Prostate specific antigen in the preoperative and postoperative evaluation of localized prostatic cancer treated with radical prostatectomy.

The usefulness of prostate specific antigen to predict final pathological stage was studied in 178 consecutive patients. Prostate specific antigen was determined preoperatively in all patients by a monoclonal immunoradiometric assay. All pathological specimens were examined for capsular penetration, seminal vesicle involvement and lymph node involvement. Prostate specific antigen correlated directly with capsular penetration (p less than 0.002), seminal vesicle involvement (p less than 0.02) and lymph node involvement (p less than 0.05). However the diagnostic accuracy of an elevated serum antigen level on an individual basis was only 55 per cent for capsular penetration and 50 per cent for seminal vesicle involvement and lymph node involvement. With a log-linear regression model, the half-life of prostate specific antigen was calculated to be 3.15 +/- 0.09 days. From the equation PSA (t) equals PSA (2) e[-0.2197(t-2)], prostate specific antigen can be used to detect residual cancer on day t in the immediate postoperative period. With respect to long-term followup, 127 patients have been monitored for longer than 2 months postoperatively with prostate specific antigen (mean followup 2 years, range 2 months to 8.6 years). Of the 101 patients who had favorable pathological findings at operation (organ-confined cancer or capsular penetration only) 92 (91 per cent) had a followup antigen concentration in the female range (0.0 to 0.2 ng. per ml.), whereas only 5 of 26 men (19 per cent) with either seminal vesicle involvement or lymph node involvement had an antigen value that was less than 0.2 ng. per ml. All patients with a documented clinical recurrence (8 of 127, 6 per cent) had an elevated followup serum prostate specific antigen concentration. These findings suggest that preoperative levels of prostate specific antigen are not sufficiently reliable to predict final pathological stage on an individual basis in patients with early prostatic cancer, and that the antigen is a sensitive tumor marker for the detection of residual disease after radical prostatectomy and subsequent recurrence of tumor on long-term followup.