|
Expectant Management Of Localized Prostate Cancer With An Intent To Cure: An Oxymoron?
(July 2004)
A thoughtful
review of this concept was presented by Kahn, Partin, and Carter in
their article, "Expectant Management of Localized Prostate Cancer"
(Urology, Nov. 2003). They set the stage for the discussion by stating:
"... expectant management with curative intent has recently been
described as a rational option for carefully selected men with
prostate-specific antigen (PSA) screen-detected cancers thought to be of
low volume." In their opinion low volume disease (< .5 cm3) was best
predicted by a PSA density of < 0.1/ng/mL/g, stage cTic, Gleason < 6, <
3 biopsy cores positive with none having > 50% cancer. This shift in
strategy is away from the concept of "watchful waiting" in which initial
therapy is deferred because the patient's predicted life expectancy is
considered shorter than the predicted time before symptom onset or
death. Recent information detailing emerging prostate cancer
demographics has made "expectant management" a relevant issue. What new
information have we learned?
As reported by
Cooperberg et al. (JCO, June 2004) the CaPSURE data shows that in the
period 1999-2001 among patients diagnosed with low-risk disease (PSA
< 10 ng/mL, Gleason sum < 6, clinical stage < T2a)
nearly two thirds were screen-detected T1c cancers, compared to 15.2% in
1989-1992; and unilaterally palpable T2a tumors have fallen from nearly
three-quarters (1989-1992) to 36% of cases (1999-2001).
It is well known
that PSA screening introduces a significant "lead time bias" and this
"bias" introduces a confounding factor in comparisons among treatment
outcomes. "Lead time" represents the interval between a diagnosis based
on PSA screening compared to a diagnosis based on clinical symptoms.
Etzioni of the Fred Hutchinson had estimated a mean lead time bias of 3
- 7 years. However, data from Rotterdam (Am J of Urol Rev, April 2004)
extends the estimate of lead time on average to 12.3 years, and a
Swedish study with 20 years of follow-up presented a median lead time
estimate of 12 years.
The large
percentage of men who have been recently diagnosed fall into this low
risk category, and an even higher proportion will do so in the future.
Data based on the 7 year follow-up of 2950 men who were in the
untreated cohort in the Prostate Cancer Prevention Trial was
reported in the article , "Prevalence of Prostate Cancer among Men with
a Prostate-Specific Antigen Level < 4 ng per Milliliter" (NEJM,
May 27, 2004). Of the 449 men with PSA values of < 4 ng/mL diagnosed as
T1c cancers, 302 (67%) had a Gleason score of 6; 47 (10.5%) had Gleason
score 5, and 12 (2.6%) Gleason score 4. Gleason scores of 7 or more were
found in only 67 (15%) of men. In total 85% of the cancer were Gleason
< 6. Based on an estimated incidence of prostate cancer in the
United States for 2003 of 220,900 men, a rough calculation would suggest
that possibly 100,000 of these men would present with low risk, T1c
prostate cancer.
The issue of
screen-detected low-risk prostate cancer is often linked with a concern
of "overdiagnosis", with its implication that treatment in the early
stage of prostate cancer is often "overtreatment", or at least treatment
rendered unnecessarily early. In contrast, when "early diagnosis" is
aligned with the emerging context of "expectant management", a different
strategy emerges, namely, postponing treatment - with all its associated
side effects - until a later time when intervention can still be
predicted to offer the same likelihood of cure as if it had been
performed at initial diagnosis.
Jan-Erik Johansson
et al. begin their article, "Natural History of Early, Localized
Prostate Cancer" (JAMA, June 9,2004) by stating, "Without understanding
the natural history of prostate cancer diagnosed at an early, localized
stage patient counseling and clinical management are difficult." They
point out that few men "diagnosed at an early clinical stage die from
prostate cancer within 10 to 15 years following diagnosis." The basis of
their article is the mean 21 year follow-up of 223 men with T0-T2 NX M0
prostate cancer, and they observe that "while most cancers had an
indolent course for the first 10 to 15 years...local tumor progression
and aggressive metastatic disease may develop in the long term. "The
prostate cancer mortality rate [for their study group] increased from 15
per 1000 person-years...during the first 15 years to 44 per 1000-person
years beyond 15 years of follow-up (P =.01). This cohort was carefully
followed with appropriate PSA testing and scans, and during the first 6
years 80% of men were re-biopsied every two years. Their data points up
the well known variation Gleason staging that may occur between
sequential needle biopsies. Of these 178 re-biopsied patients the
subsequent biopsy resulted in upgrading in 17%, downgrading in 18% (from
high to moderate differentiation), and downgrading of 13% (from moderate
to low). Three percent of men changed from high to low differentiation.
When intervention was required hormone deprivation was employed. As
expected, disease specific deaths among the 223 men occurred over time:
11 in years 0 - 4; 11 in years 5 - 9; 5 in years 1 - 14; and 8 in >
15 years. The cause specific survival for men with T1 and T2 tumors was
80% at 15 years and 56% at 20 years. 168 men died of other causes.
If the proposition
is accepted that all men presenting with low stage prostate cancer may
not require immediate intervention, then the issue becomes who are the
men with indolent disease where treatment may safely be deferred, and
are there validated guidelines available to help choose the "right" time
to initiate treatment so as to preserve the same possibility of cure
that existed at initial diagnosis?
This strategy of
"expectant management" is a "work in progress", and only partial answers
are available. The optimal candidate would be a man with Gleason score <
6, cT1c, and small volume (< .5 cc) cancer. The literature has many
reports of delayed treatment in mixed stage and grade populations with
higher risk factors, but for the purposes of this discussion these
studies are too wide in their focus. However, Carter (J Urol 167:
1231-1235,2002) chose to narrowly define his study group to conform to
Epstein's model for predicting small volume cancer: men with cT1c stage,
PSA density < 0.15 ng/ml./cm3, Gleason score < 6, less than 3
cores positive and none with > 50% cancer. By applying this model 79% of
men who had these characteristics had pathologic findings on subsequent
RP of a tumor < .5 cc, organ confined disease and not high grade. Carter
followed 81 men who initially met the Epstein criteria for small volume
cancer. DRE and PSA tests were done twice yearly and biopsies were
performed annually. The median initial PSA was 5 ng/ml, median PSA
density was 0.1, and median age was about 65 years. During the median
follow-up of 23 months 31% had disease progression, defined as having a
repeat biopsy containing any Gleason 4 grade tumor, or more than two
cores positive or one core with > 50% cancer. 12 of 13 men who underwent
prostatectomy had "curable" cancer, "defined as a cancer that was
associated with more than a 70% probability of disease-free progression
a decade after surgery". These results are the most definitive available
to begin to address the issue. The study is ongoing.
The findings on
re-biopsy adds an instructive observation that has potential for guiding
the decision to continue an "expectant" strategy. Patel et al, "An
analysis of men with clinically localized prostate cancer who deferred
definitive therapy", J Urol, April 2004, was based on a retrospective
study of 88 men diagnosed between 1984 and 2001 all of whom were judged
as likely to have small volume cancer. The median follow-up was 44
months (range up to 172). They recommended a re-biopsy at 6 months for
all, or at any time "if the patient showed DRE/TRUS or PSA abnormalities
consistent with disease progression for any man who developed worrisome
findings". Overall, the first repeat biopsy (at a median time after
diagnosis of 8 months) was negative in 61%. "Actuarial progression-free
probability at 5 and 10 years was 67% and 56%." The actuarial
probability of remaining progression free at 5 years was 83% for men who
had one or more negative follow-up biopsies, and 43% for those with
positive re-biopsy. The study group was not as narrowly defined as in
Carter. Interestingly, "Our study failed to show a correlation between
short PSA doubling time and progression. In fact, 46% of patients in the
study had a negative PSA doubling time, reflecting the limited impact of
these low volume cancers on serum PSA." The study acknowledges a
critical issue: "Seven of the 31 patients receiving treatment in this
study were unable to live with the anxiety of harboring untreated cancer
and requested treatment."
Bottom Line:
Whether "expectant management with an intent to cure" will become a
validated and well-defined option for cT1c, small volume, low grade
prostate cancer is yet to be determined, but the issue will increase in
importance with increasing PSA screening. Clinicians will need to be
compiling a mental file of the emerging data so as to be comfortable
with the issues.
«
Back to Article List |
