OBJECTIVES: No standard noninvasive diagnostic test reliably differentiates patients with organ-confined prostate cancer from those with lymph node metastases. The ability of a radiolabeled monoclonal antibody, indium-111 (111ln)-capromab pendetide, to identify sites of metastatic disease in patients at moderate to high risk of nodal involvement was investigated. METHODS: The study prospectively evaluated 160 patients with prostate cancer scheduled to undergo pelvic lymph node dissection (PLND) before or during definitive treatment. All were at relatively high risk of nodal involvement by virtue of significantly elevated baseline prostate-specific antigen (PSA) values, Gleason scores, and/or locally advanced clinical stages of disease. The histologic findings of the PLNDs were compared with the results of immunoscintigraphy, computed tomography, and magnetic resonance imaging. RESULTS: Among the 152 evaluable patientS studied with 111In-capromab pendetide before PLND, the sensitivity of immunoscintigraphy for lymph node detection was 62% and the specificity was 72%; the positive predictive value was 62% and the negative predictive value was 72%. In comparison, the sensitivity of computed tomography and magnetic resonance imaging was 4% and 15%, respectively, and the specificity was 100% for both procedures on the basis of a large number of negative interpretations. Logistic regression analysis revealed that immunoscintigraphy with 111In-capromab pendetide provided strong, independent evidence of the presence of lymph node metastases. Furthermore, the analysis indicated that certain combinations of PSA, Gleason score, and 111In-capromab pendetide were particularly effective at predicting the risk of nodal involvement. CONCLUSIONS: Immunoscintigraphy with 111In-capromab pendetide outperformed standard diagnostic imaging techniques in the detection of prostate cancer lymph node metastases and provided independent prognostic information that complemented PSA, Gleason score, and clinical stage.
ASJC Scopus subject areas