Primary extraprostatic spread or failure after prostate cancer treatment can occur locally in the prostatic fossa and/or metastasize to regional and/or distant lymphatics and/or in bone. We evaluated the ability of the ProstaScint (Cytogen Corp., Princeton, NJ) scan to identify soft tissue recurrence of prostate cancer in patients who failed primary treatment, and we monitored their responses to a secondary treatment.The 111indium-labeled monoclonal antibody (ProstaScint) was evaluated in a series of 136 consecutive scans associated with a complete set of relevant clinical and biochemical data. These scans represented 100 patients, imaged between November 1994-May 1998, who underwent a definitive prostate cancer treatment followed by evidence of recurrence. All patients had serum prostate-specific antigen (PSA), Western-blot serum prostate-specific membrane antigen (PSMA), and bone scans. Prostatic fossa and/or lymph node biopsies were available in 33 patients.Overall, no adverse reactions were related to any of the radioactive antibody infusions. The average age was 69 years (range, 48-89 years), serum PSA was 55.9 ng/ml (range, 0-2,185 ng/ml), and serum PSMA was 0.32 (relative intensity levels range, 0.04-0.55). The initial therapies given were radical prostatectomy (55 scans), prostate radiation (74 scans), and/or hormonal therapy (77 scans). Twelve patients exhibited a negative scan. Local recurrence alone was identified in 58 scans (42.6%). Lymph node metastases were identified in 66 scans (48.5%). Of these, 21 had regional metastases alone, and 45 had distant lymph node metastases. Ten scans showed skip lymph node metastases without regional failure. PSA significantly correlated with negative, pelvic, and extrapelvic scan findings (P < or = 0.02). PSMA correlated best with pelvic recurrence and extrapelvic metastases in prostatectomized patients. Thirty-four patients had repeated scans for monitoring treatment response. Of these patients, 19 (56%) showed progression on the second scan, consistent with persistent increase in PSA and PSMA levels in all but 2 patients. Another 11 patients showed no change, and 4 patients showed partial remission, suggesting a response to the salvage treatment. All 20 positive prostate biopsies and 4 of the 7 positive lymph node biopsies correlated with ProstaScint findings, whereas 4 of the 6 patients with a negative biopsy had negative scans (i.e., 89% sensitivity and 67% specificity). Bone metastases were identified in 42 bone scans; 45% of these showed no lymph node metastasis on ProstaScint. In another 24 patients (57%), bone metastases were detected on ProstaScint examinations.