The cancer-specific death rate following radical prostatectomy in pati
ents with organ-confined and specimen-confined disease is 10% at 13.5
years, less than the noncancer death rate of 20% for patients in these
disease extent categories. The median age of all patients in these ca
tegories was 65 years. Cancer remains the dominate cause of death in p
atients with margin-positive disease, being 40% at 13.5 years. Disease
detected by prostate-specific antigen (PSA) rather than digital recta
l examination appears to be of smaller volume and to have a higher pro
bability of being margin negative. Data indicate that early detection
by PSA will shift patients to a more favorable disease category at the
time of surgical intervention. Disease recurrence or persistence by P
SA detection seems to precede clinical detection of disease by 3-5 yea
rs. Disease recurrence by PSA detection does not predict survival outc
ome, probably does not differentiate between local and distant microsc
opic recurrence and is not predictive of biological aggressiveness.