How to Use Prostate-Specific Antigen in the Carcinoma Introduction

C a n c e r
C l i n
1 9 9 5 ; 4 5 : 1 4 8 - 1 6 4
How to Use Prostate-Specific Antigen in the
Early Detection or Screening for Prostatic
Michael K. Brawer, MD
Prostate cancer represents the most common male malignancy and the second
most common cause of cancer-related
mortality in American men. In 1995, it is
estimated that 244,000 men will be diagnosed with prostate cancer and 40,400
will succumb from this malignancy.1 Despite these sobering statistics, incidence
and mortality appear to be increasing.
While the autopsy prevalence of prostate
cancer far exceeds clinically manifest disease, Seidman et al2 and Scardino et al3
have demonstrated that of the 42 percent
of men older than 50 years harboring prostatic carcinoma, 9.5 percent will have a
clinical diagnosis, and 2.9 percent will
Of the three possibilities to reduce
cancer-related mortality—increased early detection, improved therapy, and reduced incidence—only increased early
detection appears feasible at present. A
major effort is under way to reduce cancer incidence in many organ systems with
chemopreventive approaches. For prostate cancer, a large-scale, multicenter trial
randomizing men to finasteride (a 5- reductase inhibitor) or placebo, in a
chemopreventive approach, is under
way.4 However, it will be many years be-
Dr. Brawer is a Professor in the Department of
Urology at the University of Washington and Chief
of the Section of Urology at the Seattle Veteran’s
Administration Medical Center.
fore the outcome of this investigation is
known. While significant strides have
been made in lessening the morbidity of
therapy directed with curative intent
(primarily radiation therapy and radical
prostatectomy), the fact that mortality in
men presenting with advanced disease
has not changed significantly in the past
several decades suggests that either most
men present with noncurable malignancy
or our therapeutic armamentarium is inadequate. As a result significant efforts
have been made to identify more men
with curable cancer.
The recognition that serum prostate-specific antigen (PSA) levels are elevated in
most men with clinically diagnosed
prostate cancer served as an impetus to
investigate the possible role of this analyte for early detection or screening. Considerable evidence suggested, however,
that this approach would not be effective,
as many reports indicated a significant elevation of PSA levels in men with benign
prostatic hyperplasia (BPH) (Table 1).5-8
As BPH is an almost universal finding in
men of an age group likely to be tested
for prostate cancer, early investigators
believed that there was no role for PSA in
early detection or screening.
One piece of evidence suggesting
a potential fallacy in this argument
stemmed from the observation that the
serum PSA level in the seminal plasma is
Ca—A cancer Journal for Clinicians
C a n c e r
C l i n
1 9 9 5 ; 4 5 : 1 4 8 - 1 6 4
Table 1
Serum PSA in Patients with Histologically Confirmed
Benign Prostatic Hyperplasia
Patients with
Patients with
Stamey et al5
Ercole et al6
Ferro et al7
Hudson et al8
PSA = prostate-specific antigen.
about a million-fold higher than the level
in the systemic circulation.9 Therefore,
there must be extraordinary barriers between the lumen of the prostatic acini and
ductule and the systemic circulation. Employing a variety of immunohistochemical and other techniques, these barriers
can be readily identified. Figure 1 demonstrates a schematic of these barriers,
which include the basal cell layer, the prostatic basement membrane, intervening
stroma, the capillary basement membrane, and the capillary endothelial cell.
In an effort to understand how PSA
arrived in the systemic circulation, we
studied a series of prostatectomy specimens from men undergoing simple
prostatectomy (transurethral resection or
simple open) for presumed BPH.10 We
observed that among the 35 patients with
an elevated serum PSA level (>4.0 ng/ml,
preoperative, Hybritech Tandem assay),
all but one had incidental carcinoma, prostatic intraepithelial neoplasia (PIN), or
foci of acute inflammation. Of the 26 men
who had only BPH or BPH associated
with chronic inflammatory cell foci, only
one had an elevated PSA level. This suggested to us that changes in some of these
barriers were necessary before PSA
Vol. 45 No. 3 may/june 1995
could leak into the capillary bed. Disruption of these barriers has been identified
in both prostatic carcinoma as well as
PSA Screening
To evaluate the role of serum PSA in an
early detection or screening strategy, Catalona et al15,16 from Washington University at St. Louis and our own group at the
University of Washington17 conducted
screening in a media-recruited cohort of
men older than 50 years. After analysis of
PSA levels by the Hybritech Tandem assay, ultrasound-guided biopsies were performed in those with a PSA level greater
than 4.0 ng/ml (Table 2).18-21 With initial
biopsy, positive predictive values (PPV)
of 30.5 to 34.4 percent and detection rates
of 2.6 to 3.1 percent were realized. Compared with mammography, for which
PPVs of 20 percent have been realized,22,23 the application of this simple
serum assay becomes exceedingly attractive for early detection.
A number of additional authors
have reported relatively consistent PPVs
of 33 to 50 percent for Hybritech PSA assay greater than 4.0 ng/ml in disparate
H o w
t o
u s e
P r o s t a t e - S p e c i f i c
Prostatic Lumenal Cell
Capillary Basement Membrane,
Endothelial Cell
Fig. 1. Schematic of the histology of prostatic
carcinoma. Note the significant tissue barriers
between the prostatic lumen and the surrounding capillary bed.
screening, referral, and mixed populations (Table 3).15-17,19,24-29
Importance of the Digital Rectal
Exam (DRE)
Despite the impressive yield for PSAbased screening alone, it is widely recognized that a significant number of men
will have prostatic carcinoma and a PSA
level less than 4.0 ng/ml. In our ultrasound-guided biopsy series, in which all
men underwent six systematic sector
biopsies, 54 (21.3 percent) of the men
with cancer had a PSA level less than 4.0
Catalona et al18 evaluated 6,630 men
with DRE and serum PSA. Two hundred
sixty-four cancers were identified. Fortyeight of the carcinomas (18.2 percent)
were in men with an abnormal DRE
alone. Thus, it is obvious that the PSA as150
A n t i g e n
say should not be used exclusively for the
detection of prostate cancer, but should
be combined with a carefully performed
The American Cancer Society recently revised its recommendation for the
annual cancer prevention check-up to include DRE and serum PSA assay for
men older than 50 years or for younger
men who are at increased risk owing to
being African American or having a significant family history.31 The Food and
Drug Administration has recently approved the Hybritech PSA assay for early
detection in conjunction with DRE.
Several caveats should be mentioned with regard to obtaining serum for
PSA testing. A number of factors have a
significant effect on serum PSA level, as
shown in Table 4. In general PSA should
be obtained in an ambulatory setting before significant prostatic manipulation.
Standard rectal examination has been
shown by several investigators to not significantly elevate the serum PSA level.30,32-34 However, more significant prostatic trauma or alteration of the hormonal
milieu will cause considerable alteration
of PSA level.
Enhancing the Specificity
of PSA Testing
Considerable efforts are being made to
improve the performance of PSA testing.
Figure 2 shows the consistent inverse relationship between sensitivity (the chance
of a test being positive when the patient
has the disease) and specificity (the
chance of a test being negative when the
patient does not have the disease). The
data depicted derive from 1,920 men undergoing systematic sector biopsy of their
prostates from our institutions.
For prostate cancer most efforts are
directed toward increasing specificity. This
stems from the likelihood that men are not
going to be tested only once in their lifetime, but will undergo serial testing, perhaps annually as suggested by the AmeriCa—A cancer Journal for Clinicians
C a n c e r
C l i n
1 9 9 5 ; 4 5 : 1 4 8 - 1 6 4
Table 2
Yield from PSA-Based Screening Studies
Number of
Detection Rate
Detection Rate
Catalona et al16
Catalona et al (Serial)16
Catalona et al
Brawer and Lange19
Brawer et al (Serial)20
Brawer et al (Serial)21
PSA = prostate-specific antigen.
can Cancer Society.31 Thus, a false-negative test, which might occur with a lesssensitive assay, is likely to be of less significance. The test result may become positive
while the malignancy is still curable.
In contrast, false-positive tests result
in a large burden in terms of increased expenditures for subsequent, unnecessary
medical procedures and increased anxiety for misdiagnosed patients. In prostate
cancer, an abnormality in PSA level or
DRE mandates transrectal ultrasound
and ultrasound-guided biopsy according
to most experts. A strategy to reduce
false positives necessitates tests that have
increased specificity.
Current approaches to improve the
specificity of PSA-based screening include PSA velocity, PSA density, agespecific PSA values, and investigation of
different circulating forms of the PSA
Carter et al35,36 reported in 1992 that an
annual increase of 0.75 ng/ml in serum
PSA level indicated men who would deVol. 45 No. 3 may/june 1995
velop prostatic carcinoma. They based
their observations on sera collected as
part of the Baltimore Longitudinal Aging
Study. Serum specimens were collected
from men over many years in a general
study of the phenomena of aging. The optimum cutoff of sensitivity and specificity
to predict which of the men ultimately developed prostatic carcinoma was a PSA
velocity of 0.75 ng/ml per year. It should
be noted that in this study a minimum of
seven years passed between determinations of PSA levels.
We were unaware of the usefulness
of this rate when we developed our strategy for serial follow-up of men with an
initial normal PSA level in our screening
study.20 We arbitrarily selected an annual
increase of 20 percent over the baseline in
subsequent years—a number not dissimilar to the observation of Schmid et al37 for
PSA doubling time in untreated prostatic
carcinoma. PPVs in the second and third
year of the series were 17.1 percent and
18.6 percent with observed detection
rates of 2.0 percent and 1.8 percent, respectively. If the PPV remained the same,
estimated detection rates for the entire
H o w
t o
u s e
P r o s t a t e - S p e c i f i c
A n t i g e n
Table 3
Positive Predictive Value for PSA Greater Than 4.0 ng/ml
Babaian and Camps24
Bazinet et al25
No. of
Positive Predictive
Value (percent)
Brawer and Lange
Brawer et al17
Catalona et al15
Catalona etal16
Catalona et al
Cooner et al26
Cooner et al27
Mettlin et al28
Rommel et al
PSA = prostate-specific antigen.
cohort including those men not returning
for evaluation would have been 6.7 percent and 3.8 percent in the second and
third year, respectively.
Catalona et al16 used a crossover to
4.0 ng/ml from an initial lower value and
noted a detection rate of 2.1 percent and
PPV of 41.9 percent. Recently, Porter et
al38 reported on an enlarged series from
our institution with one- and two-year intervals between determination of PSA
levels. With about a one-year interval between PSA-level determinations, no PSA
velocity parameter (including median
PSA velocity, median percent PSA increase per year, 0.75 ng/ml per year, or an
increase of 20 percent) was useful in selecting men with prostatic carcinoma.
Moreover, in a smaller series with at least
a two-year interval between PSA levels,
we were still unable to use PSA velocity
to separate men with or without carcinoma using any manipulation of PSA.
Littrup et al39 have reported on evi152
dence that greater intervals between determination of PSA levels may enhance
cancer detection. Analyzing the results of
the American Cancer Society National
Prostate Cancer Detection Project, they
observed that PSA velocity greater than
1.0 ng/ml per year predicted cancer. Percent change in PSA was not useful.
The discrepancy between these observations may merely be a reflection of
biologic variation masking significant
change during short-term follow-up. This
might be attenuated with longer intervals
between PSA determinations. Pearson
and Carter,36 in a review of their Baltimore Longitudinal Aging study, demonstrated that in men with prostatic carcinoma, there was a transition from a linear to
an exponential phase of PSA velocity beginning 7.3 years before diagnosis in
those men with local regional disease and
9.2 years before diagnosis in men with
more-advanced malignancy. Komatsu et
al40 have recently demonstrated signifiCa—A cancer Journal for Clinicians
C a n c e r
C l i n
1 9 9 5 ; 4 5 : 1 4 8 - 1 6 4
Table 4
Factors Affecting Serum Prostate-Specific Antigen Level
Digital rectal examination
No effect
Prostate biopsy
Prostatic massage
Prostate ultrasound
Reduction in androgen activity
Sexual activity
Urethral instrumentation
PSA Cutoff (ng/ml)
Fig. 2. The inverse relationship of prostate-specific antigen (PSA) sensitivity and specificity for
men undergoing ultrasound-guided prostate needle biopsy.
Vol. 45 No. 3 may/june 1995
H o w
t o
u s e
P r o s t a t e - S p e c i f i c
A n t i g e n
Fig. 3. Receiver operating characteristic curve of prostate-specific antigen (PSA) versus prostatespecific antigen density (PSAD). Note no enhancement of PSAD in the performance over PSA
alone. (Modified with permission from Brawer et al.44)
cant biologic variation in serum PSA levels. When specimens were drawn 15 to 183
days (mean 80 days) apart, 36.5 percent of
the patients showed an increase of more
than 20 percent, and 10 percent of patients
showed an increase of more than 0.75
ng/ml. I do not believe that current data
support using PSA velocity with the relatively short intervals between tests. Rather
a PSA level greater than 4.0 ng/ml should
be the indication for further evaluation.
Another approach to enhancing the
specificity of the serum PSA assay is PSA
density (also known as PSA index). In
this analysis, the idea is to adjust for the
contribution of PSA from BPH by dividing the serum PSA level by the volume of
the prostate. Results with this technique
were published by Benson et al.41,42
Stamey et al5 had previously demonstrated the about 10-fold increase in PSA level
arising from carcinomas compared with
the PSA level for BPH, which provides a
basis for PSA density. Furthermore,
Babaian et al43 demonstrated clearly the
relationship of prostatic volume to serum
PSA levels. Benson et al made a significant contribution when they noted highly
significant stratification of men with and
without carcinoma by calculating PSA
density, initially by measuring prostate
volume with magnetic resonance imaging41 and subsequently by using transrecCa—A cancer Journal for Clinicians
C a n c e r
C l i n
1 9 9 5 ; 4 5 : 1 4 8 - 1 6 4
Table 5
Studies of Prostate-Specific Antigen Density
No. of
Volume (cc)*
Bazinet et al25,46
21.4 (29.6)†
9.1 (8.1)
37.6 (21.4)†
51.6 (27.3)
0.63 (0.86)†
0.21 (0.25)
Benson et al41
7.0 (1.7)†
6.8 (1.8)
28.9 (14.6)†
40.1 (20.2)
0.30 (0.15)†
0.21 (0.11)
Rommel et al29,47
15.5 (21.6)†
4.9 (4.7)
42.7 (27.2)†
47.0 (31.6)
0.47 (0.11)†
0.105 (0.09)
Brawer et al44
10.4 (11.7)†
5.2 (5.0)
40.5 (16.6)
42.6 (25.6)
0.29 (0.41)†
0.14 (0.14)
Mettlin et al48
12.0 (16.0)†
2.1 (2.3)
38.9 (16.4)†
33.5 (14.2)
0.35 (0.5)†
0.08 (0.09)
Antigen Density*
*Standard deviation indicated in parentheses.
PSA: 8.0 ng/ml
Volume: 80 cc
PSAD: 0.1
PSA: 8.0 ng/ml
Volume: 40 cc
PSAD: 0.2
Fig. 4. Sampling concerns with prostate-specific antigen density (PSAD). Note that the carcinoma in the larger prostate (with a lower
PSAD) is more likely to be missed
tal ultrasound.42
We were interested in these observations and attempted to replicate the results.44 Unfortunately, we were unable to
match the performance of the reports by
Benson et al (Fig. 3). The PSA index
Vol. 45 No. 3 may/june 1995
(density) provided no increased utility
over PSA alone.
A number of possibilities exist for
the discrepancy between these studies.
Certainly, patient mix, differing ultrasound volume techniques, PSA assay
variability, as well as statistical analysis all
might contribute. Another factor is sampling (Fig. 4). If one assumes two men to
have an equivalent PSA level but widely
disparate gland volumes, a man with a
smaller prostate will have a higher PSA
density. Moreover, if each man has an
equal-volume carcinoma that is both
isoechoic (nonvisible on ultrasound) as
well as nonpalpable (T1c carcinoma),
then because of sampling considerations,
it is more likely that the cancer in the man
with the smaller gland will be identified.
Although the largest contribution to the
volume of the prostate gland is the transition zone where the minority of carcinomas arise, associated compression of the
peripheral zone and, most importantly,
lateral displacement may make sampling
H o w
t o
u s e
P r o s t a t e - S p e c i f i c
A n t i g e n
Table 6
Further Evaluation of Prostate-Specific Antigen Density
PSA Range
No. of Carcinomas
/No. of Patients
Mean PSA
Mean Volume (cc)
4.0 < PSA <10.0
0.2 - 220.0
PSA = prostate-specific antigen; CAP = carcinoma present; Ben = benign.
of the peripheral-zone neoplasm more
difficult.45 This suggests that PSA density
enhancement in predicting cancer may be
Table 5 summarizes five reports
from the literature in which PSA density
has been investigated. The reports by
Bazinet et al,25,46 Benson et al,41 and
Rommel et al29,47 demonstrate increased
stratification with PSA density. However,
in each study, the glands harboring malignancy were smaller than those without
disease in a statistically significant number. In our study44 as well as a study by
Mettlin et al,48 there was either no difference in the gland volumes or, in the case
of the experience of Mettlin et al, the cancer glands were actually larger.
Recently, we expanded our experience with PSA density, examining a series of 665 men undergoing systematic,
ultrasound-guided sector biopsy. As is
shown in Table 6, 240 men had carcinoma
detected. In this series, PSA density, unlike PSA, was useful in stratifying those
men with carcinoma in the PSA range of
4.0 to 10.0 ng/ml—the most important
subgroup for this experience. Again,
however, a statistically significant number
of men with carcinoma had smaller pros156
tates than those without. Until these issues
are resolved, I believe PSA density determination should not be used as the primary determinant for biopsy. Littrup et al39
have suggested that density determination
may have a role in a sequential decision
analysis approach to biopsy decision.
Recently, considerable attention has been
directed to the idea of age-specific serum
PSA cutoff values. This concept stems
from the observation that PSA levels increase as men get older, suggesting that
the arbitrary use of any specific cutoff for
PSA level in men of all ages may be inappropriate. It should be noted that the use
of 4.0 ng/ml as the upper limit of normal
for the Hybritech assay was based on the
observation that in men apparently free
of prostatic disease, this value represented the 95 percent confidence interval.49
Oesterling et al50 and Dalkin et al51
noted that this definition was probably inappropriate for the upper limit of normal.
They used very well-characterized patient populations apparently free of carcinoma (as evidenced by either PSA level
less than 4.0 ng/ml and a normal DRE
Ca—A cancer Journal for Clinicians
C a n c e r
C l i n
1 9 9 5 ; 4 5 : 1 4 8 - 1 6 4
Positive Predictive Value
PSA Cutoff (ng/ml)
Fig. 5. Positive predictive value for prostate-specific antigen (PSA) cutoffs of 2.5, 3.5, 4.0, 4.5,
and 6.5 ng/ml and different age groups in men undergoing ultrasound-guided prostate needle
biopsy. Note increasing yield with advancing age along with higher PSA cutoff levels.
and ultrasound or negative biopsy following abnormality on any of these tests) to
define what they termed “age-specific
cutoffs.” They reported the 95th percentile for each decade.50,51
Oesterling et al50 defined the agespecific reference range for serum PSA to
be 0.0 to 2.5 ng/ml for men aged 40 to 49
years, 0.0 to 3.5 ng/ml for men aged 50 to
59 years, 0.0 to 4.5 for men aged 60 to 69
years, and 0.0 to 6.5 for men aged 70 to 79
years. Furthermore, these authors concluded that using such age-specific cutoffs
would result in improved sensitivity for
younger men and improved specificity for
older men.
These observations are obvious.
Lowering the PSA cutoff will certainly allow an increased detection rate at the cost
of decreasing PPV, and increasing the
PSA cutoff will have the opposite effect
(Fig. 2). Figure 5 demonstrates this pheVol. 45 No. 3 may/june 1995
nomenon, using various age-specific cutoffs as suggested by Oesterling et al.50
Note that in addition to increase in PPV
with increasing PSA cutoffs, there is a
concomitant increase in cancer found as
men age—owing to the increased prevalence of prostate cancer.
Mettlin et al52 observed that among
156 cancers detected in the American
Cancer Society National Prostate Cancer
Detection Project, 35.3 percent of the patients had normal age-specific PSA cutoffs, whereas 64.7 percent of patients had
elevations. In contrast 27.5 percent of the
men with cancer had a PSA level less
than 4.0 ng/ml. This is an important cohort in which to study the efficacy of agespecific PSA cutoffs, owing to the fact
that PSA level was not an indication for
biopsy. It should be emphasized that in
this study the minimum age for entry was
55 years, thus potential enhanced sensi157
H o w
t o
u s e
P r o s t a t e - S p e c i f i c
Major Forms of PSA
PSA Complex
Free PSA
PSA Complex
Fig. 6. Schematic of the major forms of
prostate-specific antigen (PSA) in the systemic
circulation. Note that 1-antichymotrypsin may
mask three of five PSA epitopes, and all currently identified epitopes are rendered nonvisible by conventional immunoassays in the 2macroglobulin complexed form.
Cancer Detected (percent)
PSA Cutoff (ng/ml)
Fig. 7. The percentage of cancers detected at
cutoffs of 4.0, 5.0, and 6.0 ng/ml using
Hybritech Tandem and Abbott IMX assays for
men with known prostate cancer. Note stepwise reduction in the yield at each level where
the IMX is used.
tivity with a PSA cutoff less than 4.0
ng/ml based on the age-specific cutoff
could not be ascertained.
To study the impact of the effect of
A n t i g e n
age-specific cutoffs in men subjected to
early detection and screening, we compared the yield of our ultrasound-guided
needle biopsy with the cutoff of 4.0 ng/ml
for all ages and the age-specific cutoffs
from our screening cohort.53 The PPV we
applied is derived from men undergoing
ultrasound-guided biopsy for any indication, including an abnormality in PSA
(Fig. 5). Our screening study is based on
men aged 50 years and older; therefore,
the yield in a younger cohort cannot be
assessed. As anticipated, there is an agespecific cutoff enhancement in the PPV.
However, a reduction in the number of
cancers that would be detected in the
screening population was found. The detection rate was 5.3 percent using 4.0
ng/ml and 4.1 percent using age-specific
In an attempt to estimate the impact
of this difference in a population of
screened men, we analyzed the potential
years saved by applying the standard US
life table estimates. To give age-specific
cutoffs all the benefit, in our simplistic
model, we assumed that all cancers detected would be cured, that the men tested have a normal life expectancy, and
that the life expectancy for each man in
an age group was the same (i.e., a 59year-old man was assumed to have the
life expectancy of a 50-year-old man).
Moreover, we assumed that there was no
treatment-related mortality and that
missing a carcinoma would have no deleterious effect on our artificial society.
Based on 1,208 men from our
screening population, we estimated that
for this population 1,042 life years would
be saved if 4.0 ng/ml was selected as a cutoff and 856 life years would be saved if
age-specific cutoffs were used.
Other authors have also investigated
age-specific cutoffs.52,54 Catalona et al54
noted that the optimum balance of sensitivity and specificity for men older than 70
years was 5.0 ng/ml, as opposed to 4.0
ng/ml. However, they cautioned that widespread recognition of increasing unfavorCa—A cancer Journal for Clinicians
C a n c e r
C l i n
1 9 9 5 ; 4 5 : 1 4 8 - 1 6 4
Table 7
Prostate Markers of Malignant Potential
Clinical stage
Pathologic stage
Tumor volume
Prostatic acid phosphatase
Prostate-specific antigen
DNA ploidy
Nuclear morphometry
Tumor suppresser genes
Invasion markers (cathepsin, collagenase)
Basement membrane (collagen)
Growth factors
able pathology with elevation in serum
PSA makes the cutoff unwarranted.
It has been suggested that the increased specificity (and resulting decrease in sensitivity) associated with increasing PSA cutoffs is nevertheless
reasonable in older men who may derive
little benefit from early detection of prostatic carcinoma. It is my contention that
it is more reasonable to identify this
quandary prior to testing and actually not
perform screening in such patients rather
than apply a less sensitive cutoff to avoid
the dilemma of what to do with a man
with a more-limited life expectancy in
whom prostate cancer is detected. Obviously, an absolute upper age limit for
prostate cancer screening or early detection is inappropriate. However, most authorities would limit this in asymptomatic
men to those with a greater than 10-year
life expectancy.
Vol. 45 No. 3 may/june 1995
Another potential approach to enhancement of the specificity of serum PSA assays is the recognition that PSA circulates
in at least three molecular forms—free
PSA; PSA complexed with 2-macroglobulin; and the major form, PSA complexed with 1-antichymotrypsin.55-58 Figure 6 shows these three forms and the
masking of the epitopes recognized by
the antibodies in conventional serum assays for PSA. Free PSA has five epitopes
available. In general free PSA represents
the minority of identifiable PSA in the
systemic circulation. PSA complexed to
1-antichymotrypsin constitutes roughly
90 percent of the identifiable PSA in the
systemic circulation. Note that three of
the epitopes potentially available on free
PSA are masked by this protein. Finally,
PSA that is complexed to 2-macroglobulin has no epitopes available to the anti159
H o w
t o
u s e
P r o s t a t e - S p e c i f i c
bodies employed in current assays.
The recognition that free PSA
makes up a greater proportion of the total serum PSA in men without prostate
cancer has generated considerable enthusiasm. Lilja et al59 studied 89 men and
measured the ratio of free PSA to total
PSA or free PSA to complexed PSA
measured by an investigational assay at
cutoffs giving a sensitivity of 90 percent.
They observed that total PSA revealed a
specificity of 32 to 43 percent. In contrast
the specificity was 64 to 68 percent if the
ratio was used.
Stamey et al60 used gel chromatography from highly characterized patients
with and without carcinoma. They noted
that the total serum PSA recognized by
several commercial assays was 88 to 98
percent complexed with 1-antichymotrypsin in all patients with carcinoma.
Ten men with BPH had 73 to 84 percent
complexed with 1-antichymotrypsin.
Further work is obviously essential to
demonstrate whether increased specificity will be associated with examining the
ratio of free to total or free to complexed
PSA Assay Variability
There are currently six assays available
for PSA determination in the United
States, including Pros-Check PSA Assay
(Yang Laboratories, Belvue, Wash.); the
Tandem-E PSA, Tandem-R PSA, and
TOSOH assays (Hybritech Inc., San
Diego, Calif.); the IMX PSA Assay (Abbott Laboratories, North Chicago, Ill.);
and the recently improved ACS assay
(CIBA Corning, Norwood, Mass.). The
Pros-Check assay is a conventional polyclonal radioimmunoassay. The Hybritech
assays employ the Tandem monoclonalmonoclonal technology, either by a radioimmunoassay format (Tandem-R) or
immunoenzymatic format (Tandem-E).
The TOSOH assay employs the Hybritech monoclonal assays in an automated
format. The IMX and ACS assays em160
A n t i g e n
ploy polyclonal-monoclonal technology
with microparticle-capture enzyme immunoassay or chemolumenescent formats, respectively. In Europe and elsewhere, 30 or more assays are currently
Owing to the importance of knowing
whether there are differences between
assays, we have begun a series of comparisons.61,62 We examined 266 random sera
from our archival bank and ran three lots
of the Abbott IMX assay versus three lots
of the Hybritech Tandem-E assay according to the manufacturer’s specifications.
Each serum was assayed in all lots on the
same day with only one freeze/thaw. We
noted significant lot-to-lot variation with
the IMX assay.
Regression among the three Tandem-E lots had calculated slopes of 1.003,
1.033, and 1.037 (proportional bias of 0.3
to 3.7 percent). Regression of the results
from the three IMX lots demonstrated
between-lot slopes of 1.011, 1.098, and
1.109 (proportional bias of 1.1 to 10.9 percent).61 This indicates significant lot-to-lot
variability with the IMX assay.
Over the range of 2.0 to 10.0 ng/ml,
the overall bias was a 13 percent lower
reading with the IMX assay. Figure 7
demonstrates the percentage of men with
an established diagnosis of cancer who
exceeded the threshold for various cutoffs with the Tandem or IMX assay. For
each PSA cutoff, there was a stepwise reduction in yield with the IMX assay.
It is recognized that the IMX assay
preferentially identifies the free form of
PSA. For example, Stamey et al60 demonstrated that the IMX assay reads PSA
complexed with 1-antichymotrypsin at a
significantly lower value than the Tandem-R assay, but it detects the uncomplexed PSA at a higher value than the
Tandem-R assay. The authors concluded
that because 90 percent or more of the
serum PSA in cancer patients is in the
complexed form, the overall effect is a decreased value for the IMX with respect to
Ca—A cancer Journal for Clinicians
C a n c e r
C l i n
This may result from epitopic shielding as described above, making the free
PSA more readily identifiable by the antibodies employed by the IMX assay compared with the so-called equimolar response of the Tandem assay, where both
the free PSA and PSA complexed with
1-antichymotrypsin are equally identified. Moreover, the rapid-format technique of the IMX assay may preferentially allow higher signal from the free
compared with the complexed form of
PSA, owing to diffusion effects.
PSA Standardization
Recently, the Second Annual Stanford
PSA Standardization Conference was
held.63 Under the leadership of Dr.
Thomas Stamey, significant strides have
been made toward creating an international standard for serum PSA assays.
Given the rapid proliferation of available
commercial assays, this is obviously of
paramount importance. By providing a
standard calibrator determined by mass
weight of PSA, it is hoped that the various
manufacturer’s assays will be more directly comparable. It was decided at the conference that a standard calibrator with 10
percent free PSA and 90 percent complexed PSA would be most appropriate.63
The Stanford group has shown that
the use of this calibrator makes PSA assays of various manufacturers more comparable.63 While the 10 to 90 ratio may be
appropriate for many patients undergoing
PSA testing, we have observed a free
PSA range of six to 32 percent in men
with prostate cancer and a free PSA range
of six to 35 percent in men with negative
biopsies (unpublished observations). Obviously, further inquiry in this important
area is necessary and is under way.
The Need for Repeat Biopsy
One of the many unknowns surrounding
the diagnosis of prostate cancer is how to
evaluate the man with an abnormality on
Vol. 45 No. 3 may/june 1995
1 9 9 5 ; 4 5 : 1 4 8 - 1 6 4
DRE or an elevation in serum PSA who
has a negative biopsy. Little data exist to
offer guidelines on who should undergo
further evaluation.
We reported on 100 men who underwent repeat prostate needle biopsy after
initial negative biopsy.64 Carcinoma was
detected in 20 men. Cancer was found in
four of 17 (23.5 percent) who had atypia
on the initial biopsy, five of 14 (35.7 percent) who had PIN, and 10 of 69 (14.5
percent) who had neither of these suspicious findings on the initial biopsy. Unfortunately, PSA levels or PSA velocity
did not offer statistically significant stratification of who had carcinoma on the repeat biopsy. Clearly, these data indicate
that men with an initial biopsy revealing
PIN or atypia should undergo repeat
biopsy. In addition, despite lack of definitive evidence, a man with a rapidly rising
PSA, or a grossly abnormal DRE, and a
benign biopsy may well be a candidate
for repeat biopsy.
Screening Issues
Considerable controversy surrounds the
entire issue of screening or early detection of prostatic carcinoma. Certainly
valid arguments against screening can be
made, based on scientific issues (e.g.,
length- and lead-time bias and the problem of overdetection) and ethical issues
(e.g., commitment of significant resources
that might be better used elsewhere), as
well as legal concerns.
Feightner65 reviewed the recommendation of the Canadian Task Force on the
Periodic Health Examination against the
use of PSA for population-based screening. He stated that given the absence of
proof of definitive therapy, including radiation therapy and radical prostatectomy, in controlling or decreasing prostate
cancer-specific mortality, widespread
screening cannot be recommended.
Woolf66 analyzed the appropriateness of
PSA screening employing four criteria:
burden of suffering, effectiveness of
H o w
t o
u s e
P r o s t a t e - S p e c i f i c
A n t i g e n
screening, potential harms of screening,
and economics of screening. He cautioned that while the burden of suffering
with carcinoma is obvious, currently
there are no data to suggest that screening is effective in reduction of cancer-related mortality.
A number of important trials are being conducted, yet the rising rate of
prostate cancer incidence and mortality
makes early detection and attempts at curative therapy the standard of care in
most settings. The Prostate, Lung, Colon,
and Ovarian trial67 will attempt to demonstrate the benefit of early detection of
prostatic carcinoma, using DRE and
PSA. Although this study is hampered by
the fact that no therapeutic approach is
mandated and the screening interval may
be too short, the biggest problem may be
identification and maintenance of a control (nontested) population.
The Prostate Intervention Versus
Observation Trial is currently under way
to investigate whether radical prostatectomy is effective in reducing cancer mortality.68 This study, which will randomize
2,000 men to radical prostatectomy versus observation, will in effect ultimately
prove or disprove the utility of early detection efforts. If curative therapy prolongs life, the efficacy of efforts to detect
earlier cancers will almost certainly be realized. However, if treatment is shown to
be ineffective without advances in our
therapeutic armamentarium, emphasis
on early detection will decrease.
It is likely that even with the opportunity of universal screening, not every
man would undergo such testing. Never-
theless, the economic implications are
staggering. In addition, we may diagnose
cancer in many men unlikely to derive
benefit. In my opinion, the most critical
issue in prostate cancer today is the development of reliable markers of malignant potential. Table 7 depicts a partial
listing of some of the markers currently
Ultimately, it is the primary care
provider who must make the decision
whether the early detection of prostate
cancer will enhance his or her patient’s
well-being. Given such parameters as patient age, intercurrent illness, social situation, and patient desires, the primary
physician is well suited to counsel the patient with regard to benefits as well as potential risks that might arise from early
detection efforts. If, after such counsel, it
is determined that the patient will derive
benefit from early detection of prostatic
carcinoma, then a carefully performed
DRE with an attempt to identify subtle
changes, such as minimal induration or
perhaps even asymmetry, should be carried out and serum PSA measured. In my
opinion, any abnormality on DRE or a
serum PSA level greater than 4.0 ng/ml
should be an indication for referral to a
urologist. I do not believe that the current
data support the use of PSA velocity,
PSA density, age-specific PSA cutoffs, or
PSA isoform levels to determine who
should undergo further testing (i.e., biopsy). If, however, the primary provider and
patient feel that no benefit is to accrue by
early detection efforts, then no testing
should be performed.
1. Wingo PA, Tong T, Bolden S: Cancer statistics,
1995. CA Cancer J Clin 1995;45:8-30.
2. Seidman H, Mushinski MH, Gelb SK, Silverberg
E: Probabilities of eventually developing or dying
of cancer: United States, 1985. CA 1985;35:36-56.
3. Scardino PT, Weaver R, Hudson MA: Early
detection of prostate cancer. Human Pathology
4. Brawer MK, Ellis WJ: Chemoprevention for prostatic carcinoma. Cancer 1994. In Press.
5. Stamey TA, Yang N, Hay AR, et al: Prostate-specific antigen as a serum marker for adenocarcinoma
of the prostate. N Engl J Med 1987;317:909-916.
6. Ercole CJ, Lange PH, Mathisen M, et al: Prostate
specific antigen and prostatic acid phosphatase in
Ca—A cancer Journal for Clinicians
C a n c e r
C l i n
the monitoring and staging of patients with prostatic cancer. J Urol 1987;138:1181-1184.
7. Ferro MA, Barnes I, Roberts JBM, Smith PJB:
Tumour markers in prostatic carcinoma: A comparison of prostate-specific antigen with acid phosphatase. Br J Urol 1987;60:69-73.
8. Hudson MA, Bahnson RR, Catalona WJ:
Clinical use of prostate specific antigen in patients
with prostate cancer. J Urol 1989;142:1011-1017.
9. Sensabaugh GF, Crim D: Isolation and characterization of a semen-specific protein from human
seminal plasma: A potential new marker for semen
identification. J Forensic Sci 1978;23:106-115.
10. Brawer MK, Rennels MA, Nagle RB, et al:
Serum prostate specific antigen and prostate
pathology in men having simple prostatectomy. Am
J Clin Pathol 1989;92:760-764.
11. Brawer MK, Bostwick DM, Peehl DM, et al:
Keratin immunoreactivity in the benign and neoplastic human prostate. Cancer Res 1985;45:36633667.
12. Bostwick DM, Brawer MK: Prostatic intraepithelial neoplasia and early invasion in prostate
cancer. Cancer 1987;59:788-794.
13. Bigler SA, Brown M, Deering RE, Brawer MK:
Immunohistochemistry of type VII collagen in
human prostatic tissue. J Urol 1994;151(Suppl):
14. Fuchs ME, Brawer MK, Rennels MA, Nagle
RB: The relationship of basement membrane to
histologic grade of human prostatic carcinoma.
Modern Path 1989;2:105-111.
15. Catalona WJ, Smith DS, Ratliff TL, et al:
Measurement of prostate-specific antigen in serum
as a screening test for prostate cancer. N Engl J
Med 1991;324:1156-1161.
16. Catalona WJ, Smith DS, Ratliff TL, Basler JW:
Detection of organ-confined prostate cancer is
increased through prostate-specific antigen-based
screening. JAMA 1993;270:948-954.
17. Brawer MK, Chetner MP, Beatie J, et al:
Screening for prostatic carcinoma with prostate
specific antigen. J Urol 1992 147:841-845.
18. Catalona WJ, Richie JP, Ahmann FR, et al:
Comparison of digital rectal examination and
serum prostate specific antigen in the early detection of prostate cancer: Results of a multicenter
clinical trial of 6,630 men. J Urol 1994;151:12831290.
19. Brawer MK, Lange PH: PSA in the screening,
staging and follow up of early-stage prostate cancer:
A review of recent developments. World J Urol
20. Brawer MK, Beattie J, Wener MH, et al:
Screening for prostatic carcinoma with prostate
specific antigen: Results of the second year. J Urol
21. Brawer MK, Beatie J, Wener MH: PSA as the
initial test in prostate carcinoma screening: Results
of the third year. J Urol 1993;149(Suppl):299A.
22. Kinne DW, Kopans DB: Physical examination
and mammography in the diagnosis of breast disease, in Harris JR et al (ed): Breast Disease.
Vol. 45 No. 3 may/june 1995
1 9 9 5 ; 4 5 : 1 4 8 - 1 6 4
Philadelphia, JB Lippincott Company, 1987.
23. Moskowitz M: Cost-benefit determinations in
screening mammography. Cancer 1987;60(Suppl):
24. Babaian RJ, Camps JL: The role of prostatespecific antigen as part of the diagnostic triad and as
a guide when to perform a biopsy. Cancer 1991;
25. Bazinet M, Meshref AW, Trudel C, et al:
Prospective evaluation of prostate-specific antigen
density and systematic biopsies for early detection
of prostatic carcinoma. Urology 1994;43:44-52.
26. Cooner WH, Mosley BR, Rutherford CL Jr:
Clinical application of transrectal ultrasonography
and prostate specific antigen in the search for
prostate cancer. J Urol 1988;139:758-761.
27. Cooner WH, Mosley BR, Rutherford CL Jr, et
al: Prostate cancer detection in a clinical urological
practice by ultrasonography, digital rectal examination and prostate specific antigen. J Urol
28. Mettlin C, Lee F, Drago J, et al: The American
Cancer Society National Prostate Cancer Detection
Project: Findings on the detection of early prostate
cancer in 2425 men. Cancer 1991;67:2949-2958.
29. Rommel FM, Augusta VE, Breslin JA, et al:
The use of prostate specific antigen and prostate
specific antigen density in the diagnosis of prostate
cancer in a community based urology practice. J
Urol 1994;151:88-93.
30. Ellis WJ, Chetner MP, Preston SD, Brawer MK:
Diagnosis of prostatic carcinoma: The yield of
serum prostate specific antigen, digital rectal examination and transrectal ultrasound. J Urol 1994;
31. Mettlin C, Jones G, Averette H, et al: Defining
and updating the ACS guidelines for the cancer
related check-up: Prostate and endometrial cancer.
CA Cancer J Clin 1993;43:42-46.
32. Brawer MK, Schifman RB, Ahmann FR, et al:
The effect of digital rectal examination on serum
levels of prostate specific antigen. Arch Pathol Lab
Med 1988;112:1110-1112.
33. Crawford ED, Schutz M, Clejan S, et al: The
effect of digital rectal examination on prostate specific antigen. J Urol 1991;145:398A.
34. Yuan JJ, Catalona WJ: Effect of digital rectal
examination, prostate massage, transrectal ultrasonography and needle biopsy of the prostate on
serum prostate specific antigen levels. J Urol
35. Carter HB, Morrell CH, Pearson JD, et al:
Estimation of prostatic growth using serial prostatespecific antigen measurements in men with and
without prostate disease. Cancer Res 1992;52:33233328.
36. Pearson JD, Carter HB: Natural history of
changes in prostate specific antigen in early stage
prostate cancer. J Urol 1994;152:1743-1748.
37. Schmid HP, McNeal JE, Stamey TA:
Observations on the doubling time of prostate cancer: The use of serial prostate-specific antigen in
patients with untreated disease as a measure of
H o w
t o
u s e
P r o s t a t e - S p e c i f i c
increasing cancer volume. Cancer 1993;71:20312040.
38. Porter JR, Hayward R, Brawer MK: The significance of short-term PSA change in men undergoing ultrasound-guided prostate biopsy. J Urol
39. Littrup PJ, Kane RA, Mettlin CJ, et al: Costeffective prostate cancer detection. Cancer 1994;
40. Komatsu K, Wehner N, Prestigiacomo AF, et al:
Variation of serum prostate specific antigen in 814
men from a screening population: Intra-individual
assay variation is greater than the repeat assay variation. J Urol 1994;151(Suppl):401A.
41. Benson MC, Whang IS, Olsson CA, et al: The
use of prostate specific antigen density to enhance
the predictive value of intermediate levels of serum
prostate specific antigen. J Urol 1992;147:817-821.
42. Benson MC, Whang IS, Pantuck A, et al:
Prostate specific antigen density: A means of distinguishing benign prostatic hypertrophy and prostate
cancer. J Urol 1992;147:815-816.
43. Babaian RJ, Fritsche HA, Evans RB: PSA and
prostate gland volume: Correlation and clinical
application. J Clin Lab 1990;4: 135-137.
44. Brawer MK, Aramburu EAG, Chen GL, et al:
The inability of prostate specific antigen index to
enhance the predictive value of prostate specific
antigen in the diagnosis of prostatic carcinoma. J
Urol 1993;150:369-373.
45. Stamey TA: Making the most out of six systematic sextant biopsies. Urology 1995;45:2-12.
46. Bazinet M: Personal communication. 1994.
47. Rommel FM: Personal communication. 1994.
48. Mettlin C, Littrup PJ, Kane RA, et al: Relative
sensitivity and specificity of serum prostate specific
antigen level compared with age-referenced PSA,
PSA density, and PSA change. Cancer 1994;
49. Myrtle J, Klimley P, Ivor L, Bruni J: Clinical
utility of prostate specific antigen (PSA) in the
management of prostate cancer. Advances in
Cancer Diagnostics. 1986.
50. Oesterling JE, Jacobsen SJ, Chute CG, et al:
Serum prostate-specific antigen in a communitybased population of healthy men. JAMA 1993;270:
51. Dalkin BL, Ahmann FR, Kopp JB: Prostate
specific antigen levels in men older than 50 years
without clinical evidence of prostatic carcinoma. J
Urol 1993;150:1837-1839.
52. Mettlin C, Murphy GP, Lee F, et al:
Characteristics of prostate cancer detected in the
American Cancer Society-National Prostate
Cancer Detection Project. J Urol 1994;152:17371740.
53. Petteway J, Brawer MK: Age specific vs. 4.0
ng/ml as a PSA cutoff in the screening population:
Impact on cancer detection. J Urol 1995;153
54. Catalona WJ, Hudson MA, Scardino PT, et al:
A n t i g e n
Selection of optimal prostate specific antigen cutoffs for early detection of prostate cancer: Receiver
operating characteristic curves. J Urol 1994;
151(Suppl): 449A.
55. Christensson A, Bjork T, Nilsson O, et al: Serum
prostate specific antigen complexed to 1-antichymotrypsin as an indicator of prostate cancer. J Urol
56. Lilja H: Significance of different molecular
forms of serum PSA: The free, noncomplexed form
of PSA versus that complexed to 1-antichymotrypsin. Urol Clin N Am 1993;20:681-686.
57. Stenman U, Leinonen J, Alfthan H, et al: A
complex between PSA and 1-antichymotrypsin is
the major form of prostate-specific antigen in serum
of patients with prostatic cancer: Assay of the complex improves clinical sensitivity for cancer. Cancer
Res 1991;51:222-226.
58. Lilja H, Christensson A, Dahlen U, et al:
Prostate-specific antigen in serum occurs predominantly in complex with 1-antichymotrypsin. Clin
Chem 1991;37:1618-1625.
59. Lilja H, Bjork T, Abrahamsson P, et al:
Improved separation between normals, BPH and
carcinoma of the prostate by measuring free, complexed and total concentrations of prostate specific
antigen. J Urol 1994; 151(Suppl):400A.
60. Stamey TA, Chen Z, Prestigiacomo A: Serum
prostate specific antigen binding 1-antichymotrypsin: Influence of cancer volume, location and
therapeutic selection of resistant clones. J Urol
61. Brawer MK, Wener MH, Daum PR, Close B:
Method to method variation in assays for PSA. J
Urol 1994;151(Suppl):450A.
62. Brawer MK, Daum P, Petteway JC, Wener MH:
Assay variability in serum PSA determination.
Prostate 1995. in press.
63. Murphy GP: The second Stanford conference
on international standardization of PSA assays.
Cancer 1995;75:122-128.
64. Ellis WJ, Brawer MK: Repeat prostate needle
biopsy: Who needs it? J Urol 1995. in press.
65. Feightner JW: The early detection and treatment of prostate cancer: The perspective of the
Canadian Task Force on the periodic health examination. J Urol 1994;152:1682-1684.
66. Woolf SH: Public health perspective: The health
policy implications of screening for prostate cancer.
J Urol 1994;152:1685-1688.
67. Gohagan JK, Prorok PC, Kramer BS, Cornett
JE: Prostate cancer screening in the Prostate, Lung,
Colorectal and Ovarian Cancer Screening Trial of
the National Cancer Institute. J Urol 1994;
68. Wilt TJ, Brawer MK: The prostate cancer intervention versus observation trial: A randomized trial
comparing radical prostatectomy versus expectant
management for the treatment of clinically localized prostate cancer. J Urol 1994;152:1910-1914.
Ca—A cancer Journal for Clinicians