Screening for cancer: are biomarkers of value?

– February/March 2011 22 Tumour markers
Role of PSA isoforms in improving
PSA specificity in asymptomatic men
An overview of the European Randomised Study of Screening for
Prostate Cancer (ERSPC) side studies indicates that biomarker panels
would enhance accuracy of PCa screening in asymptomatic men [1].
by Dr Chris Bangma
The diagnosis of prostate cancer (PCa)
continues to challenge both clinicians
and their laboratory colleagues. With
the incidence of PCa rising progressively
each year, it is now the most common
form of malignancy amongst men
in Europe. [2]. Prostate-specific antigen
(PSA) has been the main driver for early
detection of prostate cancer. Increasingly,
over the last decade, we have seen how
awareness of the PSA test has generated a
storm of diagnostic activity, particularly in
countries like the US, even in men with few
or no symptoms.
This tumour marker has attracted extraordinary
scientific attention regarding its
biochemical characteristics, epidemiological
and clinical value, alongside huge
commercial interest and increasing debate,
even in the lay press [3, 4]. The level of
PSA at initial screening is highly indicative
of PCa being diagnosed later in life
– however, approximately half of cancers
detected in this way are found to be indolent
[5]. It has always been recognised that
many prostate cancer patients died with,
rather than because of, their disease, but
it was not foreseen that the impact of PSA
screening of asymptomatic men would
result in such a large number of indolent
(slow growing) cancers being detected.
The ERSPC believes that the benefits of
screening would be significantly increased
by the introduction of new biomarkers
such as PSA isoforms [Figure 1], and
Complexed PSA
Total PSA
Other
inactive
PSA
Truncated
proPSA
Free PSA
Figure 1. SA and PSA Isoforms – overview.
Adapted by Chris Bangma from original table
published in European Journal of Cancer, Oct
2010 [1].
BPSA
the subsequent development of a panel of
screening tests. The additional analysis it
carried out of candidate markers provides
early encouragement. So far, the most
promising results have been obtained from
the analysis of free PSA, proPSA, nicked
PSA and hK2 [6]. The availability of a
panel of such markers could increase the
specificity and sensitivity of screening for
prostate cancer, enabling new strategies to
be developed for monitoring and asessing
patient risk. This would reduce unnecessary
biopsies, with surgical intervention
only offered to men at the highest risk.
The first breakthrough in assessing the
effectiveness of PSA testing on PCa mortality
assessment came in March 2009 with the
publication of the ERSPC study (involving
an overall follow-up of up to 12 years) [7].
It reported an initial reduction of at least
20% in mortality, which rose to approximately
30% when adjustment was made for
non-attendance of men enrolled in the
study [8].
Solving the dilemma of
overdiagnosis
The likelihood of being diagnosed with prostate
cancer in later life is highly dependent
on the level of PSA at initial screening. PSA
increases more quickly over time in men who
get detectable cancer. However, the ERSPC
study highlighted the serious limitations of
current PSA testing – that of overdiagnosis.
In order to save one life, 1410 men would
have to be tested and 48 men diagnosed with
prostate cancer. To achieve this, an unacceptably
high number of biopsies would
subsequently be found to be negative.
Current clinical practice uses a serum PSA
value of more than 4 ng/mL to indicate
abnormality; this is likely to lead to a prostatic
biopsy. The case is less clear cut with
men with a serum PSA between 2-4ng/mL.
After undergoing a biopsy, 20 -30% of these
men will be seen to have PCa, and of those,
approximately one third of these cancers
are indolent and unlikely to cause death [9].
Many patients have therefore undergone an
unnecessary, invasive procedure– at a cost
to both the men themselves and the health
service involved. Based on the annual incidence,
it is estimated that the annual number
of unnecessary biopsies is around 750,000 in
the US alone [10, 11].
Throughout the duration of the study, the
eight ERSPC countries involved used total
PSA value as the sole indicator for biopsies.
However, the study involved such a large
sample (182,000 participants initially) that it
was possible to carry out a number of additional
country-specific studies specifically
focusing on different candidate markers.
Alongside evaluating changes in PSA concentrations
over time (PSA velocity, PSA
doubling), various PSA-isoforms, kallikreins
and molecular markers were assessed retrospectively
in different ERSPC cohorts of men
both with and without prostate cancer. The
objective was to see what additional information
these candidate markers could add to
the total PSA result hopefully enabling more
selective PCa screening models to be developed
for asymptomatic men.
PSA is a kallikrein-like serine protease (also
known as kallikrein-related peptidase 3),
which is cleaved into smaller fragments as
part of its biological profile. Family members
include kallikrein-related peptidase 2 (hK2),
which is genetically and structurally very
similar to PSA and also primarily localised to
the epithelial cells of the prostate. hK2 rapidly
converts the inactive precursor form of PSA,
proPSA, to active PSA and appears to be more
strongly associated with prostate tumours
than total PSA, being highly expressed in
poorly differentiated cancer cells.
Candidate markers —
a brief glimpse at the possibilities
From the side studies, the most promising
potential biomarkers included % free PSA
and free PSA forms (intact/nicked PSA
and proPSA) and hK2 [6]. ProPSA is found
in normal prostatic epithelium together
with truncated forms. Mikolajczyk et al.
identified a clinically important form, (-2)
proPSA, containing 239 amino acids. This
form is found in prostate cancer tissue as
well as serum [12]. Indications are that the