 15 g/dL for women and men, respectively. Locoregional progression-free survival was significantly
shorter in patients receiving Epoetin beta (HR 1.62, 95% CI: 1.22, 2.14; p = 0.0008) with a median of
406 days Epoetin beta vs. 745 days placebo. Overall survival was significantly shorter in patients
receiving Epoetin beta (HR 1.39, 95% CI: 1.05, 1.84; p = 0.02). 38
Decreased locoregional control:
Cancer Study 6 (DAHANCA 10) was conducted in 522 patients with primary squamous cell carcinoma
of the head and neck receiving radiation therapy randomized to darbepoetin alfa with radiotherapy or
radiotherapy alone. An interim analysis on 484 patients demonstrated that locoregional control at 5
years was significantly shorter in patients receiving darbepoetin alfa (RR 1.44, 95% CI: 1.06, 1.96; p =
0.02). Overall survival was shorter in patients receiving darbepoetin alfa (RR 1.28, 95% CI: 0.98, 1.68;
p = 0.08).
ESA APPRISE Oncology Program
Page 196
Prescribers and hospitals must enroll in and comply with the ESA APPRISE Oncology Program to
prescribe and/or dispense EPOGEN ® to patients with cancer. To enroll, visit www.esa-apprise.com or
call 1-866-284-8089 for further assistance. Additionally, prescribers and patients must provide written
acknowledgement of a discussion of the risks associated with EPOGEN ® .
Pure Red Cell Aplasia
Cases of pure red cell aplasia (PRCA) and of severe anemia, with or without other cytopenias,
associated with neutralizing antibodies to erythropoietin have been reported in patients treated with
EPOGEN ® . This has been reported predominantly in patients with CRF receiving ESAs by
subcutaneous administration. PRCA has also been reported in patients receiving ESAs while
undergoing treatment for hepatitis C with interferon and ribavirin. Any patient who develops a sudden
loss of response to EPOGEN ® , accompanied by severe anemia and low reticulocyte count, should be
evaluated for the etiology of loss of effect, including the presence of neutralizing antibodies to
erythropoietin (see PRECAUTIONS: Lack or Loss of Response). If anti-erythropoietin antibody-
13

Page 197
associated anemia is suspected, withhold EPOGEN ® and other ESAs. Contact Amgen (1-800-
77AMGEN) to perform assays for binding and neutralizing antibodies. EPOGEN ® should be
permanently discontinued in patients with antibody-mediated anemia. Patients should not be switched
to other ESAs as antibodies may cross-react (see ADVERSE REACTIONS: Immunogenicity).
Albumin (Human)
EPOGEN ® contains albumin, a derivative of human blood. Based on effective donor screening and
product manufacturing processes, it carries an extremely remote risk for transmission of viral diseases.
A theoretical risk for transmission of Creutzfeldt-Jakob disease (CJD) also is considered extremely
remote. No cases of transmission of viral diseases or CJD have ever been identified for albumin.
Chronic Renal Failure Patients
Hypertension: Patients with uncontrolled hypertension should not be treated with EPOGEN ® ; blood
pressure should be controlled adequately before initiation of therapy. Although there do not appear to
be any direct pressor effects of EPOGEN ® , blood pressure may rise during EPOGEN ® therapy. During
the early phase of treatment when the hematocrit is increasing, approximately 25% of patients on
dialysis may require initiation of, or increases in, antihypertensive therapy. Hypertensive
encephalopathy and seizures have been observed in patients with CRF treated with EPOGEN ® .
Special care should be taken to closely monitor and aggressively control blood pressure in patients
treated with EPOGEN ® . Patients should be advised as to the importance of compliance with
antihypertensive therapy and dietary restrictions. If blood pressure is difficult to control by initiation of
appropriate measures, the hemoglobin may be reduced by decreasing or withholding the dose of
EPOGEN ® . A clinically significant decrease in hemoglobin may not be observed for several weeks.
It is recommended that the dose of EPOGEN ® be decreased if the hemoglobin increase exceeds 1 g/dL
in any 2-week period, because of the possible association of excessive rate of rise of hemoglobin with
an exacerbation of hypertension. In CRF patients on hemodialysis with clinically evident ischemic heart
disease or congestive heart failure, the dose of EPOGEN ® should be carefully adjusted to achieve and
maintain hemoglobin levels between 10-12 g/dL (see WARNINGS: Increased Mortality, Serious
Cardiovascular Events, Thromboembolic Events, and Stroke, and DOSAGE AND ADMINISTRATION:
Chronic Renal Failure Patients).
Seizures: Seizures have occurred in patients with CRF participating in EPOGEN ® clinical trials.
In adult patients on dialysis, there was a higher incidence of seizures during the first 90 days of therapy
(occurring in approximately 2.5% of patients) as compared with later timepoints.
Given the potential for an increased risk of seizures during the first 90 days of therapy, blood pressure
and the presence of premonitory neurologic symptoms should be monitored closely. Patients should
be cautioned to avoid potentially hazardous activities such as driving or operating heavy machinery
during this period.
While the relationship between seizures and the rate of rise of hemoglobin is uncertain, it is
recommended that the dose of EPOGEN ® be decreased if the hemoglobin increase exceeds 1 g/dL in
any 2-week period.
Thrombotic Events: During hemodialysis, patients treated with EPOGEN ® may require increased
anticoagulation with heparin to prevent clotting of the artificial kidney (see ADVERSE REACTIONS for
more information about thrombotic events).
Other thrombotic events (eg, myocardial infarction, cerebrovascular accident, transient ischemic attack)
have occurred in clinical trials at an annualized rate of less than 0.04 events per patient year of
14

EPOGEN ® therapy. These trials were conducted in adult patients with CRF (whether on dialysis or not)
in whom the target hematocrit was 32% to 40%. However, the risk of thrombotic events, including
vascular access thrombosis, was significantly increased in adult patients with ischemic heart disease or
congestive heart failure receiving EPOGEN ® therapy with the goal of reaching a normal hematocrit
(42%) as compared to a target hematocrit of 30%. Patients with pre-existing cardiovascular disease
should be monitored closely.
Zidovudine-treated HIV-infected Patients
In contrast to CRF patients, EPOGEN ® therapy has not been linked to exacerbation of hypertension,
seizures, and thrombotic events in HIV-infected patients. However, the clinical data do not rule out an
increased risk for serious cardiovascular events.
PRECAUTIONS
The parenteral administration of any biologic product should be attended by appropriate precautions in
case allergic or other untoward reactions occur (see CONTRAINDICATIONS). In clinical trials, while
transient rashes were occasionally observed concurrently with EPOGEN ® therapy, no serious allergic
or anaphylactic reactions were reported (see ADVERSE REACTIONS for more information regarding
allergic reactions).
The safety and efficacy of EPOGEN ® therapy have not been established in patients with a known
history of a seizure disorder or underlying hematologic disease (eg, sickle cell anemia, myelodysplastic
syndromes, or hypercoagulable disorders).
In some female patients, menses have resumed following EPOGEN ® therapy; the possibility of
pregnancy should be discussed and the need for contraception evaluated.
Page 198
Hematology
Exacerbation of porphyria has been observed rarely in patients with CRF treated with EPOGEN ® .
However, EPOGEN ® has not caused increased urinary excretion of porphyrin metabolites in normal
volunteers, even in the presence of a rapid erythropoietic response. Nevertheless, EPOGEN ® should
be used with caution in patients with known porphyria.
In preclinical studies in dogs and rats, but not in monkeys, EPOGEN ® therapy was associated with
subclinical bone marrow fibrosis. Bone marrow fibrosis is a known complication of CRF in humans and
may be related to secondary hyperparathyroidism or unknown factors. The incidence of bone marrow
fibrosis was not increased in a study of adult patients on dialysis who were treated with EPOGEN ® for
12 to 19 months, compared to the incidence of bone marrow fibrosis in a matched group of patients
who had not been treated with EPOGEN ® .
Hemoglobin in CRF patients should be measured twice a week; zidovudine-treated HIV-infected and
cancer patients should have hemoglobin measured once a week until hemoglobin has been stabilized,
and measured periodically thereafter.
Lack or Loss of Response
If the patient fails to respond or to maintain a response to doses within the recommended dosing range,
the following etiologies should be considered and evaluated:
1. Iron deficiency: Virtually all patients will eventually require supplemental iron therapy (see IRON
EVALUATION).
2. Underlying infectious, inflammatory, or malignant processes.
3. Occult blood loss.
15

Page 199
4. Underlying hematologic diseases (ie, thalassemia, refractory anemia, or other myelodysplastic
disorders).
5. Vitamin deficiencies: Folic acid or vitamin B12.
6. Hemolysis.
7. Aluminum intoxication.
8. Osteitis fibrosa cystica.
9. Pure Red Cell Aplasia (PRCA) or anti-erythropoietin antibody-associated anemia: In the absence of
another etiology, the patient should be evaluated for evidence of PRCA and sera should be tested
for the presence of antibodies to erythropoietin (see WARNINGS: Pure Red Cell Aplasia).
See DOSAGE AND ADMINISTRATION: Chronic Renal Failure Patients for management of patients
with an insufficient hemoglobin response to EPOGEN ® therapy.
Iron Evaluation
During EPOGEN ® therapy, absolute or functional iron deficiency may develop. Functional iron
deficiency, with normal ferritin levels but low transferrin saturation, is presumably due to the inability to
mobilize iron stores rapidly enough to support increased erythropoiesis. Transferrin saturation should
be at least 20% and ferritin should be at least 100 ng/mL.
Prior to and during EPOGEN ® therapy, the patient’s iron status, including transferrin saturation (serum
iron divided by iron binding capacity) and serum ferritin, should be evaluated. Virtually all patients will
eventually require supplemental iron to increase or maintain transferrin saturation to levels which will
adequately support erythropoiesis stimulated by EPOGEN ® . All surgery patients being treated with
EPOGEN ® should receive adequate iron supplementation throughout the course of therapy in order to
support erythropoiesis and avoid depletion of iron stores.
Drug Interaction
No evidence of interaction of EPOGEN ® with other drugs was observed in the course of clinical trials.
Carcinogenesis, Mutagenesis, and Impairment of Fertility
Carcinogenic potential of EPOGEN ® has not been evaluated. EPOGEN ® does not induce bacterial
gene mutation (Ames Test), chromosomal aberrations in mammalian cells, micronuclei in mice, or gene
mutation at the HGPRT locus. In female rats treated IV with EPOGEN ® , there was a trend for slightly
increased fetal wastage at doses of 100 and 500 Units/kg.
Pregnancy Category C
EPOGEN ® has been shown to have adverse effects in rats when given in doses 5 times the human
dose. There are no adequate and well-controlled studies in pregnant women. EPOGEN ® should be
used during pregnancy only if potential benefit justifies the potential risk to the fetus.
In studies in female rats, there were decreases in body weight gain, delays in appearance of abdominal
hair, delayed eyelid opening, delayed ossification, and decreases in the number of caudal vertebrae in
the F1 fetuses of the 500 Units/kg group. In female rats treated IV, there was a trend for slightly
increased fetal wastage at doses of 100 and 500 Units/kg. EPOGEN ® has not shown any adverse
effect at doses as high as 500 Units/kg in pregnant rabbits (from day 6 to 18 of gestation).
Nursing Mothers
Postnatal observations of the live offspring (F1 generation) of female rats treated with EPOGEN ® during
gestation and lactation revealed no effect of EPOGEN ® at doses of up to 500 Units/kg. There were,
however, decreases in body weight gain, delays in appearance of abdominal hair, eyelid opening, and
decreases in the number of caudal vertebrae in the F1 fetuses of the 500 Units/kg group. There were
no EPOGEN ® -related effects on the F2 generation fetuses.
16

It is not known whether EPOGEN ® is excreted in human milk. Because many drugs are excreted in
human milk, caution should be exercised when EPOGEN ® is administered to a nursing woman.
Pediatric Use
See WARNINGS: Pediatrics.
Pediatric Patients on Dialysis: EPOGEN ® is indicated in infants (1 month to 2 years), children (2 years
to 12 years), and adolescents (12 years to 16 years) for the treatment of anemia associated with CRF
requiring dialysis. Safety and effectiveness in pediatric patients less than 1 month old have not been
established (see CLINICAL EXPERIENCE: CHRONIC RENAL FAILURE, PEDIATRIC PATIENTS ON
DIALYSIS). The safety data from these studies show that there is no increased risk to pediatric CRF
patients on dialysis when compared to the safety profile of EPOGEN ® in adult CRF patients (see
ADVERSE REACTIONS and WARNINGS). Published literature 27-30 provides supportive evidence of
the safety and effectiveness of EPOGEN ® in pediatric CRF patients on dialysis.
Pediatric Patients Not Requiring Dialysis: Published literature 30,31 has reported the use of EPOGEN ® in
133 pediatric patients with anemia associated with CRF not requiring dialysis, ages 3 months to 20
years‚ treated with 50 to 250 Units/kg SC or IV‚ QW to TIW. Dose-dependent increases in hemoglobin
and hematocrit were observed with reductions in transfusion requirements.
Pediatric HIV-infected Patients: Published literature 32,33 has reported the use of EPOGEN ® in 20
zidovudine-treated anemic HIV-infected pediatric patients ages 8 months to 17 years‚ treated with 50 to
400 Units/kg SC or IV‚ 2 to 3 times per week. Increases in hemoglobin levels and in reticulocyte
counts‚ and decreases in or elimination of blood transfusions were observed.
Pediatric Cancer Patients on Chemotherapy: The safety and effectiveness of EPOGEN ® were
evaluated in a randomized, double-blind, placebo-controlled, multicenter study (see CLINICAL
EXPERIENCE, Weekly (QW) Dosing, Pediatric Patients).
Page 200
Geriatric Use
Among 1051 patients enrolled in the 5 clinical trials of EPOGEN ® for reduction of allogeneic blood
transfusions in patients undergoing elective surgery 745 received EPOGEN ® and 306 received
placebo. Of the 745 patients who received EPOGEN ® , 432 (58%) were aged 65 and over, while 175
(23%) were 75 and over. No overall differences in safety or effectiveness were observed between
geriatric and younger patients. The dose requirements for EPOGEN ® in geriatric and younger patients
within the 4 trials using the TIW schedule were similar. Insufficient numbers of patients were enrolled in
the study using the weekly dosing regimen to determine whether the dosing requirements differ for this
schedule.
Of the 882 patients enrolled in the 3 studies of chronic renal failure patients on dialysis, 757 received
EPOGEN ® and 125 received placebo. Of the 757 patients who received EPOGEN ® , 361 (47%) were
aged 65 and over, while 100 (13%) were 75 and over. No differences in safety or effectiveness were
observed between geriatric and younger patients. Dose selection and adjustment for an elderly patient
should be individualized to achieve and maintain the target hematocrit (See DOSAGE AND
ADMINISTRATION).
Insufficient numbers of patients age 65 or older were enrolled in clinical studies of EPOGEN ® for the
treatment of anemia associated with pre-dialysis chronic renal failure, cancer chemotherapy, and
Zidovudine-treatment of HIV infection to determine whether they respond differently from younger
subjects.
17

Page 201
Information for Patients
Patients should be informed of the increased risks of mortality, serious cardiovascular events,
thromboembolic events, and increased risk of tumor progression or recurrence (see WARNINGS). In
those situations in which the physician determines that a patient or their caregiver can safely and
effectively administer EPOGEN ® at home, instruction as to the proper dosage and administration should
be provided. Patients should be instructed to read the EPOGEN ® Medication Guide and Patient
Instructions for Use and should be informed that the Medication Guide is not a disclosure of all possible
side effects. Patients should be informed of the possible side effects of EPOGEN ® and of the signs and
symptoms of allergic drug reaction and advised of appropriate actions. If home use is prescribed for a
patient, the patient should be thoroughly instructed in the importance of proper disposal and cautioned
against the reuse of needles, syringes, or drug product. A puncture-resistant container should be
available for the disposal of used syringes and needles, and guidance provided on disposal of the full
container.
Chronic Renal Failure Patients
Patients with CRF Not Requiring Dialysis
Blood pressure and hemoglobin should be monitored no less frequently than for patients maintained on
dialysis. Renal function and fluid and electrolyte balance should be closely monitored.
Hematology
Sufficient time should be allowed to determine a patient’s responsiveness to a dosage of EPOGEN ®
before adjusting the dose. Because of the time required for erythropoiesis and the red cell half-life, an
interval of 2 to 6 weeks may occur between the time of a dose adjustment (initiation, increase,
decrease, or discontinuation) and a significant change in hemoglobin.
For patients who respond to EPOGEN ® with a rapid increase in hemoglobin (eg, more than 1 g/dL in
any 2-week period), the dose of EPOGEN ® should be reduced because of the possible association of
excessive rate of rise of hemoglobin with an exacerbation of hypertension.
The elevated bleeding time characteristic of CRF decreases toward normal after correction of anemia in
adult patients treated with EPOGEN ® . Reduction of bleeding time also occurs after correction of
anemia by transfusion.
Laboratory Monitoring
The hemoglobin should be determined twice a week until it has stabilized in the suggested hemoglobin
range and the maintenance dose has been established. After any dose adjustment, the hemoglobin
should also be determined twice weekly for at least 2 to 6 weeks until it has been determined that the
hemoglobin has stabilized in response to the dose change. The hemoglobin should then be monitored
at regular intervals.
A complete blood count with differential and platelet count should be performed regularly. During
clinical trials, modest increases were seen in platelets and white blood cell counts. While these
changes were statistically significant, they were not clinically significant and the values remained within
normal ranges.
In patients with CRF, serum chemistry values (including blood urea nitrogen [BUN], uric acid,
creatinine, phosphorus, and potassium) should be monitored regularly. During clinical trials in adult
patients on dialysis, modest increases were seen in BUN, creatinine, phosphorus, and potassium. In
some adult patients with CRF not on dialysis treated with EPOGEN ® , modest increases in serum uric
acid and phosphorus were observed. While changes were statistically significant, the values remained
within the ranges normally seen in patients with CRF.
18

Page 202
Diet
The importance of compliance with dietary and dialysis prescriptions should be reinforced. In
particular, hyperkalemia is not uncommon in patients with CRF. In US studies in patients on dialysis,
hyperkalemia has occurred at an annualized rate of approximately 0.11 episodes per patient-year of
EPOGEN ® therapy, often in association with poor compliance to medication, diet, and/or dialysis.
Dialysis Management
Therapy with EPOGEN ® results in an increase in hematocrit and a decrease in plasma volume which
could affect dialysis efficiency. In studies to date, the resulting increase in hematocrit did not appear to
adversely affect dialyzer function 8,9 or the efficiency of high flux hemodialysis. 10 During hemodialysis,
patients treated with EPOGEN ® may require increased anticoagulation with heparin to prevent clotting
of the artificial kidney.
Patients who are marginally dialyzed may require adjustments in their dialysis prescription. As with all
patients on dialysis, the serum chemistry values (including BUN, creatinine, phosphorus, and
potassium) in patients treated with EPOGEN ® should be monitored regularly to assure the adequacy of
the dialysis prescription.
Renal Function
In adult patients with CRF not on dialysis, renal function and fluid and electrolyte balance should be
closely monitored. In patients with CRF not on dialysis, placebo-controlled studies of progression of
renal dysfunction over periods of greater than 1 year have not been completed. In shorter term trials in
adult patients with CRF not on dialysis, changes in creatinine and creatinine clearance were not
significantly different in patients treated with EPOGEN ® compared with placebo-treated patients.
Analysis of the slope of 1/serum creatinine versus time plots in these patients indicates no significant
change in the slope after the initiation of EPOGEN ® therapy.
Zidovudine-treated HIV-infected Patients
Hypertension
Exacerbation of hypertension has not been observed in zidovudine-treated HIV-infected patients
treated with EPOGEN ® . However, EPOGEN ® should be withheld in these patients if pre-existing
hypertension is uncontrolled, and should not be started until blood pressure is controlled. In doubleblind
studies, a single seizure has been experienced by a patient treated with EPOGEN ® . 23
Cancer Patients on Chemotherapy
Hypertension
Hypertension, associated with a significant increase in hemoglobin, has been noted rarely in patients
treated with EPOGEN ® . Nevertheless, blood pressure in patients treated with EPOGEN ® should be
monitored carefully, particularly in patients with an underlying history of hypertension or cardiovascular
disease.
Seizures
In double-blind, placebo-controlled trials, 3.2% (n = 2/63) of patients treated with EPOGEN ® TIW and
2.9% (n = 2/68) of placebo-treated patients had seizures. Seizures in 1.6% (n = 1/63) of patients
treated with EPOGEN ® TIW occurred in the context of a significant increase in blood pressure and
hematocrit from baseline values. However, both patients treated with EPOGEN ® also had underlying
CNS pathology which may have been related to seizure activity.
In a placebo-controlled, double-blind trial utilizing weekly dosing with EPOGEN ® , 1.2% (n = 2/168) of
safety-evaluable patients treated with EPOGEN ® and 1% (n = 1/165) of placebo-treated patients had
seizures. Seizures in the patients treated with weekly EPOGEN ® occurred in the context of a
19

significant increase in hemoglobin from baseline values however significant increases in blood pressure
were not seen. These patients may have had other CNS pathology.
Thrombotic Events
In double-blind, placebo-controlled trials, 3.2% (n = 2/63) of patients treated with EPOGEN ® TIW and
11.8% (n = 8/68) of placebo-treated patients had thrombotic events (eg, pulmonary embolism,
cerebrovascular accident), (see WARNINGS: Increased Mortality, Serious Cardiovascular Events,
Thromboembolic Events, and Stroke).
In a placebo-controlled, double-blind trial utilizing weekly dosing with EPOGEN ® , 6.0% (n = 10/168) of
safety-evaluable patients treated with EPOGEN ® and 3.6% (n = 6/165) (p = 0.444) of placebo-treated
patients had clinically significant thrombotic events (deep vein thrombosis requiring anticoagulant
therapy, embolic event including pulmonary embolism, myocardial infarction, cerebral ischemia, left
ventricular failure and thrombotic microangiopathy). A definitive relationship between the rate of
hemoglobin increase and the occurrence of clinically significant thrombotic events could not be
evaluated due to the limited schedule of hemoglobin measurements in this study.
The safety and efficacy of EPOGEN ® were evaluated in a randomized, double-blind, placebocontrolled,
multicenter study that enrolled 222 anemic patients ages 5 to 18 receiving treatment for a
variety of childhood malignancies. Due to the study design (small sample size and the heterogeneity of
the underlying malignancies and of anti-neoplastic treatments employed), a determination of the effect
of EPOGEN ® on the incidence of thrombotic events could not be performed. In the EPOGEN ® arm, the
overall incidence of thrombotic events was 10.8% and the incidence of serious or life-threatening
events was 7.2%.
Surgery Patients
Hypertension
Blood pressure may rise in the perioperative period in patients being treated with EPOGEN ® .
Therefore, blood pressure should be monitored carefully.
Page 203
ADVERSE REACTIONS
Immunogenicity
As with all therapeutic proteins, there is the potential for immunogenicity. Neutralizing antibodies to
erythropoietin, in association with PRCA or severe anemia (with or without other cytopenias), have
been reported in patients receiving EPOGEN ® (see WARNINGS: Pure Red Cell Aplasia) during postmarketing
experience.
There has been no systematic assessment of immune responses, i.e., the incidence of either binding or
neutralizing antibodies to EPOGEN ® , in controlled clinical trials.
Where reported, the incidence of antibody formation is highly dependent on the sensitivity and
specificity of the assay. Additionally, the observed incidence of antibody (including neutralizing
antibody) positivity in an assay may be influenced by several factors including assay methodology,
sample handling, timing of sample collection, concomitant medications, and underlying disease. For
these reasons, comparison of the incidence of antibodies across products within this class
(erythropoietic proteins) may be misleading.
Chronic Renal Failure Patients
In double-blind, placebo-controlled studies involving over 300 patients with CRF, the events reported in
greater than 5% of patients treated with EPOGEN ® during the blinded phase were:
20

Percent of Patients Reporting Event
Patients Treated With Placebo-treated
EPOGEN ® Patients
Event (n = 200) (n = 135)
Hypertension 24% 19%
Headache 16% 12%
Arthralgias 11% 6%
Nausea 11% 9%
Edema 9% 10%
Fatigue 9% 14%
Diarrhea 9% 6%
Vomiting 8% 5%
Chest Pain 7% 9%
Skin Reaction 7% 12%
(Administration Site)
Asthenia 7% 12%
Dizziness 7% 13%
Clotted Access 7% 2%
Significant adverse events of concern in patients with CRF treated in double-blind, placebo-controlled
trials occurred in the following percent of patients during the blinded phase of the studies:
Seizure 1.1% 1.1%
CVA/TIA 0.4% 0.6%
MI 0.4% 1.1%
Death 0% 1.7%
Page 204
In the US EPOGEN ® studies in adult patients on dialysis (over 567 patients), the incidence (number of
events per patient-year) of the most frequently reported adverse events were: hypertension (0.75),
headache (0.40), tachycardia (0.31), nausea/vomiting (0.26), clotted vascular access (0.25), shortness
of breath (0.14), hyperkalemia (0.11), and diarrhea (0.11). Other reported events occurred at a rate of
less than 0.10 events per patient per year.
Events reported to have occurred within several hours of administration of EPOGEN ® were rare, mild,
and transient, and included injection site stinging in dialysis patients and flu-like symptoms such as
arthralgias and myalgias.
In all studies analyzed to date, EPOGEN ® administration was generally well-tolerated, irrespective of
the route of administration.
21

Page 205
Pediatric CRF Patients: In pediatric patients with CRF on dialysis, the pattern of most adverse events
was similar to that found in adults. Additional adverse events reported during the double-blind phase in
>10% of pediatric patients in either treatment group were: abdominal pain, dialysis access
complications including access infections and peritonitis in those receiving peritoneal dialysis, fever,
upper respiratory infection, cough, pharyngitis, and constipation. The rates are similar between the
treatment groups for each event.
Hypertension: Increases in blood pressure have been reported in clinical trials, often during the first
90 days of therapy. On occasion, hypertensive encephalopathy and seizures have been observed in
patients with CRF treated with EPOGEN ® . When data from all patients in the US phase 3 multicenter
trial were analyzed, there was an apparent trend of more reports of hypertensive adverse events in
patients on dialysis with a faster rate of rise of hematocrit (greater than 4 hematocrit points in any 2week
period). However, in a double-blind, placebo-controlled trial, hypertensive adverse events were
not reported at an increased rate in the group treated with EPOGEN ® (150 Units/kg TIW) relative to the
placebo group.
Seizures: There have been 47 seizures in 1010 patients on dialysis treated with EPOGEN ® in clinical
trials, with an exposure of 986 patient-years for a rate of approximately 0.048 events per patient-year.
However, there appeared to be a higher rate of seizures during the first 90 days of therapy (occurring in
approximately 2.5% of patients) when compared to subsequent 90-day periods. The baseline
incidence of seizures in the untreated dialysis population is difficult to determine; it appears to be in the
range of 5% to 10% per patient-year. 34-36
Thrombotic Events: In clinical trials where the maintenance hematocrit was 35 ± 3% on EPOGEN ® ,
clotting of the vascular access (A-V shunt) has occurred at an annualized rate of about 0.25 events per
patient-year, and other thrombotic events (eg, myocardial infarction, cerebral vascular accident,
transient ischemic attack, and pulmonary embolism) occurred at a rate of 0.04 events per patient-year.
In a separate study of 1111 untreated dialysis patients, clotting of the vascular access occurred at a
rate of 0.50 events per patient-year. However, in CRF patients on hemodialysis who also had clinically
evident ischemic heart disease or congestive heart failure, the risk of A-V shunt thrombosis was higher
(39% vs 29%, p < 0.001), and myocardial infarctions, vascular ischemic events, and venous thrombosis
were increased, in patients targeted to a hematocrit of 42 ± 3% compared to those maintained at 30 ±
3% (see WARNINGS).
In patients treated with commercial EPOGEN ® , there have been rare reports of serious or unusual
thromboembolic events including migratory thrombophlebitis, microvascular thrombosis, pulmonary
embolus, and thrombosis of the retinal artery, and temporal and renal veins. A causal relationship has
not been established.
Allergic Reactions: There have been no reports of serious allergic reactions or anaphylaxis
associated with EPOGEN ® administration during clinical trials. Skin rashes and urticaria have been
observed rarely and when reported have generally been mild and transient in nature.
There have been rare reports of potentially serious allergic reactions including urticaria with associated
respiratory symptoms or circumoral edema, or urticaria alone. Most reactions occurred in situations
where a causal relationship could not be established. Symptoms recurred with rechallenge in a few
instances, suggesting that allergic reactivity may occasionally be associated with EPOGEN ® therapy. If
an anaphylactoid reaction occurs, EPOGEN ® should be immediately discontinued and appropriate
therapy initiated.
22

Zidovudine-treated HIV-infected Patients
In double-blind, placebo-controlled studies of 3 months duration involving approximately 300
zidovudine-treated HIV-infected patients, adverse events with an incidence of ≥ 10% in either patients
treated with EPOGEN ® or placebo-treated patients were:
Percent of Patients Reporting Event
Patients Treated With Placebo-treated
EPOGEN ® Patients
Event (n = 144) (n = 153)
Pyrexia 38% 29%
Fatigue 25% 31%
Headache 19% 14%
Cough 18% 14%
Diarrhea 16% 18%
Rash 16% 8%
Congestion, 15% 10%
Respiratory
Nausea 15% 12%
Shortness of Breath 14% 13%
Asthenia 11% 14%
Skin Reaction, 10% 7%
Medication Site
Dizziness 9% 10%
In the 297 patients studied, EPOGEN ® was not associated with significant increases in opportunistic
infections or mortality. 23 In 71 patients from this group treated with EPOGEN ® at 150 Units/kg TIW,
serum p24 antigen levels did not appear to increase. 25 Preliminary data showed no enhancement of
HIV replication in infected cell lines in vitro. 23
Peripheral white blood cell and platelet counts are unchanged following EPOGEN ® therapy.
Page 206
Allergic Reactions: Two zidovudine-treated HIV-infected patients had urticarial reactions within 48
hours of their first exposure to study medication. One patient was treated with EPOGEN ® and one was
treated with placebo (EPOGEN ® vehicle alone). Both patients had positive immediate skin tests
against their study medication with a negative saline control. The basis for this apparent pre-existing
hypersensitivity to components of the EPOGEN ® formulation is unknown, but may be related to HIVinduced
immunosuppression or prior exposure to blood products.
Seizures: In double-blind and open-label trials of EPOGEN ® in zidovudine-treated HIV-infected
patients, 10 patients have experienced seizures. 23 In general, these seizures appear to be related to
underlying pathology such as meningitis or cerebral neoplasms, not EPOGEN ® therapy.
23

Cancer Patients on Chemotherapy
In double-blind, placebo-controlled studies of up to 3 months duration involving 131 cancer patients,
adverse events with an incidence > 10% in either patients treated with EPOGEN ® or placebo-treated
patients were as indicated below:
Percent of Patients Reporting Event
Patients Treated With Placebo-treated
EPOGEN ® Patients
Event (n = 63) (n = 68)
Pyrexia 29% 19%
Diarrhea 21%* 7%
Nausea 17%* 32%
Vomiting 17% 15%
Edema 17%* 1%
Asthenia 13% 16%
Fatigue 13% 15%
Shortness of Breath 13% 9%
Parasthesia 11% 6%
Upper Respiratory
Infection
11% 4%
Dizziness 5% 12%
Trunk Pain
* Statistically significant
3%* 16%
Although some statistically significant differences between patients being treated with EPOGEN ® and
placebo-treated patients were noted, the overall safety profile of EPOGEN ® appeared to be consistent
with the disease process of advanced cancer. During double-blind and subsequent open-label therapy
in which patients (n = 72 for total exposure to EPOGEN ® ) were treated for up to 32 weeks with doses
as high as 927 Units/kg, the adverse experience profile of EPOGEN ® was consistent with the
progression of advanced cancer.
Three hundred thirty-three (333) cancer patients enrolled in a placebo-controlled double-blind trial
utilizing Weekly dosing with EPOGEN ® for up to 4 months were evaluable for adverse events.
The incidence of adverse events was similar in both the treatment and placebo arms.
Surgery Patients
Adverse events with an incidence of ≥ 10% are shown in the following table:
Page 207
24

Percent of Patients Reporting Event
Patients Patients Placebo- Patients Patients
Treated Treated treated Treated Treated
With With Patients With With
EPOGEN ®
EPOGEN ®
EPOGEN ®
EPOGEN ®
300 U/kg 100 U/kg 600 U/kg 300 U/kg
Event (n = 112) a
(n = 101) a
(n = 103) a
(n = 73) b
(n = 72) b
Pyrexia 51% 50% 60% 47% 42%
Nausea 48% 43% 45% 45% 58%
Constipation 43% 42% 43% 51% 53%
Skin Reaction, 25% 19% 22% 26% 29%
Medication
Site
Vomiting 22% 12% 14% 21% 29%
Skin Pain 18% 18% 17% 5% 4%
Pruritus 16% 16% 14% 14% 22%
Insomnia 13% 16% 13% 21% 18%
Headache 13% 11% 9% 10% 19%
Dizziness 12% 9% 12% 11% 21%
Urinary Tract 12% 3% 11% 11% 8%
Infection
Hypertension 10% 11% 10% 5% 10%
Diarrhea 10% 7% 12% 10% 6%
Deep Venous 10% 3% 5% 0% c
Thrombosis
Dyspepsia 9% 11% 6% 7% 8%
Anxiety 7% 2% 11% 11% 4%
Edema 6% 11% 8% 11% 7%
a Study including patients undergoing orthopedic surgery treated with EPOGEN ® or placebo for 15 days
b Study including patients undergoing orthopedic surgery treated with EPOGEN ® 600 Units/kg weekly x 4 or 300
Units/kg daily x 15
c Determined by clinical symptoms
Thrombotic/Vascular Events: In three double-blind, placebo-controlled orthopedic surgery studies,
the rate of deep venous thrombosis (DVT) was similar among Epoetin alfa and placebo-treated patients
in the recommended population of patients with a pretreatment hemoglobin of > 10 g/dL to ≤ 13
g/dL. 17,19,26 However, in 2 of 3 orthopedic surgery studies the overall rate (all pretreatment hemoglobin
groups combined) of DVTs detected by postoperative ultrasonography and/or surveillance venography
was higher in the group treated with Epoetin alfa than in the placebo-treated group (11% vs. 6%). This
finding was attributable to the difference in DVT rates observed in the subgroup of patients with
pretreatment hemoglobin > 13 g/dL.
0% c
Page 208
25

In the orthopedic surgery study of patients with pretreatment hemoglobin of > 10 g/dL to ≤ 13 g/dL
which compared two dosing regimens (600 Units/kg weekly x 4 and 300 Units/kg daily x 15), 4 subjects
in the 600 Units/kg weekly EPOGEN ® group (5%) and no subjects in the 300 Units/kg daily group had a
thrombotic vascular event during the study period. 18
In a study examining the use of Epoetin alfa in 182 patients scheduled for coronary artery bypass graft
surgery, 23% of patients treated with Epoetin alfa and 29% treated with placebo experienced
thrombotic/vascular events. There were 4 deaths among the Epoetin alfa-treated patients that were
associated with a thrombotic/vascular event (see WARNINGS).
OVERDOSAGE
The expected manifestations of EPOGEN ® overdosage include signs and symptoms associated with an
excessive and/or rapid increase in hemoglobin concentration, including any of the cardiovascular
events described in WARNINGS and listed in ADVERSE REACTIONS. Patients receiving an
overdosage of EPOGEN ® should be monitored closely for cardiovascular events and hematologic
abnormalities. Polycythemia should be managed acutely with phlebotomy, as clinically indicated.
Following resolution of the effects due to EPOGEN ® overdosage, reintroduction of EPOGEN ® therapy
should be accompanied by close monitoring for evidence of rapid increases in hemoglobin
concentration (>1 gm/dL per 14 days). In patients with an excessive hematopoietic response, reduce
the EPOGEN ® dose in accordance with the recommendations described in DOSAGE AND
ADMINISTRATION.
DOSAGE AND ADMINISTRATION
IMPORTANT: See BOXED WARNINGS and WARNINGS: Increased Mortality,
Serious Cardiovascular Events, Thromboembolic Events, and Stroke.
Chronic Renal Failure Patients
The recommended range for the starting dose of EPOGEN ® is 50 to 100 Units/kg TIW for adult
patients. The recommended starting dose for pediatric CRF patients on dialysis is 50 Units/kg TIW.
Individualize dosing to achieve and maintain hemoglobin levels between 10-12 g/dL. The dose of
EPOGEN ® should be reduced as the hemoglobin approaches 12 g/dL or increases by more than 1 g/dL
in any 2-week period. If hemoglobin excursions outside the recommended range occur, the EPOGEN ®
dose should be adjusted as described below.
EPOGEN ® may be given either as an IV or SC injection. In patients on hemodialysis, the IV route is
recommended (see WARNINGS: Pure Red Cell Aplasia). While the administration of EPOGEN ® is
independent of the dialysis procedure, EPOGEN ® may be administered into the venous line at the end
of the dialysis procedure to obviate the need for additional venous access. In adult patients with CRF
not on dialysis, EPOGEN ® may be given either as an IV or SC injection.
Patients who have been judged competent by their physicians to self-administer EPOGEN ® without
medical or other supervision may give themselves either an IV or SC injection. The table below
provides general therapeutic guidelines for patients with CRF:
Individually titrate to achieve and maintain hemoglobin levels between 10 to 12 g/dL.
Page 209
26

Starting Dose:
Adults 50 to 100 Units/kg TIW; IV or SC
Pediatric Patients 50 Units/kg TIW; IV or SC
Increase Dose by 25% If: 1. Hemoglobin is < 10 g/dL and has not increased by
1 g/dL after 4 weeks of therapy or
2. Hemoglobin decreases below 10 g/dL
Reduce Dose by 25% When: 1. Hemoglobin approaches 12 g/dL or,
2. Hemoglobin increases > 1 g/dL in any 2-week period
During therapy, hematological parameters should be monitored regularly. Doses must be
individualized to ensure that hemoglobin is maintained at an appropriate level for each patient.
Page 210
For patients whose hemoglobin does not attain a level within the range of 10 to 12 g/dL despite the use
of appropriate EPOGEN ® dose titrations over a 12-week period:
� do not administer higher EPOGEN ® doses and use the lowest dose that will maintain a
hemoglobin level sufficient to avoid the need for recurrent RBC transfusions,
� evaluate and treat for other causes of anemia (see PRECAUTIONS: Lack or Loss of
Response), and
� thereafter, hemoglobin should continue to be monitored and if responsiveness improves,
EPOGEN ® dose adjustments should be made as described above; discontinue EPOGEN ® if
responsiveness does not improve and the patient needs recurrent RBC transfusions.
Pretherapy Iron Evaluation: Prior to and during EPOGEN ® therapy, the patient’s iron stores,
including transferrin saturation (serum iron divided by iron binding capacity) and serum ferritin, should
be evaluated. Transferrin saturation should be at least 20%, and ferritin should be at least 100 ng/mL.
Virtually all patients will eventually require supplemental iron to increase or maintain transferrin
saturation to levels that will adequately support erythropoiesis stimulated by EPOGEN ® .
Dose Adjustment: The dose should be adjusted for each patient to achieve and maintain hemoglobin
levels between 10 to 12 g/dL.
Increases in dose should not be made more frequently than once a month. If the hemoglobin is
increasing and approaching 12 g/dL, the dose should be reduced by approximately 25%. If the
hemoglobin continues to increase, dose should be temporarily withheld until the hemoglobin begins to
decrease, at which point therapy should be reinitiated at a dose approximately 25% below the previous
dose. If the hemoglobin increases by more than 1 g/dL in a 2-week period, the dose should be
decreased by approximately 25%.
If the increase in the hemoglobin is less than 1 g/dL over 4 weeks and iron stores are adequate (see
PRECAUTIONS: Laboratory Monitoring), the dose of EPOGEN ® may be increased by approximately
25% of the previous dose. Further increases may be made at 4-week intervals until the specified
hemoglobin is obtained.
Maintenance Dose: The maintenance dose must be individualized for each patient on dialysis. In the
US phase 3 multicenter trial in patients on hemodialysis, the median maintenance dose was 75
Units/kg TIW, with a range from 12.5 to 525 Units/kg TIW. Almost 10% of the patients required a dose
of 25 Units/kg, or less, and approximately 10% of the patients required more than 200 Units/kg TIW to
maintain their hematocrit in the suggested target range. In pediatric hemodialysis and peritoneal
27

dialysis patients, the median maintenance dose was 167 Units/kg/week (49 to 447 Units/kg per week)
and 76 Units/kg per week (24 to 323 Units/kg/week) administered in divided doses (TIW or BIW),
respectively to achieve the target range of 30% to 36%.
If the transferrin saturation is greater than 20%, the dose of EPOGEN ® may be increased. Such dose
increases should not be made more frequently than once a month, unless clinically indicated, as the
response time of the hemoglobin to a dose increase can be 2 to 6 weeks. Hemoglobin should be
measured twice weekly for 2 to 6 weeks following dose increases. In adult patients with CRF not on
dialysis, the dose should also be individualized to maintain hemoglobin levels between 10 to 12 g/dL.
EPOGEN ® doses of 75 to 150 Units/kg/week have been shown to maintain hematocrits of 36% to 38%
for up to 6 months.
Lack or Loss of Response: If a patient fails to respond or maintain a response, an evaluation for
causative factors should be undertaken (see WARNINGS: Pure Red Cell Aplasia, PRECAUTIONS:
Lack or Loss of Response, and PRECAUTIONS: Iron Evaluation). If the transferrin saturation is less
than 20%, supplemental iron should be administered.
Zidovudine-treated HIV-infected Patients
Prior to beginning EPOGEN ® , it is recommended that the endogenous serum erythropoietin level be
determined (prior to transfusion). Available evidence suggests that patients receiving zidovudine with
endogenous serum erythropoietin levels > 500 mUnits/mL are unlikely to respond to therapy with
EPOGEN ® .
In zidovudine-treated HIV-infected patients the dosage of EPOGEN ® should be titrated for each patient
to achieve and maintain the lowest hemoglobin level sufficient to avoid the need for blood transfusion
and not to exceed the upper safety limit of 12 g/dL.
Starting Dose: For adult patients with serum erythropoietin levels ≤ 500 mUnits/mL who are receiving
a dose of zidovudine ≤ 4200 mg/week, the recommended starting dose of EPOGEN ® is 100 Units/kg as
an IV or SC injection TIW for 8 weeks. For pediatric patients, see PRECAUTIONS: PEDIATRIC USE.
Increase Dose: During the dose adjustment phase of therapy, the hemoglobin should be monitored
weekly. If the response is not satisfactory in terms of reducing transfusion requirements or increasing
hemoglobin after 8 weeks of therapy, the dose of EPOGEN ® can be increased by 50 to 100 Units/kg
TIW. Response should be evaluated every 4 to 8 weeks thereafter and the dose adjusted accordingly
by 50 to 100 Units/kg increments TIW. If patients have not responded satisfactorily to an EPOGEN ®
dose of 300 Units/kg TIW, it is unlikely that they will respond to higher doses of EPOGEN ® .
Maintenance Dose: After attainment of the desired response (ie, reduced transfusion requirements or
increased hemoglobin), the dose of EPOGEN ® should be titrated to maintain the response based on
factors such as variations in zidovudine dose and the presence of intercurrent infectious or
inflammatory episodes. If the hemoglobin exceeds the upper safety limit of 12 g/dL, the dose should be
discontinued until the hemoglobin drops below 11 g/dL. The dose should be reduced by 25% when
treatment is resumed and then titrated to maintain the desired hemoglobin.
Cancer Patients on Chemotherapy
Page 211
Only prescribers enrolled in the ESA APPRISE Oncology Program may prescribe and/or dispense
EPOGEN ® (see WARNINGS: ESA APPRISE Oncology Program).
Although no specific serum erythropoietin level has been established which predicts which patients
would be unlikely to respond to EPOGEN ® therapy, treatment of patients with grossly elevated serum
28

erythropoietin levels (eg, > 200 mUnits/mL) is not recommended. Therapy should not be initiated at
hemoglobin levels ≥ 10 g/dL. The hemoglobin should be monitored on a weekly basis in patients
receiving EPOGEN ® therapy until hemoglobin becomes stable. The dose of EPOGEN ® should be
titrated for each patient to achieve and maintain the lowest hemoglobin level sufficient to avoid the need
for blood transfusion (see recommended Dose Modifications, below).
Recommended Dose: The initial recommended dose of EPOGEN ® in adults is 150 Units/kg SC TIW
or 40,000 Units SC Weekly. The initial recommended dose of EPOGEN ® in pediatric patients is 600
Units/kg IV weekly. Discontinue EPOGEN ® following the completion of a chemotherapy course (see
BOXED WARNINGS: Cancer).
Dose Modification
TIW Dosing
Starting Dose:
Adults 150 Units/kg SC TIW
Reduce Dose by 25% when: Hemoglobin reaches a level needed to avoid transfusion or
increases > 1 g/dL in any 2-week period.
Withhold Dose if: Hemoglobin exceeds a level needed to avoid transfusion.
Restart at 25% below the previous dose when the
hemoglobin approaches a level where transfusions may be
required.
Increase Dose to 300 Units/kg TIW if: Response is not satisfactory (no reduction in transfusion
requirements or rise in hemoglobin) after 4 weeks to
achieve and maintain the lowest hemoglobin level
sufficient to avoid the need for RBC transfusion.
Discontinue: If after 8 weeks of therapy there is no response as measured by
hemoglobin levels or if transfusions are still required.
Weekly Dosing
Starting Dose:
Adults
Pediatrics
40,000 Units SC
600 Units/kg IV (maximum 40,000 Units)
Reduce Dose by 25% when: Hemoglobin reaches a level needed to avoid transfusion or
increases > 1 g/dL in any 2-weeks.
Withhold Dose if: Hemoglobin exceeds a level needed to avoid transfusion and
restart at 25% below the previous dose when the hemoglobin
approaches a level where transfusions may be required.
Increase Dose if:
For Adults: 60,000 Units SC
Weekly
For Pediatrics: 900 Units/kg IV
(maximum 60,000 Units) if:
Page 212
Response is not satisfactory (no increase in hemoglobin by ≥ 1
g/dL after 4 weeks of therapy, in the absence of a RBC
transfusion) to achieve and maintain the lowest hemoglobin
level sufficient to avoid the need for RBC transfusion.
29

Discontinue: If after 8 weeks of therapy there is no response as measured by
hemoglobin levels or if transfusions are still required.
Surgery Patients
Prior to initiating treatment with EPOGEN ® a hemoglobin should be obtained to establish that it is > 10
to ≤ 13 g/dL. 17 The recommended dose of EPOGEN ® is 300 Units/kg/day subcutaneously for 10 days
before surgery, on the day of surgery, and for 4 days after surgery.
An alternate dose schedule is 600 Units/kg EPOGEN ® subcutaneously in once weekly doses (21, 14,
and 7 days before surgery) plus a fourth dose on the day of surgery. 18
All patients should receive adequate iron supplementation. Iron supplementation should be initiated no
later than the beginning of treatment with EPOGEN ® and should continue throughout the course of
therapy. Deep venous thrombosis prophylaxis should be strongly considered (see BOXED
WARNINGS).
PREPARATION AND ADMINISTRATION OF EPOGEN ®
1. Do not shake. It is not necessary to shake EPOGEN ® . Prolonged vigorous shaking may denature
any glycoprotein, rendering it biologically inactive.
2. Protect the solution from light. Parenteral drug products should be inspected visually for
particulate matter and discoloration prior to administration. Do not use any vials exhibiting
particulate matter or discoloration.
3. Using aseptic techniques, attach a sterile needle to a sterile syringe. Remove the flip top from the
vial containing EPOGEN ® , and wipe the septum with a disinfectant. Insert the needle into the vial,
and withdraw into the syringe an appropriate volume of solution.
4. Single-dose: 1 mL vial contains no preservative. Use one dose per vial; do not re-enter the vial.
Discard unused portions.
Multidose: 1 mL and 2 mL vials contain preservative. Store at 2° to 8° C after initial entry and
between doses. Discard 21 days after initial entry.
5. Do not dilute or administer in conjunction with other drug solutions. However, at the time of SC
administration, preservative-free EPOGEN ® from single-use vials may be admixed in a syringe with
bacteriostatic 0.9% sodium chloride injection, USP, with benzyl alcohol 0.9% (bacteriostatic saline)
at a 1:1 ratio using aseptic technique. The benzyl alcohol in the bacteriostatic saline acts as a
local anesthetic which may ameliorate SC injection site discomfort. Admixing is not necessary
when using the multidose vials of EPOGEN ® containing benzyl alcohol.
HOW SUPPLIED
EPOGEN ® , containing Epoetin alfa, is available in the following packages:
1 mL Single-dose, Preservative-free Solution
2000 Units/mL (NDC 55513-126-10)
3000 Units/mL (NDC 55513-267-10)
4000 Units/mL (NDC 55513-148-10)
10,000 Units/mL (NDC 55513-144-10)
40,000 Units/mL (NDC 55513-823-10)
Page 213
30

Supplied in dispensing packs containing 10 single-dose vials.2 mL Multidose, Preserved Solution
10,000 Units/mL (NDC 55513-283-10)
1 mL Multidose, Preserved Solution
20,000 Units/mL (NDC 55513-478-10)
Supplied in dispensing packs containing 10 multidose vials.
STORAGE
Store at 2 o to 8 o C (36 o to 46 o F). Do not freeze or shake. Protect from light.
REFERENCES
1. Egrie JC, Strickland TW, Lane J, et al. Characterization and Biological Effects of Recombinant
Human Erythropoietin. Immunobiol. 1986;72:213-224.
2. Graber SE, Krantz SB. Erythropoietin and the Control of Red Cell Production. Ann Rev Med.
1978;29:51-66.
3. Eschbach JW, Adamson JW. Anemia of End-Stage Renal Disease (ESRD). Kidney Intl.
1985;28:1-5.
Page 214
4. Eschbach JW, Egrie JC, Downing MR, et al. Correction of the Anemia of End-Stage Renal Disease
with Recombinant Human Erythropoietin. NEJM. 1987;316:73-78.
5. Eschbach JW, Abdulhadi MH, Browne JK, et al. Recombinant Human Erythropoietin in Anemic
Patients with End-Stage Renal Disease. Ann Intern Med. 1989;111:992-1000.
6. Eschbach JW, Egrie JC, Downing MR, et al. The Use of Recombinant Human Erythropoietin (r-
HuEPO): Effect in End-Stage Renal Disease (ESRD). In: Friedman, Beyer, DeSanto, Giordano,
eds. Prevention of Chronic Uremia. Philadelphia, PA: Field and Wood Inc; 1989: 148-155.
7. Egrie JC, Eschbach JW, McGuire T, Adamson JW. Pharmacokinetics of Recombinant Human
Erythropoietin (r-HuEPO) Administered to Hemodialysis (HD) Patients. Kidney Intl. 1988;33:262.
8. Paganini E, Garcia J, Ellis P, et al. Clinical Sequelae of Correction of Anemia with Recombinant
Human Erythropoietin (r-HuEPO); Urea Kinetics, Dialyzer Function and Reuse. Am J Kid Dis.
1988;11:16.
9. Delano BG, Lundin AP, Golansky R, et al. Dialyzer Urea and Creatinine Clearances Not
Significantly Changed in r-HuEPO Treated Maintenance Hemodialysis (MD) Patients. Kidney Intl.
1988;33:219.
10. Stivelman J, Van Wyck D, Ogden D. Use of Recombinant Erythropoietin (r-HuEPO) with High Flux
Dialysis (HFD) Does Not Worsen Azotemia or Shorten Access Survival. Kidney Intl. 1988;33:239.
11. Lim VS, DeGowin RL, Zavala D, et al. Recombinant Human Erythropoietin Treatment in Pre-
Dialysis Patients: A Double-Blind Placebo Controlled Trial. Ann Int Med.
1989;110:108-114.
12. Stone WJ, Graber SE, Krantz SB, et al. Treatment of the Anemia of Pre-Dialysis Patients with
Recombinant Human Erythropoietin: A Randomized, Placebo-Controlled Trial. Am J Med Sci.
1988;296:171-179.
13. Braun A, Ding R, Seidel C, Fies T, Kurtz A, Scharer K. Pharmacokinetics of recombinant human
erythropoietin applied subcutaneously to children with chronic renal failure. Pediatr Nephrol.
1993;7:61-64.
14. Geva P, Sherwood JB. Pharmacokinetics of recombinant human erythropoietin (rHuEPO) in
pediatric patients on chronic cycling peritoneal dialysis (CCPD). Blood. 1991;78 (Suppl 1):91a.
31

15. Jabs K, Grant JR, Harmon W, et al. Pharmacokinetics of Epoetin alfa (rHuEPO) in pediatric
hemodialysis (HD) patients. J Am Soc Nephrol. 1991;2:380.
16. Kling PJ, Widness JA, Guillery EN, Veng-Pedersen P, Peters C, DeAlarcon PA. Pharmacokinetics
and pharmacodynamics of erythropoietin during therapy in an infant with renal failure. J Pediatr.
1992;121:822-825.
17. deAndrade JR and Jove M. Baseline Hemoglobin as a Predictor of Risk of Transfusion and
Response to Epoetin alfa in Orthoperdic Surgery Patients. Am. J. of Orthoped. 1996;25 (8):533-
542.
18. Goldberg MA and McCutchen JW. A Safety and Efficacy Comparison Study of Two Dosing
Regimens of Epoetin alfa in Patients Undergoing Major Orthopedic Surgery. Am. J. of Orthoped.
1996;25 (8):544-552.
19. Faris PM and Ritter MA. The Effects of Recombinant Human Erythropoietin on Perioperative
Transfusion Requirements in Patients Having a Major Orthopedic Operation. J. Bone and Joint
surgery. 1996;78-A:62-72.
20. Amgen Inc., data on file.
21. Eschbach JW, Kelly MR, Haley NR, et al. Treatment of the Anemia of Progressive Renal Failure
with Recombinant Human Erythropoietin. NEJM. 1989;321:158-163.
22. The US Recombinant Human Erythropoietin Predialysis Study Group. Double-Blind, Placebo-
Controlled Study of the Therapeutic Use of Recombinant Human Erythropoietin for Anemia
Associated with Chronic Renal Failure in Predialysis Patients. Am J Kid Dis. 1991;18:50-59.
23. Ortho Biologics, Inc., data on file.
Page 215
24. Danna RP, Rudnick SA, Abels RI. Erythropoietin Therapy for the Anemia Associated with AIDS
and AIDS Therapy and Cancer. In: MB Garnick, ed. Erythropoietin in Clinical Applications - An
International Perspective. New York, NY: Marcel Dekker; 1990:301-324.
25. Fischl M, Galpin JE, Levine JD, et al. Recombinant Human Erythropoietin for Patients with AIDS
Treated with Zidovudine. NEJM. 1990;322:1488-1493.
26. Laupacis A. Effectiveness of Perioperative Recombinant Human Eyrthropoietin in Elective Hip
Replacement. Lancet. 1993;341:1228-1232.
27. Campos A, Garin EH. Therapy of renal anemia in children and adolescents with recombinant
human erythropoietin (rHuEPO). Clin Pediatr (Phila). 1992;31:94-99.
28. Montini G, Zacchello G, Baraldi E, et al. Benefits and risks of anemia correction with recombinant
human erythropoietin in children maintained by hemodialysis. J Pediatr. 1990;117:556-560.
29. Offner G, Hoyer PF, Latta K, Winkler L, Brodehl J, Scigalla P. One year's experience with
recombinant erythropoietin in children undergoing continuous ambulatory or cycling peritoneal
dialysis. Pediatr Nephrol. 1990;4:498-500.
30. Muller-Wiefel DE, Scigalla P. Specific problems of renal anemia in childhood. Contrib Nephrol.
1988;66:71-84.
31. Scharer K, Klare B, Dressel P, Gretz N. Treatment of renal anemia by subcutaneous erythropoietin
in children with preterminal chronic renal failure. Acta Paediatr. 1993;82:953-958.
32. Mueller BU, Jacobsen RN, Jarosinski P, et al. Erythropoietin for zidovudine-associated anemia in
children with HIV infection. Pediatr AIDS and HIV Infect: Fetus to Adolesc. 1994;5:169-173.
33. Zuccotti GV, Plebani A, Biasucci G, et al. Granulocyte-colony stimulating factor and erythropoietin
therapy in children with human immunodeficiency virus infection. J Int Med Res. 1996;24:115-121.
32

34. Raskin NH, Fishman RA. Neurologic Disorders in Renal Failure (First of Two Parts). NEJM.
1976;294:143-148.
35. Raskin NH and Fishman RA. Neurologic Disorders in Renal Failure (Second of Two Parts). NEJM.
1976;294:204-210.
36. Messing RO, Simon RP. Seizures as a Manifestation of Systemic Disease. Neurologic Clinics.
1986;4:563-584.
37. Besarab A, Bolton WK, Browne JK, et al. The effects of normal as compared with low hematocrit
values in patients with cardiac disease who are receiving hemodialysis and epoetin. NEJM.
1998;339:584-90.
38. Henke, M, Laszig, R, Rübe, C., et al. Erythropoietin to treat head and neck cancer patients with
anaemia undergoing radiotherapy: randomized, double-blind, placebo-controlled trial. The Lancet.
2003; 362:1255-1260.
39. Widness JA, Veng-Pedersen P, Peters C, Pereira LM, Schmidt RL, Lowe SL. Erythropoietin
Pharmacokinetics in Premature Infants: Developmental, Nonlinearity, and Treatment Effects. J
Appl Physiol. 1996;80 (1):140-148.
40. Singh AK, Szczech L, Tang KL, et al. Correction of Anemia with Epoetin Alfa in Chronic Kidney
Disease, N Engl j Med. 2006; 355:2085-98.
41. Bohlius J, Wilson J, Seidenfeld J, et at. Recombinant HumanErythropoietins and Cancer Patients:
Updated Meta-Analysis of 57 Studies Including 9353 Patients. J Natl Cancer Inst. 2006; 98:708-14.
42. D’Ambra MN, Gray RJ, Hillman R, et al. Effect of Recombinant Human Erythropoietin on
Transfusion Risk in Coronary Bypass Patients. Ann Thorac Surg. 1997; 64: 1686-93.
43. Leyland-Jones B, Semiglazov V, Pawlicki M, et al. Maintaining Normal Hemoglobin Levels With
Epoetin Alfa in Mainly Nonanemic Patients With Metastatic Breast Cancer Receiving First-Line
Chemotherapy: A Survival Study. JCO. 2005; 23(25): 1-13.
This product’s label may have been revised after this insert was used in production. For further product
information and the current package insert, please visit www.amgen.com or call our medical information
department toll-free at 1-800-77AMGEN (1-800-772-6436).
Manufactured by:
Amgen Manufacturing, Limited, a subsidiary of Amgen Inc.
One Amgen Center Drive
Thousand Oaks, CA 91320-1799
©1989-2010 Amgen Inc. All rights reserved.
1xxxxxx – v24
Revised: 02/2010
Page 216
33

Technical Appendix 4 – Response to Voting Questions
Technical Appendix 4 – Response to Voting Questions
Page 217

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Evidence for Panel Voting Questions:
Erythropoiesis-stimulating agents (ESA) were developed to fill the need unmet by red
blood cell (RBC) transfusions and to avoid their inherent risks. Red blood cell
transfusions and adjunctive iron are not therapeutically equivalent to ESAs for the
treatment of anemia in chronic kidney disease (CKD) patients. Transfusions are
transiently effective in the treatment of chronic anemia and are associated with a
number of significant risks, including: acute side effects, cumulative exposure to
blood-borne infections, and importantly, allo-sensitization to foreign antigens. Allosensitization
can decrease a patient’s chances for successful transplantation and
degrade the function of the transplanted kidney. Kidney transplantation is the
preferred modality for patients with end stage renal disease because of its superior
survival benefits, superior quality of life and decreased cost. Transfusions are
therefore not considered appropriate as first-line chronic treatment for anemia in
CKD patients and instead are reserved for treatment of acute anemia.
Consequently, any evaluation of the evidence to determine the impact of ESA use on
health care outcomes must, as we have done, include an assessment of the
evidence regarding the impact of ESA on transfusion avoidance.
Amgen Inc., the United States (US) license holder of Epoetin alfa and darbepoetin
alfa, developed ESAs as supportive therapies to stimulate RBC production in order
to elevate and maintain hemoglobin concentrations in patients with chronic renal
failure (CRF). The FDA-approved EPOGEN ® (Epoetin alfa) Package Insert provides
an example of the importance of transfusion avoidance as a core ESA clinical
outcome:
Page 218

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
“EPOGEN ® is indicated for the treatment of anemia associated with CRF,
including patients on dialysis and patients not on dialysis. EPOGEN ® is
indicated to elevate or maintain the red blood cell level (as manifested by the
hematocrit or hemoglobin determinations) and to decrease the need for
transfusions in these patients.
Non-dialysis patients with symptomatic anemia considered for therapy should
have a hemoglobin less than 10 g/dL.
EPOGEN ® is not intended for patients who require immediate correction of
severe anemia. EPOGEN ® may obviate the need for maintenance
transfusions but is not a substitute for emergency transfusion.”
“EPOGEN ® -treated patients experienced improvements in exercise tolerance
and patient-reported physical functioning at month 2 that was maintained
throughout the study.”
For the reasons stated above, Amgen’s response to the MEDCAC questions will only
address the evidence supporting the use of ESA therapy to maintain or raise
hemoglobin levels of anemic CKD patients.
Iron is a necessary cofactor for RBC formation and is clearly necessary therapy for
the patient who is iron deficient. However, iron alone does not replace erythropoietin
in treating the anemia of CKD that results from endogenous erythropoietin
deficiency. While Amgen has carefully considered the most frequently cited
literature, we have not conducted any randomized controlled studies, nor are we
aware of any studies or systemic literature that assess the efficacy or safety of using
iron to raise or maintain the hemoglobin level to affect any of the health outcomes
listed in the table below in patients with CKD on dialysis (dialysis) and CKD not on
dialysis (CKD-NOD). In addition, no long-term safety studies of parenteral iron have
been conducted as part of the registration programs for these products.
We summarize the evidence supporting each panel voting question below. The level
of confidence is a qualitative assessment and to date there is no standardized
method of grading. Because there is no standard and well-accepted grading level,
Amgen has chosen to leave this column blank.
For further clarity in addressing the panel voting questions, we have separated the
summary of evidence into two categories based on the hemoglobin target range
when reported in the literature. This reference is based on the FDA-approved target
Page 219

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
hemoglobin range (10-12 g/dL), which was based on original registration trials that
employed a hematocrit range 35 ± 3% (equivalent to hemoglobin of 10.7-12.7 g/dL).
If a study addressing a specific question targeted a range above the FDA-approved
target hemoglobin, then the study was summarized in a separate row in the table in
order to provide the MEDCAC panel the opportunity to assess it separately. This
separation by labeled range is used in responding to questions 1 – 4.
In reviewing the evidence, we note that the hemoglobin enrollment criteria for the
randomized and single arm studies supporting the registration of EPOGEN ® all used
a pre-treatment hemoglobin of < 10 g/dL. Subsequent studies used a pre-treatment
baseline hemoglobin of < 11 g/dL. Thus, the impact of ESA therapy on these
outcomes will be summarized using the baseline hemoglobin levels specific to the
available evidence for each outcome in 5b below.
Finally, there are significant differences between dialysis and CKD-NOD patients:
e.g., dialysis patients have more co-morbidities, greater intensity of care necessary
for management, higher mortality rate, increased severity of erythropoietin deficiency
and of anemia, and ongoing blood loss. Therefore, as described in our evidence
review, we provide separate responses to the MEDCAC voting questions regarding
the use of ESAs for the treatment of anemia for dialysis and CKD-NOD patients.
Page 220

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 1: How confident are you that there is sufficient evidence to determine whether using a medical intervention (e.g., blood transfusion, iron therapy, or
ESAs) to maintain or raise the hemoglobin or hematocrit levels of anemic CKD patients affects each of the health outcomes below?
Health
Outcome
Exercise
Tolerance
Vascular
events
(including
stroke, MI,
CHF)
Confidence
Level
Dialysis Patients CKD-NOD Patients
Brief Response
For ESAs there is a double-blind
placebo-controlled RCT as well
as single-arm trial data
demonstrating significant
improvements when treating to
approximately the FDA-approved
target hemoglobin (Hb) range (10-
12 g/dL).
Target Hb above the label
range: There is data from one
open-label RCT with defined and
adjudicated endpoints showing
elevated risks of vascular events
when targeting a Hb level of 14
g/dL (above the FDA-approved
target Hb range [10-12 g/dL])
compared to a target of 10 g/dL.
There is one double-blind placebo
controlled trial that demonstrated
an excess in the adverse event of
cerebrovascular disorder when
targeting a Hb range of 11.5-13.5
g/dL.
Supporting
Evidence
Section: 1.4
Figure: 5 and 6
Appendix 1,
Table 4
Section : 1.5,
1.8
Appendix 1,
Table 5
Appendix 2,
Besarab et al
1998
Parfrey 2005
Confidence
Level
Brief Response
For ESAs there is single-arm and nonexperimental
studies data showing
significant improvement when treating
patients to approximately the FDAapproved
target Hb range (10-12 g/dL).
Target Hb above the label range:
There are data from two open-label and
one double-blind placebo-controlled
RCTs with defined and adjudicated
endpoints showing elevated risks of
vascular events when targeting a Hb
level of > 13 g/dL, which is above the
FDA-approved target Hb range (10-
12 g/dL).
Supporting
Evidence
Section: 2.3
Figure: 16, 17
Appendix 1,
Table 8
Section : 2.4, 2.6
Appendix 1,
Table 5
Appendix 2,
Singh 2006;
Pfeffer 2009
Drueke 2006
Page 221

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 1: How confident are you that there is sufficient evidence to determine whether using a medical intervention (e.g., blood transfusion, iron therapy, or
ESAs) to maintain or raise the hemoglobin or hematocrit levels of anemic CKD patients affects each of the health outcomes below?
Health
Outcome
Patient
Perceived
QoL
Survival
time
Confidence
Level
Dialysis Patients CKD-NOD Patients
Brief Response
Target Hb to the label range:
There are no RCT data evaluating
these health outcomes when
treating to the FDA-approved
target Hb range (10-12 g/dL).*
There is one double-blind placebo
controlled RCT with defined and
adjudicated endpoints and singlearm
study data showing
significant improvement when
treating patients to approximately
the FDA-approved Hb target
range (10-12 g/dL).
There is data from one open-label
RCT with defined and adjudicated
endpoints showing elevated risks
of vascular events when targeting
a Hb level of 14 g/dL (above the
FDA-approved target Hb range
[10-12 g/dL]) compared to a
target of 10 g/dL.
Supporting
Evidence
None available
Section: 1.4
Figure: 5 and 6
Appendix 1,
Table 4
Section 1.5
Figure : 9
Appendix 1,
Table 5
Besarab et al
1998
Confidence
Level
Brief Response
Target Hb to the label range: There
are no RCT data evaluating these
health outcomes when treating to the
FDA-approved target Hb range (10-
12 g/dL).
There are single-arm and nonexperimental
studies data showing
significant improvement when treating
patients to approximately the FDAapproved
Hb target range (10-12 g/dL).
There are data from one placebo
controlled double-blind RCT and two
open-label and RCTs with defined and
adjudicated endpoints showing elevated
risks of vascular events when targeting
a Hb level of ≥ 13 g/dL, which is above
the FDA-approved target Hb range (10-
12 g/dL).
Supporting
Evidence
None available
Section: 2.3
Figure: 16, 17
Appendix 1,
Table 8
Section : 2.4, 2.6
Appendix 1,
Table 5
Singh 2006;
Pfeffer 2009
Drueke 2006
Page 222

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 1: How confident are you that there is sufficient evidence to determine whether using a medical intervention (e.g., blood transfusion, iron therapy, or
ESAs) to maintain or raise the hemoglobin or hematocrit levels of anemic CKD patients affects each of the health outcomes below?
Health
Outcome
Transfusion
avoidance
Confidence
Level
Dialysis Patients CKD-NOD Patients
Brief Response
There are no RCT data evaluating
survival when treating to the FDAapproved
target Hb range (10-12
g/dL). The mortality rate in US
dialysis patients in general clinical
practice has declined over the
past 20 years as evaluated by
USRDS, which provides near
complete surveillance of all US
dialysis patients.
A registrational double-blind
placebo-controlled RCT provide
substantial evidence
demonstrating that treatment with
EPOGEN ® to approximately the
FDA-approved target Hb range
(10-12 g/dL) markedly reduced
transfusions (> 90% reduction) in
dialysis patients.
There is substantial evidence that
transfusion avoidance reduces
the likelihood of allo-sensitization,
which can delay or preclude
kidney transplantation and reduce
transplant survival. Successfully
Supporting
Evidence
Confidence
Level
Brief Response
Appendix 2, There are no RCT data evaluating
survival when treating to the FDAapproved
target Hb range (10-12 g/dL).
Section: 1.3,
1.4, 1.5, 1.7
Figure: 4, 7, 8,
11, 12
Appendix 1,
Tables 1a, 1b,
2, 3
Appendix 2,
Ibrahim et al.
AJKD 2008
Appendix 3
One open-label single-arm study
demonstrated a 70% reduction in
transfusions over a six-month period
with ESA treatment to approximately
the FDA-approved target Hb range (10-
12 g/dL).
There is substantial evidence that
transfusion avoidance reduces the
likelihood of allo-sensitization, which
can delay or preclude kidney
transplantation and reduce transplant
survival. Successfully transplanted
patients have significantly greater 5-
and 10-year survival.
Supporting
Evidence
Appendix 2
Section: 2.1, 2.2,
2.3
Figure: 14, 15
Appendix 1,
Tables 1a, 1b, 2,
3
Appendix 2,
Ibrahim et al.
2009
Appendix 3
Page 223

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 1: How confident are you that there is sufficient evidence to determine whether using a medical intervention (e.g., blood transfusion, iron therapy, or
ESAs) to maintain or raise the hemoglobin or hematocrit levels of anemic CKD patients affects each of the health outcomes below?
Health
Outcome
Confidence
Level
Dialysis Patients CKD-NOD Patients
Brief Response
transplanted patients have
significantly greater 5- and 10year
survival.
Supporting
Evidence
Confidence
Level
Brief Response
Supporting
Evidence
Page 224

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 2: For any health outcome listed in Question 1 for which the panel indicates at least intermediate confidence (mean score ≥ 2.5) in the sufficiency of
evidence, how confident are you that maintaining or raising hemoglobin or hematocrit levels of anemic CKD patients improves each such health outcome?
Health
Outcome
Exercise
Tolerance
Vascular
events
(including
stroke, MI,
CHF)
Confidence
Level
Dialysis Patients CKD-NOD Patients
Brief Response
For ESAs there are RCT and
single-arm trial data
demonstrating significant
improvements when treating to
approximately the FDA-approved
target Hb range (10-12 g/dL).
Target Hb above the label
range: There is data from one
open-label RCT with defined and
adjudicated endpoints showing
elevated risks of vascular events
when targeting a Hb level of 14
g/dL (above the FDA-approved
target Hb range [10-12 g/dL])
compared to a target of 10 g/dL.
There is one double-blind placebo
controlled trial that demonstrated
an excess in the adverse event of
Cerebrovascular Disorder when
targeting a Hb ≥ 13 g/dL
Supporting
Evidence
Section: 1.4
Figure: 5 and 6
Appendix 1,
Table 4
Section : 1.5,
1.8
Appendix 1,
Table 5
Besarab et al
1998
Parfrey 2005
Confidence
Level
Brief Response
For ESAs there is single-arm and nonexperimental
studies data showing
significant improvement when treating
patients to approximately the FDAapproved
target Hb range (10-12 g/dL).
Target Hb above the label range:
There are data from one placebo
controlled double-blind RCT and two
open-label and RCTs with defined and
adjudicated endpoints showing elevated
risks of vascular events when targeting
a Hb level of ≥ 13 g/dL, which is above
the FDA-approved target Hb range (10-
12 g/dL).
Supporting
Evidence
Section: 2.3
Figure: 16, 17
Appendix 1,
Table 8
Section : 2.4, 2.6
Appendix 1,
Table 5
Singh 2006;
Pfeffer 2009
Drueke 2006
Page 225

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 2: For any health outcome listed in Question 1 for which the panel indicates at least intermediate confidence (mean score ≥ 2.5) in the sufficiency of
evidence, how confident are you that maintaining or raising hemoglobin or hematocrit levels of anemic CKD patients improves each such health outcome?
Health
Outcome
Patient
Perceived
QoL
Survival
time
Confidence
Level
Dialysis Patients CKD-NOD Patients
Brief Response
Target Hb to the label range:
There are no RCT data evaluating
these health outcomes when
treating to the FDA-approved
target Hb range (10-12 g/dL).
There are RCT and single-arm
study data showing significant
improvement when treating
patients to approximately the
FDA-approved Hb target range
(10-12 g/dL).
There is data from one RCT with
defined and adjudicated
endpoints showing elevated risks
of mortality when targeting a Hb
level of 14 g/dL (above the FDAapproved
target Hb range [10-12
g/dL]) compared to a target of 10
g/dL.
Supporting
Evidence
Appendix 2
Section: 1.4
Figure: 5 and 6
Appendix 1,
Table 4
Section 1.5
Figure : 9
Appendix 1,
Table 5
Besarab et al
1998
Confidence
Level
Brief Response
Target Hb to the label range: There
are no RCT data evaluating these
health outcomes when treating to the
FDA-approved target Hb range (10-12
g/dL).
There are single-arm and nonexperimental
studies data showing
significant improvement when treating
patients to approximately the FDAapproved
Hb target range (10-12 g/dL).
There are data from one placebo
controlled double-blind RCT and two
open-label and RCTs with defined and
adjudicated endpoints showing elevated
risks of vascular events when targeting
a Hb level of ≥ 13 g/dL, which is above
the FDA-approved target Hb range (10-
12 g/dL).
Supporting
Evidence
Appendix 2
Section: 2.3
Figure: 16, 17
Appendix 1,
Table 8
Section : 2.4, 2.6
Appendix 1,
Table 5
Singh 2006;
Pfeffer 2009
Page 226

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 2: For any health outcome listed in Question 1 for which the panel indicates at least intermediate confidence (mean score ≥ 2.5) in the sufficiency of
evidence, how confident are you that maintaining or raising hemoglobin or hematocrit levels of anemic CKD patients improves each such health outcome?
Health
Outcome
Transfusion
avoidance
Confidence
Level
Dialysis Patients CKD-NOD Patients
Brief Response
There are no RCT data evaluating
survival when treating to the FDAapproved
target Hb range (10-
12 g/dL). The mortality rate in US
dialysis patients in general clinical
practice has declined over the
past 20 years as evaluated by
USRDS, which provides near
complete surveillance of all US
dialysis patients.
RCTs provide substantial
evidence demonstrating that
treatment with EPOGEN ® to
approximately the FDA-approved
target Hb range (10-12 g/dL)
markedly reduced transfusions (>
90% reduction) in dialysis
patients.
There is substantial evidence that
transfusion avoidance reduces
the likelihood of allo-sensitization,
which can delay or preclude
kidney transplantation and reduce
transplant survival. Successfully
transplanted patients have
Supporting
Evidence
Confidence
Level
Brief Response
Appendix 2, There are no RCT data evaluating
survival when treating to the FDAapproved
target Hb range (10-12 g/dL).
Section: 1.3,
1.4, 1.5, 1.7
Figure: 4, 7, 8,
11, 12
Appendix 1,
Tables 1a, 1b,
2, 3
Appendix 2,
Ibrahim et al.
AJKD 2008
Appendix 3
There is data from one placebocontrolled
RCT demonstrating a near
halving of transfusions when patients
were treated to a Hb of 13 g/dL as
compared to placebo with rescue at
9 g/dL. 13 g/dL is above the FDA
approved range of 10-12 g/dL.
One open-label single-arm study
demonstrated a 70% reduction in
transfusions over a six-month period
with ESA treatment to approximately
the FDA-approved target Hb range (10-
12 g/dL).
There is substantial evidence that
transfusion avoidance reduces the
Supporting
Evidence
Appendix 2,
Pfeffer 2009
Section: 2.1, 2.2,
2.3
Figure: 14, 15
Appendix 1,
Tables 1a, 1b, 2,
3
Appendix 2,
Ibrahim et al.
2009
Appendix 3
Page 227

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 2: For any health outcome listed in Question 1 for which the panel indicates at least intermediate confidence (mean score ≥ 2.5) in the sufficiency of
evidence, how confident are you that maintaining or raising hemoglobin or hematocrit levels of anemic CKD patients improves each such health outcome?
Health
Outcome
Confidence
Level
Dialysis Patients CKD-NOD Patients
Brief Response
significantly greater 5- and 10year
survival.
Supporting
Evidence
Confidence
Level
Brief Response
likelihood of allo-sensitization, which
can delay or preclude kidney
transplantation and reduce transplant
survival. Successfully transplanted
patients have significantly greater 5-
and 10-year survival.
Supporting
Evidence
Page 228

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 3: For any health outcome addressed in Question 2 for which the panel indicates at least intermediate confidence (mean score > 2.5), how
confident are you that there is sufficient evidence to determine whether the use of ESAs to maintain or raise hemoglobin or hematocrit levels of CKD
patients improves each such health outcome?
Answer 3: Amgen’s response to question # 2 (above) provides a review of the evidence addressing the impact of raising hemoglobin levels with ESAs on
the specific health outcomes.
Page 229

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 4: How confident are you that there is sufficient evidence to determine whether the use of ESAs to maintain or raise the hemoglobin or hematocrit levels of
anemic CKD patients worsens any health outcome listed in Question 1?
Health
Outcome
Worsens
Exercise
Tolerance
Worsens
Vascular
events
including
stroke, MI,
CHF
Confidence
Level
Dialysis Patients CKD-NOD Patients
Brief Response
There are RCT and single-arm study
data showing significant improvement
in exercise when treating patients to the
FDA-approved labeled Hb range of 10-
12 g/dL
There is data from one RCT with
defined and adjudicated endpoints
showing elevated risks of mortality
when targeting a Hb level of 14 g/dL
(above the FDA-approved target Hb
range [10-12 g/dL]) compared to a
target of 10 g/dL.
There is one double-blind placebo
controlled trial that demonstrated an
excess in the adverse event of
cerebrovascular disorder when
targeting a Hb ≥ 13 g/dL
Supporting
Evidence
Section: 1.4
Figure: 5 and 6
Appendix 1,
Table 4
Section: 1.5,
1.8
Appendix 1,
Table 5
Appendix 2,
Besarab 1998
Parfrey 2005
Confidence
Level
Brief Response
For ESAs there are single-arm and
non-experimental study data
showing significant improvement
when treating patients to the FDAapproved
labeled Hb range of 10-12
g/dL
There are data from one placebo
controlled double-blind RCT and
two open-label RCTs with defined
and adjudicated endpoints showing
elevated risks of vascular events
when targeting a Hb level of ≥ 13
g/dL, which is above the FDAapproved
target Hb range (10-12
g/dL).
There is one double-blind placebo
controlled trial that demonstrated an
excess in the adverse event of
Cerebrovascular Disorder when
targeting a Hb ≥ 13 g/dL
Supporting
Evidence
Section: 2.3
Figure: 16, 17
Appendix 1,
Table 8
Section: 2.4, 2.6
Appendix 1,
Table 5
Appendix 2,
Singh 2006;
Pfeffer 2009
Drueke 2006
Page 230

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 4: How confident are you that there is sufficient evidence to determine whether the use of ESAs to maintain or raise the hemoglobin or hematocrit levels of
anemic CKD patients worsens any health outcome listed in Question 1?
Health
Outcome
Worsens
Vascular
events
including
stroke, MI,
CHF
Worsens
Patient
Perceived QoL
Worsens
Survival time
(RCT)
Confidence
Level
Dialysis Patients CKD-NOD Patients
Brief Response
There are no RCT data evaluating
these health outcomes when treating to
the label Hb range (10-12 g/dL).
There is a double-blind placebo
controlled RCT and single-arm study
data showing significant improvement
when treating patients to the FDAapproved
labeled Hb range of 10-12
g/dL.
There is data from one RCT showing
elevated risks of mortality when
targeting a Hb level of 14 g/dL (above
the FDA-approved target Hb range (10-
12 g/dL) compared to a target of 10
g/dL.
Supporting
Evidence
Section: 1.4
Figure: 5 and 6
Appendix 1,
Table 4
Section 1.5
Figure: 9
Appendix 1,
Table 5
Appendix 2,
Besarab et al
1998
Confidence
Level
Brief Response
There are no RCT data evaluating
these health outcomes when
treating to the label Hb range (10-
12 g/dL).
There are single-arm and nonexperimental
studies data showing
significant improvement when
treating patients to the FDAapproved
labeled range of 10-12
g/dL.
There are data from one placebo
controlled double-blind RCT and
two open-label RCTs with defined
and adjudicated endpoints showing
elevated risks of vascular events
when targeting a Hb level of ≥ 13
g/dL, which is above the FDAapproved
target Hb range (10-12
g/dL).
Supporting
Evidence
Section: 2.3
Figure: 16, 17
Appendix 1,
Table 8
Section: 2.4, 2.6
Appendix 1,
Table 5
Appendix 2,
Singh 2006;
Drueke 2006
Page 231

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 4: How confident are you that there is sufficient evidence to determine whether the use of ESAs to maintain or raise the hemoglobin or hematocrit levels of
anemic CKD patients worsens any health outcome listed in Question 1?
Health
Outcome
Worsens
Survival time
Transfusion
avoidance
Confidence
Level
Dialysis Patients CKD-NOD Patients
Brief Response
There are no RCT data evaluating
survival when treating to the label range
(10-12 g/dL).
The mortality rate in US dialysis
patients in general clinical practice has
declined over the past 20 years as
evaluated by USRDS, which provides
near complete surveillance of all US
dialysis patients.
For ESAs there is a placebo-controlled
double-blind RCT and a single-arm
study data showing significant
improvement in transfusion
avoidance/reduction when treating
patients to the FDA-approved labeled
Hb range of 10-12 g/dL.
Supporting
Evidence
Section 1.5
Figure : 9
Appendix 1,
Table 5
Appendix 1,
Table 3
Appendix 2,
Eschbach,
1989
Confidence
Level
Brief Response
There are no RCT CV outcome
data evaluating survival when
treating to the label Hb range (10-
12 g/dL).
For ESAs there are single-arm and
non-experimental study data
showing significant improvement
when treating patients to the FDAapproved
labeled Hb range of 10-12
g/dL.
Supporting
Evidence
None available
Section: 2.3
Figure: 16, 17
Appendix 1,
Table 8
Page 232

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 5a: Please discuss any impact of the following factors on the conclusion reached above.
Answer 5a: There are significant differences between CKD not on dialysis and CKD on dialysis patients: e.g., co-morbidities, intensity of care necessary for
management, mortality rate, severity of erythropoietin deficiency and of anemia, and ongoing blood loss. Therefore, as described in our evidence review, we
provided separate responses regarding the use of ESAs for the treatment of anemia for CKD on dialysis and CKD not on dialysis patients for questions 1 -4 above,
and for question 5 b below.
Question 5b: For any health outcome listed in Question 1 for which the panel indicates at least intermediate confidence (mean score ≥ 2.5) in the sufficiency of
evidence, how confident are you that maintaining or raising hemoglobin or hematocrit levels of anemic CKD patients improves each such health outcome, and
whether this is different for different pre-treatment hemoglobin levels (

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 5b: For any health outcome listed in Question 1 for which the panel indicates at least intermediate confidence (mean score ≥ 2.5) in the sufficiency of
evidence, how confident are you that maintaining or raising hemoglobin or hematocrit levels of anemic CKD patients improves each such health outcome, and
whether this is different for different pre-treatment hemoglobin levels (

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 5b: For any health outcome listed in Question 1 for which the panel indicates at least intermediate confidence (mean score ≥ 2.5) in the sufficiency of
evidence, how confident are you that maintaining or raising hemoglobin or hematocrit levels of anemic CKD patients improves each such health outcome, and
whether this is different for different pre-treatment hemoglobin levels (

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 5b: For any health outcome listed in Question 1 for which the panel indicates at least intermediate confidence (mean score ≥ 2.5) in the sufficiency of
evidence, how confident are you that maintaining or raising hemoglobin or hematocrit levels of anemic CKD patients improves each such health outcome, and
whether this is different for different pre-treatment hemoglobin levels (

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 5b: For any health outcome listed in Question 1 for which the panel indicates at least intermediate confidence (mean score ≥ 2.5) in the sufficiency of
evidence, how confident are you that maintaining or raising hemoglobin or hematocrit levels of anemic CKD patients improves each such health outcome, and
whether this is different for different pre-treatment hemoglobin levels (

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Question 5b: For any health outcome listed in Question 1 for which the panel indicates at least intermediate confidence (mean score ≥ 2.5) in the sufficiency of
evidence, how confident are you that maintaining or raising hemoglobin or hematocrit levels of anemic CKD patients improves each such health outcome, and
whether this is different for different pre-treatment hemoglobin levels (

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
5c. Whether an appropriate target hemoglobin or hematocrit has been set for the
CKD patients?
10-12 g/dL in dialysis, as specified in product labeling. Please see responses to
questions 1-4 where we separately discuss the risks and benefits for dialysis and CKD-
NOD patients
10-12 g/dL in CKD-NOD, as specified in product labeling. Please see responses to
questions 1-4 where we separately discuss the risks and benefits for dialysis and CKD-
NOD patients
5d. Whether the ESA dosing strategy has been implemented to minimize the
rapidity of hemoglobin or hematocrit rise and/or oscillations in their levels.
In dialysis, there are no prospective RCTs evaluating dosing algorithms other than those
in the Amgen labels for EPOGEN ® and Aranesp ® . There are no RCTs specifically
evaluating algorithms that modify rate of rise or variability.
In CKD-NOD, there are no prospective RCTs evaluating dosing algorithms other than
those in the Amgen labels for EPOGEN ® and Aranesp ® . There are no RCTs specifically
evaluating algorithms that modify rate of rise or variability.
5e. Whether the CKD patient has demonstrated a blunted or “non-response” to
interventions to raise hemoglobin or hematocrit.
Page 239
Post hoc analysis of RCTs targeting hemoglobin concentrations ≥ 13 g/dL, suggest that
hyporesponsive dialysis patients are at increased risk of cardiovascular events and
mortality. None of these analyses can confirm or refute that this increased risk is related
to ESA treatment or dose. While there is no RCT evidence for guiding management,
Amgen has proactively modified its label to incorporate prudent dosing guidance for the
hyporesponsive patient.
Post hoc analysis of RCTs targeting hemoglobin concentrations ≥ 13 g/dL, suggest that
hyporesponsive CKD-NOD patients are at increased risk of cardiovascular events and
mortality. None of these analyses can confirm or refute that this increased risk is related
to ESA treatment or dose. While there is no RCT evidence for guiding management,

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
Amgen has proactively modified its label to incorporate prudent dosing guidance for the
hyporesponsive patient.
5f. Whether the CKD patient has been evaluated to determine the etiology (cause)
of the anemia.
Amgen’s ESAs are to be used for the indicated causes of anemia of which the anemia of
dialysis is being discussed herein. The product labeling indicates that iron deficiency
should be repaired at the time of and during the course of therapy. The label also
outlines guidance for patients who lose or have an insufficient hemoglobin response.
These warnings apply to dialysis.
Amgen’s ESAs are to be used for the indicated causes of anemia, of which, the anemia
of CKD-NOD is being discussed herein. The product labeling indicates that iron
deficiency should be repaired at the time of and during the course of therapy. The label
also outlines guidance for patients who lose or have an insufficient hemoglobin
response. These warnings apply to CKD-NOD.
5g. Whether the CKD patient demonstrates cardiac, cerebral or other vascular
comorbidities.
Amgen considers that safety signals derived from RCTs should be applied to all patients
in the absence of compelling and high-quality evidence to identify a specific, at-risk
subgroup. As such, safety signals with regard to cardiovascular morbidity and mortality
are applied to all dialysis patients.
Amgen considers the safety signals derived from RCTs should be applied to all patients
in the absence of compelling and high-quality evidence to identify a specific subgroup at
risk. As such, safety signals with regard to cardiovascular morbidity and mortality are
applied to all CKD-NOD patients.
6. What clinical trials designs would be most desirable to fill in any indentified
evidence gaps?
The totality of the evidence informing the benefit and risks of ESA in CKD patients was
last reviewed at the September 2007 joint meeting between the FDA CRDAC & Drug
Safety and Risk Management Advisory Committee. Three areas were identified for
Page 240

SUMMARY OF THE EVIDENCE ADDRESSING PANEL VOTING QUESTIONS
further study: hyporesponsiveness, variability and target hemoglobin.
Hyporesponsiveness and hemoglobin variability, and dosing strategies that may
minimize variability, are most relevant to dialysis patients. It was agreed that the
completion of TREAT would be an important milestone in our understanding of the use
of ESAs in CKD-NOD.
Page 241
Placebo controlled studies of optimal hemoglobin targets are most relevant to CKD-NOD
patients as a randomized placebo controlled trial is not feasible in the dialysis
population. Progress has been made, including the completion of TREAT and its
communication to FDA, and we look forward to an upcoming DAC meeting with the FDA
to guide future studies.