Longer-term PCa treatment comparisons are still fairly rare.  Recently
Kupelian, et. al. published a paper comparing the freedom from
biochemical recurrence between several types of PCa treatment
(International Journal of Radiation Oncology Biology Physics, Volume
58, Issue 1, 1 January 2004, Pages 25-33, Radical prostatectomy,
external beam radiotherapy <72 Gy, external beam radiotherapy >72 Gy,
permanent seed implantation, or combined seeds/external beam
radiotherapy for stage T1–T2 prostate cancer).  The authors concluded
that biochemical recurrence-free rates at 7 years were similar between
RP (76%), EBRT @ 72 Gy (81%), seed implants (SI, 75%) and SI+EBRT
(77%).  This analysis involved 2,991 patients treated at the Cleveland
Clinic or MSK at Mercy Medical Center.  This paper was peer-reviewed
prior to publication.

However, it is important to note that different definitions of failure
were used for comparing the RP (PSA >0.2 ng/ml) and the RT (ASTRO, 3
consecutive PSA rises after nadir) therapies.  Imagine that we take a
walk together and we decide to determine whether the stones we see are
"hot."  I have a thermometer and say anything over 70F is what I will
call "hot".  You say, oh, I'll just touch them with my finger to
determine if they are "hot."  After we both measure the same 100
stones, it wouldn't be surprising if we had a different count on how
many were "hot", and in this case we are measuring the same underlying
population of stones.  Suppose you take a walk in your neighborhood
and sample 100 stones and I do the same in my neighborhood.  Odds are
that the discrepancy between our counts would be further magnified.
Critz has a nice PCa study that makes this same point (J. Urol., vol.
167, 1310-13, 2002, A Standard Definition of Disease Freedom Is Needed
for Prostate Cancer: Undetectable PSA Compared with the ASTO Consensus
Definition).  In this paper Critz uses his RCOG SI+EBRT (~ 160 Gy)
technique to treat 591 men.  He then uses both ASTRO and 0.2 ng/ml
definitions to determine failure and compares the results (BTW, RCOG
uses 0.2 ng/ml as their definition of failure, at 160 Gy they believe
they have "eliminated" the prostate much like surgery, in fact they
call their treatment Prost-R-Scission).  Median follow-up was 6 years
(range 5-8).  There were 64 failures using the ASTO definition, 92
failures using 0.2 ng/ml, a difference of 28 failures (44%)!  When
projected out to 8 years using standard Kaplan-Meier statistical
methods this led to an 89% recurrence-free rate using ASTRO and an 84%
recurrence-free rate using the 0.2 ng/ml PSA cutpoint, 5 percentage
points of difference favoring ASTRO.  Even larger differences were
seen with lower risk groups.  Going back to the "stones" analogy for a
moment, we compared the same underlying population of stones.  In
Kupelian's paper, the RP and various RT subgroups had slightly
different underlying populations which would only serve to make the
comparisons even more suspect.

At last week's AUA meeting, Dr. Catalona, another noted PCA
researcher, made the same point (AUA 2004 Meeting, Program#/Poster#:
1179, COMPARISON OF RADICAL PROSTATECTOMY, RADIOTHERAPY, HORMONAL
THERAPY, AND WATCHFUL WAITING FOR SCREEN-DETECTED PROSTATE CANCER: AN
UPDATE).  His team compared the 7-year biochemical recurrence-free
survival for over 3,000 PCa patients that had selected different
treatments.  The results, by risk group, are presented in the
following Table:

             7-year Survival Estimate Ranges (%)
Treatment, Median Follow-up (mo.), Low-risk, Intermed-risk, High-risk
Group
Radical Prostatectomy    68        87        80             59
Brachytherapy            42        90-98     74-83          41
External Beam Radiation  63        60-83     47-60          33-54
Hormonal Therapy         50        23-71     50-64          25-40
WW                       47        22-56     16-37          0

They too used 0.2 ng/ml to determine failure for the surgery patients
and the ASTRO definition for all others.  Remember now, the ASTRO
definition requires 3 consecutive PSA rises after a man's PSA nadirs.
If an RT man nadired around 2 years post-treatment and he then
measured his PSA once every year, he couldn't fail before year 5
according to ASTRO no matter what!  Catalona concludes, "Radical
prostatectomy and brachytherapy had similar progression rates that are
lower than other treatments. However, assuming that all PSA rises that
have not yet been twice verified will ultimately become progressions,
RRP has a lower progression rate than other treatments with
intermediate term follow-up."

As others have pointed out, when we have 20-30 years of data, we will
know the "truth" because at those times, interpretive differences
caused by differences in the ASTRO and 0.2 ng/ml failure definitions
will be leveled out.

The question remains what is a man making his decision today to do?  I
don't pretend to have the answer, but let me put a few generalizations
forward for comment.

1.  Several studies have now suggested that for low-risk men,
treatment differences between RP and various RT therapies are small.
I suspect that this is true, maybe because of the ASTRO - 0.2 ng/ml
comparison issue, RP might be 5-10 points better in the 7-10 year
range.  Is that difference large enough to affect decision making?
2.  Men continue to fail out past 10 years post-RP, there is some hint
in the available data that this might not be the case for SI+EBRT.
There is some conjecture that with the RCOG SI+EBRT treatment, men
stop failing after 8 years.  If true, this would tend to reduce any RP
advantage seen at 10 years, at least in comparison to SI+EBRT.
3. I was surprised to see how effective RP was for higher-risk men in
Catalona's Table.
4. Although not considered in these papers, other recently published
data shows cryo to be, on average, an ineffective treatment both in
terms of biochemical relapse and QOL issues (J. Urol., 2003; 170(4):
1126-1130).
5. Both Kupelian's and Catalona's numbers for the various treatments
can be noticeably improved by an artist.  For example, Catalona's
study (which is a database study) shows 87%
biochemical-recurrence-free survival at 7 years.  The Hopkins'
nomogram shows roughly 95% biochemical-relapse-free survival at 10
years.  Similarly, I suspect that if Drs. Onik and Lee published their
cryo numbers, they would be significantly better than those in the
above-referenced J. Urol cryo article.
6.  Watchful waiting, particularly when substantial lifestyle changes
and active disease surveillance are practiced, probably hasn't
received adequate consideration, particularly amongst younger men.
Moul, et. al. has shown that in about 25% of the low-risk men he
studied, there was no significant disease progression after 8 years of
WW (J. Clin. Oncology, Vol. 21, 4001-8, 2003).  His data also showed
that the number of men moving out of WW clearly plateaued at around 5
years, suggesting that those remaining in the WW group at this time
might be able to remain on a WW regimen much longer!  If low-risk men
are further selected based on the Hopkins' low-volume disease
criteria, I suspect the number of men who successfully practice active
WW out to 8 years and beyond would increase further.
7.  Treatment selection usually factors in QOL issues as well as
freedom from disease recurrence.  I'll post some data on an
interesting QOL comparison of treatments shortly.

In terms of a disclaimer, I had an RRP in 02/03.  I don't think RP is
right for everyone and I believe whatever choice a man makes IS the
right choice for him.  I am a bit of an academician and because the
data appeared so muddled and confused for me when I was making my
decision, I continue to try and keep current hoping that some of what
I sort through and write about may prove useful to those who
follow...best wishes and good health, Ron