|
Lead time bias in relation to
screening with PSA (November 2002)
PSA screening has been and remains a
controversial issue. Clearly, if early diagnosis and treatment lowers
ultimate mortality from PC the debate is over. However, what about overall
mortality versus PC specific death rate (PCSD) and the consequences of
treatment morbidity in men who then die from other causes? Where does
"lead time bias"(the time by which the PSA advances the diagnosis) fit
into the picture and how long is it? The five-year survival for patients
with PC in 1950 was 40% and now is 95%. Is this real gain or just the
consequence of lead-time bias? Some new data sheds some light on this
tricky issue.
A much quoted article, "A Randomized Trial Comparing Radical Prostatectomy
With Watchful Waiting In Early Prostate Cancer" appeared in the September
12 issue of the N Eng J Med. It reported that the overall mortality in 696
Swedish men at a median follow-up of 6.2 years was not significantly
different for the RP group versus the observation group. The men's average
was 64 years, PSA's <50 ng/mL, expected future life 10 years or more, and
bone scans negative. If nodes were positive at RP the man was excluded
from the study. Similar treatment was given to the two groups when local
or distant disease developed. Important points for this discussion are: 1)
Screening accounted for only 5.2% of diagnoses (the study period began in
l987 at the very beginning of the PSA period) and 40% of men were
symptomatic at diagnosis; 2) 12% of men presented with T1C and 12% T1B
disease respectively, and 75% were T2, and, 3) 46% of PSA's were >10 ng/mL,
and, 4) 60% of Gleason score were 2-6, 23% 7, and 5% 8-10.
The outcome: at 6.2 year median F/U PCSD was 4.6% in the RP group and 8.9%
in the WW group. Early data at eight years showed a PCSD rate of 7.1% for
the RP group versus l3.6% for WW. HOWEVER, at 6.2 years of F/U, the
OVERALL mortality was not significantly different. (Editor: This is the
point that I fear most lay readers might take away from the study]
Interestingly, at only the five year point there was no difference between
PCSD in either group. It was only at 8 years that the full survival
advantage in PCSD was seen favoring the RP group, i.e. 7.1% versus 13.6%.
At 8 years there was a 6% absolute reduction in both overall and disease
specific death rate favoring the RP group but this was incomplete data. At
2 years there was no difference in metastatic disease, but by 8 years the
RP group showed a 14% comparative reduction. And by 8 years local
recurrence developed in 60% for the WW group versus 20% for RP.
Bottom Line: The question here is - Is this study really about EARLY
prostate cancer?
«
Back to Article List |
