is There Any Evidence-based Medicine for the Very Old?

Heart Failure 2005
The octogenarian failing heart
Lisbon, 14 June 2005
Pharmacological Treatment: is There
Any Evidence-based Medicine for
the Very Old?
Marco Metra
Cattedra di Cardiologia
Università di Brescia

No, not even for the
slightly old !

Mean Age of the Patients with Heart Failure in
Epidemiological Studies
Author (year)
Chae (1999)
EPICA, Portugal (2000)
Ghali (1997)
Lowe (1998)
MacIntyre (2000)
McCullogh (2002)
Mosterd, Netherlands (2000)
Senni, Minnesota (1998)
SEOSI (1997)
EuroHeart Failure (2003)
No. of patients
221
551
1133
579
66 547
29 686
24 868
216
3 921
46 788
Age (years), M+SD
77.9 + 5
70 + 11
77.6 + 7.9
77
♀: 78; ♂: 72
♀: 73.7; ♂: 69.2
♀: 77.7; ♂: 72.9
77.3 + 12.1
67 + 12
71

Trial
CONSENSUS I
SOLVD-T
DIG
MERIT-HF
CIBIS-II
COPERNICUS
RALES
EPHESUS
ELITE-II
Val-HeFT
CHARM
Mean Age and Proportions of Elderly
Patients in Multicenter Trials
No. of patients
253
6 797
7 788
3 991
2 647
2 289
1 633
6 642
3 152
5 010
7 601
Mean Age
71
61
63
64
61
63
61
64
71
62
66
% > 70 years
50%
15%
27%
32%
…
…
…
58%
…
23% (>75)

Factors Distinguishing HF in the Elderly from
HF at Middle Age
Prevalence
Sex
Etiology
Clinical features
LVEF
Comorbidities
Physician
RCTs
Therapy
Middle age
women
CAD
Typical
Reduced
Few
Cardiologist
Many
Evidence-based
Elderly
˜ 10%
Women > men
Hypertension
Atypical
Normal
Multiple
Primary care
Few
Empiric
Rich MW, Am J Med 2005; 118:342

Factors Distinguishing HF in the Elderly from
HF at Middle Age
Prevalence
Sex
Etiology
Clinical features
LVEF
Comorbidities
Physician
RCTs
Therapy
Middle age
women
CAD
Typical
Reduced
Few
Cardiologist
Many
Evidence-based
Elderly
˜ 10%
Women > men
Hypertension
Atypical
Normal
Multiple
Primary care
Few
Empiric
Rich MW, Am J Med 2005; 118:342

Baseline Characteristics of Patients Recruited in COMET
Subdivided According to Age (mean, 62 ys.; range, 18-90)
Female, %
BMI, m + sd
Dyspnoea score > 4, %
Fatigue score > 4, %
NT-ProBNP, median
Creatinine, median
Diabetes, %
Atrial fibrillation, %
Paced rhythm, %
< 60 years
(n = 1171)
17%
28 + 5
13%
13%
804
92
20%
15%
2%
60-70 years
(n = 1119)
19%
27 + 4
18%
16%
1314
102
29%
20%
7%
> 70 years
(n = 739)
26%
25 + 4
27%
26%
1826
114
25%
27%
12%
All these variables were significantly different between the different age subgroups
Cleland et al., Cardiovasc Drugs Ther 2004; 18:139

% of patients
70
60
50
40
30
20
10
0
Clinical Characteristics of the Patients with HF
According to Age: IMPROVEMENT Study
8256 patients in 15 countries, 1999-2000
62
Male sex
p

% of patients
35
30
25
20
15
10
5
0
Concomitant Diseases of the Patients with HF
According to Age: IMPROVEMENT Study
8256 patients in 15 countries, 1999-2000
Renal dysfunction
p

Most Common Non-cardiac Diseases for Patients >
65 years with CHF: An analysis of 122 630 subjects
Hypertension
Diabetes
COPD
Ocular disorders
Hypercholest.
Atherosclerosis
Arthritis
COPD
Thyroid disorders
Complicated HBP
Alzheimer's
Depression
Renal failure
9
8
7
11
14
14
26
24
21
20
19
0 10 20 30 40 50 60
% of patients
31
55
Braunstein et al., J Am Coll Cardiol 2003;42:1226

Probability
Non-cardiac Comorbidity Increases Preventable
Hospitalizations among Medicare Beneficiaries with CHF
A survey of 122 630 Subjects Aged > 65 years
1
0,8
0,6
0,4
0,2
0
Any hospItalisation
ACSC
CHF ACSC
0 1 2 3 4 5 6 7 8 9 10 Overall
ACSC = Ambulatory Care Sensitive
(e.g. preventable) Condition Braunstein et al., JACC 2003;42:1226

Incidence of Precipitating Factors as Causes of
Hospital Admissions in Patients with Heart Failure
% Patients
40
30
20
10
0
Causes of hospitalizations
(179 pts.; age, 75+10 years)
Poor Inadequate
Hypertension
compliance treatment
Ischemia Arrhythmias
Michaelsen et al., Heart 1998;80:437

Management of HF in the Elderly:
General Measures
• Educate the patients and their family ↑ compliance!
• Limit sodium intake (< 1.5 Gm/day)
• Monitor body weight daily
• Avoid smoking
• Only moderate alcohol consumption (e.g. < 1-2 drinks/day)
• Low-intensity aerobic exercise (3-5 times/week)
• Treat concomitant diseases
– Hypertension, and/or diabetes, CAD…
– anemia, thyroid disease, sleep disorders, depression…
• Minimize use of NSAIDs

Effect of Antihypertensive Therapy in the
Primary Prevention of HF in Older Adults
Trial (year)
EWPHE (1985)
Coope (1986)
STOP-HTN (1991)
SHEP (1991)
STONE (1996)
Syst-Eur (1997)
Syst-China (1998)
No.
840
884
1627
4736
1632
4695
2934
Age, ys
>60
60-79
70-84
>60
60-79
>60
>60
Risk reduction
for HF
↓22%
↓ 32%
↓ 51%
↓ 55%
↓ 68%
↓ 36%
↓ 38%
Rich MW, Am J Med 2005; 118:342

Pharmacological therapy of heart failure due to Left
Ventricular Systolic Dysfunction
NYHA I reduce / stop diuretic
NYHA II
NYHA III
NYHA IV
For Survival/Morbidity
mandatory therapy
Cont. ACE inhibitor/ARB if ACE
inhibitor intolerant, continue
aldosterone antagonist if post-MI
add beta-blocker if post-MI
ACE inhibitor as first-line
treatment/ARB if ACE inhibitor
intolerant
add beta-blocker
and aldosterone antagonist if post MI
ACE inhibitor plus ARB or ARB
alone if ACE intolerant
beta- blocker
add aldosterone
antagonist
Continue ACE inhibitor/ARB
beta-blocker
Aldosterone antagonist
For Symptoms
+/- diuretic
depending on fluid
retention
+ diuretics + digitalis
If still symptomatic
+diuretics + digitalis
+ consider temporary
inotropic support
Swedberg, Eur Heart J, June 2005; 26: 1115 - 1140.

Drug
ACEi
Loop diuretics
Beta-blockers
Aldosterone
antagonists
ARBs
Digoxin
Treatment of HF in the Elderly:
What Do the Guidelines Say
Evidence
(+)
0
+
(+)
+
(+)
Efficacy
+
+
+
+
+
+
Side
effects
↑
↑
↑
↑↑
↑
↑
Tolerance
↓
↓
↓
↓
↓
↓

Drug
ACEi
Loop diuretics
Beta-blockers
Aldosterone
antagonists
ARBs
Digoxin
Treatment of HF in the Elderly:
What Do the Guidelines Say
Evidence
(+)
0
+
(+)
+
(+)
Efficacy
+
+
+
+
+
+
Side
effects
↑
↑
↑
↑↑
↑
↑
Tolerance
↓
↓
↓
↓
↓
↓

Drug
ACEi
Loop diuretics
Beta-blockers
Aldosterone
antagonists
ARBs
Digoxin
Treatment of HF in the Elderly:
What Do the Guidelines Say
Evidence
(+)
0
+
(+)
+
(+)
Efficacy
+
+
+
+
+
+
Side
effects
↑
↑
↑
↑↑
↑
↑
Tolerance
↓
↓
↓
↓
↓
↓

Elderly Age is an Independent Predictor of Lack of
Prescription of ACE Inhibitors: Results from the
EuroHeart Failure Survey
Age >70 ys.
Male gender
IHD
0.77
1.34
2.45
0 0.5 1 1.5 2 2.5 3 3.5
Odds ratio (95% CI)
Prescription less likely ← → Prescription more likely
Komajda et al. Eur Heart J 2003; 464-475

% of patients
100
80
60
40
20
0
ACEi/ARBs Prescriptions in the Patients with HF
According to Age: IMPROVEMENT Study
8256 patients in 15 countries, 1999-2000
69
Males
p=0.002
72
67
57
84
Years
% of patients
100
80
60
40
20
0
Females
p=0.090
68 71 68 65
84
Years
Muntwyler et al., Eur J Heart Fail 2004; 6:663

Effects of ACE inhibitors on outcome in patients
subdivided on the basis of their age: Meta-analysis of
the SOLVD, SAVE, AIRE and TRACE trials
Age
(years)
Death/CHF/MI
75
0.4 0.6 0.8 1 1.2 1.4
Risk ratio (95% CI)
No of events/No of patients
495/3165
994/4315
1227/4194
454/1066
878/3165
1534/4315
1761/4194
590/1066
Flather et al., Lancet 2000; 355:1575

Drug
ACEi
Loop diuretics
Beta-blockers
Treatment of HF in the Elderly:
What Do the Guidelines Say
Evidence
(+)
0
+
Efficacy
+
+
+
Side
effects
↑
↑
↑
Tolerance
↓
↓
↓

Elderly Age is an Independent Predictor of Lack of
Prescription of Beta-blockers: Results from the
EuroHeart Failure Survey
COPD
Cardiologist vs GIM
Ischemic Heart Disease
Age > 70 ys.
Male gender
0.35
0.55
1.16
2.69
2.63
0 0.5 1 1.5 2 2.5 3 3.5
Odds ratio (95% CI)
Prescription less likely ← → Prescription more likely
Komajda et al. Eur Heart J 2003; 464-475

% of patients
100
80
60
40
20
0
Beta-blockers Prescriptions in the Patients with HF
According to Age: IMPROVEMENT Study
8256 patients in 15 countries, 1999-2000
44
Males
p

BRING UP
Reasons for Not Commencing β-Blockade Blockade
Age 70 years
NYHA III-IV III IV
Pulmonary rales
Periph. Periph.
oedema
Atrial fibrillation
EF not available
EF < 30%
On
blockers
(n = 771)
24%
26%
10%
12%
14%
5%
30%
blockers
started
(n = 865)
25%
29%
20%
13%
17%
3%
30%
No
blockers
(n = 1455)
43%
42%
26%
21%
21%
6%
35%
P value
0.001
0.001
0.001
0.001
0.001
0.008
0.017
Maggioni A et al. Heart 2003

BRING UP
Independent Predictors of Blocker
Tolerability
Age
(as a continuous variable)
NYHA class
(III-IV (III IV v I-II) I II)
Systolic blood pressure
(as a continuous variable)
Heart rate
(as a continuous variable)
Ejection fraction
(not available v available)
OR
0.97
0.62
1.02
1.01
0.46
95% CI
0.96 - 0.97
0.51 - 0.75
1.01 - 1.02
1.01 - 1.02
0.28 - 0.76
P value
0.0001
0.0001
0.0001
0.0001
0.0022
Maggioni A et al. Heart 2003

%
100
75
50
25
0
84.3
70- 80 years
(n=297)
P

COLA II
Multivariate predictors of tolerability
Advanced age
Low LVEF
Low DBP
OAD present
Diabetes present
Courtesy of H. Krum
0.45
0.72
0.98
0.97
0.1 1.0
Tolerated worse
1.58
2.0
Tolerated better

%
Effect of Bisoprolol on Mortality in Different
Age Subgroups
30
25
20
15
10
5
0
Mortality
***
***
> 71 years < 71 years
Placebo Bisoprolol
< 71 years
> 71 years
0 0.2 0.4 0.6 0.8 1
Hazard ratio (95% CI)
Erdmann E et al. Eur J Heart Fail 2001; 3:469

Efficacy of Metoprolol CR/XL in Elderly Patients
with Heart Failure: the MERIT-HF Study
Total Mortality
65 years
Total Mortality/CV hosp
65 years
Total Mortality/WHF hosp
65 years
Cardiac death/non-fatal MI
65 years
0 0.2 0.4 0.6 0.8 1 1.2
Risk ratio (95% CI)
Deedwania E et al. Eur Heart J 2004; 25:1300

Beta-blockers Reduce Mortality also in Elderly Patients:
Meta-analysis of >12,000 Patients in Large-scale Trials
(BEST, Carvedilol US, CIBIS-II, COPERNICUS, MERIT-HF)
Mortality
Non elderly
(n=8112)
Elderly
(n=4617)
0.66
0.76
0.2 0.4 0.6 0.8 1
Risk ratio (95% CI)
Dulin, Haas, Abraham & Krum. Am J Cardiol 2005; 95:896

SENIORS Trial: Inclusion Criteria
• Age 70 years
• A clinical diagnosis of chronic heart failure (HF) and either
of:
a) documented LVEF 35% within previous 6 months
or
b) hospital admission within previous 1 year for
congestive HF
• Written consent prior to enrolment into the study

SENIORS: Baseline Characteristics
Age (mean, yrs)
Male (n, %)
IV 19 (1.8) 24 (2.3)
* 7 patients with missing LVEF at baseline
Nebivolol
76.1
657 (61.6)
Placebo
76.1
686 (64.7)
LVEF 35 % (n, %) * 683 (64.3) 686 (64.8)
LVEF (mean, %) 36.0 36.0
NYHA Class (n, %)
I 32 (3.0) 29 (2.7)
II 603 (56.5) 597 (56.3)
III 413 (38.7) 411 (38.7)

Event free
Survival %
100
90
80
70
60
50
SENIORS Trial: All Cause Mortality or CV Hospital
Admission (Primary Outcome)
Nebivolol
Placebo
Hazard Ratio 0.86 [0.74;0.99]
p = 0.039
0 6 12 18 24 30 36
Time (months)
No. of events: Nebivolol 332 (31.1%); Placebo 375 (35.3%)

LVEF
35 %
> 35 %
Age
70-75 y
> 75 y
Total
Death or CV Hospitalisation by Subgroup
Nebivolol
219 (32.1%)
110 (28.9%)
148 (27.5%)
184 (34.8%)
332 (31.1%)
Placebo
249 (36.3%)
125 (33.6%)
176 (33.5%)
199 (37.1%)
375 (35.3%)
Favours
Nebivolol
Hazard ratio and 95% CI
Favours
Placebo
0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20

Drug
ACEi
Loop diuretics
Beta-blockers
Aldosterone
antagonists
ARBs
Digoxin
Treatment of HF in the Elderly:
What Do the Guidelines Say
Evidence
(+)
0
+
(+)
+
(+)
Efficacy
+
+
+
+
+
+
Side
effects
↑
↑
↑
↑↑
↑
↑
Tolerance
↓
↓
↓
↓
↓
↓

Elderly Age is an Independent Predictor of Lack of
Prescription of Spironolactone: Results from the
EuroHeart Failure Survey
Age > 70 ys.
Cardiologist vs GIM
Ischemic Heart Disease
SVT/AF
Male gender
…
0.76
0.78
1.39
1.28
1.31
0 0.5 1 1.5 2 2.5 3 3.5
Odds ratio (95% CI)
Prescription less likely ← → Prescription more likely
Komajda et al. Eur Heart J 2003; 464-475

Spironolactone-induced hyperkaliemia and renal
insufficiency in patients with heart failure
Variable
Serum potassium (baseline)
Age
Beta-blocker use
Predictors of change in serum creatinine
Thiazide diuretic
P value
Predictors of changes in serum potassium

Drug
ACEi
Loop diuretics
Beta-blockers
Aldosterone
antagonists
ARBs
Digoxin
Treatment of HF in the Elderly:
What Do the Guidelines Say
Evidence
(+)
0
+
(+)
+
(+)
Efficacy
+
+
+
+
+
+
Side
effects
↑
↑
↑
↑↑
↑
↑
Tolerance
↓
↓
↓
↓
↓
↓

Effects of Losartan Compared with Captopril on
Mortality in Patients with Symptomatic Heart Failure:
ELITE II Results
Probability of survival
1
0,8
0,6
0,4
0,2
Losartan
Captopril
Losartan, 280/1578 (17.7%)
Captopril, 250/1574 (15.9%)
P=0.16
Total mortality
0
0 100 200 300 400 500 600 700
Follow-up (days)
(Pitt et al., Lancet 2000; 355:1582)

10
5
0
-5
-10
-15
-20
-25
-30
-35
Lack of Influence of Age on the Effects of
Valsartan in Val-HeFT Trial
Risk Ratio Valsartan vs. Placebo
Primary end-points Secondary end-points
All-cause
death
1st morbid
event CV death HF hosp.
P=0.07 P=0.09
> 65 years < 65 years
P=0.012
P

Drug
ACEi
Loop diuretics
Beta-blockers
Aldosterone
antagonists
ARBs
Digoxin
Treatment of HF in the Elderly:
What Do the Guidelines Say
Evidence
(+)
0
+
(+)
+
(+)
Efficacy
+
+
+
+
+
+
Side
effects
↑
↑
↑
↑↑
↑
↑
Tolerance
↓
↓
↓
↓
↓
↓

%
40
30
20
10
Lack of effect of age on the effects on digoxin on
outcome: Results from the DIG study
HF hospitalizations
P=0.0001
Digoxin
Placebo
80
Age (years)
%
40
30
20
10
HF mortality or HF
hospitalizations
P=0.0001
80
Age (years)
Rich et al. J Am Coll Cardiol 2001; 38:806

% of patients
50
40
30
20
10
Oral Anticoagulant Prescriptions in the Patients with HF
According to Age: IMPROVEMENT Study
8256 patients in 15 countries, 1999-2000
0
19
Males
p

Randomised Controlled Trial of Cardiac
Rehabilitation in Elderly Patients with HF
Age, m+SD
6-min walk
distance, mts
MLHF score
NYHA class
Total hospital
admissions
Days in hospital
Deaths
Baseline
72 + 7
259
44
2.53
Control
24 weeks
252
37
2.48
33
187
4
Rehabilitation
Baseline
72 + 6
275
41
2.21
24
weeks
320***
23**
2.01***
11
41
5

Effects of a Multidisciplinary Intervention to Prevent the
Readmission of Patients with Heart Failure
(Washington University HF disease management trial)
• Study group
– 282 patients, selected among 1306
– Mean age 79 ys; female, 65%
– NYHA: 2.4 +1.1; EF: 41+13%
• Randomization
– Intervention vs control groups
• Personnel
– Nurse, dietician, social assistant,
geriatrician, cardiologist
• Follow-up
– 90 days
• Results
– n.s. mortality
– hospitalizations
– costs ( $153/month/pt.)
– compliance and quality of life
60
50
40
30
20
10
0
Control
Intervention
n.s.
Mortality
% patients Days/patient
0.03
> 1
0.01
> 2
18
16
14
12
10
8
6
4
2
Hospitalizations
0.04
0
Days/patient
Rich et al., NEJM 1995;333:1190

Randomized Controlled Trial of Specialist Nurse
Intervention in Heart Failure
• Study group
– 165 / 801 pts with an
– mean age: 75 ys, female, 49%
– Emergency admission for acute HF
– Echo LV systolic dysfunction
– NHYAII/III/IV, 43/76/81
• Randomization
– Intervention vs control groups
• Personnel
– Home visit by a nurse
– Telephone contacts as needed
• Methods
– Patient’s education
– Treatment optimization
– Laboratory monitoring
• Results
– Similar deaths (31% vs 30%)
– 39% death or HF hospitalization
– 28% death or all cause
hospitalizations
% event-free survival
100
90
80
70
60
50
40
30
20
10
0
Intervention
Usual care
P=0.033
0 2 4 6 8 10 12
Follow-up (months)
Blue, McMurray, et al. BMJ 2001;323:715

Effects of Disease Management Programs on
Hospitalization Rates in Heart Failure
Multidisciplinary specialized team
Rich, 1993
Rich, 1995
Cline
Stewart, 1998
Ekman, 1998
Serxner
Jaarsma, 1999
Stewart, 1999
Naylor, 1999
Primary care
Naylor, 1994
Weinberger
TOTAL
McAlister et al., Am J Med 2001; 110:378
0 0,5 1 1,5 2
Risk ratio (95% CI)
Subtotal:
RR: 0.77 (0.68-0.86)
Total:
RR: 0.87 (0.79-0.96)

Efficacy of Disease Management Programmes
in Reducing Hospital readmissions in Older
(>70 ys) Patients with HF
Randomised trials
CV or HF readmissions (n=3160)
All-cause readmissions (n=4440)
All-cause readmissions (n=3936)
(Weinberger excluded)
Readmission or death (n=2985)
0.7
0.85
0.82
0.88
0.4 0.5 0.6 0.7 0.8 0.9 1
Risk ratio (95% CI)
Gonseth et al., Eur Heart J 2004; 25:1570