|
The Prostate Specific
Antigen Has Become Less Informative Over The Past 20 Years. (December
2004)
Drs. Stamey, McNeal et al have offered a significant and
consequential observation that the serum PSA level, especially in the
range of < 10 ng/mL, no longer is reflective of prostate cancer, but
only of benign prostate hypertrophy. Their important report, “The
Prostate Specific Antigen Era in the United States is Over for Prostate
Cancer: What Happened in the Last 20 Years?” (J Urol, Oct. 2004)
documents their observations based on 1317 radical prostatectomy
specimens. The issue under evaluation was “how well preoperative serum
prostate specific antigen (PSA) reflects the largest cancer in
consecutive untreated radical prostatectomies during the last 20 years at
Stanford University.”
At its essence this article presents a challenge to the validity of the
premise underlying PSA screening. Stamey maintains that “current
evidence from the last 10 years is convincing that the relationship
between prostate cancer and serum PSA is tenuous at best, especially with
serum PSA less than 10 ng/mL...”. They interpret their data as showing
that there are “serious limitations in the relationship of serum PSA to
prostate cancer volume and Gleason grade 4/5 cancer”, rendering PSA
currently “misleading in the diagnosis of prostate cancer”.
As is well recognized, the characteristics of diagnosed
prostate cancer have markedly changed over 20 years. The researchers
categorized their findings into four 5-year periods. In the earliest
5-year period, beginning in 8/1983, the PSA level at diagnosis was highly
significantly related to the largest cancer, and to the presence of
capsular penetration, positive lymph nodes, seminal vesicle invasion, and
to the percent of Gleason grade 4/5 in the largest cancer.
In the most recent period, 1/1999-7/2003, the statistically
significant relationship in these associations has been lost. The
differences in prostate cancer characteristics between the first and last
period, no doubt due to extensive screening, are not surprising and reveal
the changing face of prostate cancer. The percent of palpable cancers on
DRE decreased from 90.8% to 16.7%; the mean serum PSA at diagnosis dropped
from 24.7 to 8.14 ng/mL; and the mean volume of the largest cancer
shrank from 5.33 to 2.44 cc. Positive lymph node discovery decreased from
12.5% to 0.0%, and seminal vesicle invasion decreased from 23% to 5.4%. Notably,
however, there was no significant change in prostate weight, 46.5 gm at
first and 43.5 gm in the last period, which continued to be reflected by
the PSA level.
The “nuts and bolts” of Stamey’s argument lie in the
calculations reflected in Table 4: ”Comparison of Pearson correlations
of preoperative serum PSA with radical prostatectomy morphology in the
first and last periods”. [Courtesy of Google: the Pearson coefficient
measures the strength of a linear relationship between two variables, with
a value of 0 (range -0.3 to +0.30) showing little or no association; +0.3
to +0.7 a weak association; and +0.7 to 1.0 a strong positive
association.] The table shows that for the largest cancer the coefficient
decreased from 0.659 to 0.148; for capsular penetration it declined from
0.539 to 0.033; for % seminal vesicle invasion - 0.437 to 0.069; and
finally, for % Gleason grade 4/5 in the largest cancer the decrease was
0.274 to 0.031. By the last 5-year period there was “no
correlation of [the preoperative] serum PSA with any morphological
variable except prostate weight.
An observation relevant to the increasing awareness that
adverse histology may be present very early in the life history of some
cancers was the finding that despite the observed transition over
time to smaller, lower PSA cancers, 83% of which were eventually T1c
compared to 7% at first, the percent with Gleason grade 4/5 showed a 13%
increase from 31% to 35%!
In the background of Stamey’s argument - and referred to
in the article - is the data obtained by Sakr (Eur Urol,30:138,1996) who
examined the prostates of accidental death victims of a wide range of
ages. He found that the prevalence of invasive prostate cancer
increased from 8% in 20 year old men to 80% in men in their 70’s.
Interestingly, the presence of HGPIN showed narly identical increasing
figures for prevalence with aging.
Conclusions from the Stanford data suggest that in the
current setting the PSA value no longer gives useful guidance for the
detection of “significant” cancers. “This means that any excuse to
biopsy the prostate has an excellent, age dependent chance of being
positive.” The authors’ overall conclusion was that “What is
urgently needed is a serum marker [in the PSA range of 2 to 10 ng/mL] for
prostate cancer that is truly proportional to the volume and grade of this
ubiquitous cancer, and solid observations on who should and who should not
be treated...”.
Bottom Line:: If we are
persuaded by Stamey’s argument that the “PSA today as a basis for
diagnosing and treating prostate cancer is related only to the amount of
benign prostatic hypertrophy in the prostate”, where do we go from here?
«
Back to Article List
|
