The subrenal capsule tumor implant assay as a predictor of clinical response to chemotherapy: 3 years of experience.

The clinical response to chemotherapy of a series of female patients with advanced pelvic malignancies was compared to the response of their tumors to the same agents in the murine subrenal capsule implant assay. A total of 194 different patients were studied in 242 different assays; 89.3% of the assays were evaluable. There were 83 prospective assays (assays performed before the patient received the chemotherapy) of 66 different patients for which clinical correlations were available. In these assays the sensitivity (frequency of positive test results in responding patients) was 85.0%, the specificity (frequency of negative test results in nonresponding patients) was 57.1%, and the efficiency (percentage correctly classified) was 63.9%. There were 100 retrospective assays (assays performed after the patient had been treated with the chemotherapy) of 69 different patients for which clinical correlations were available. In these assays the sensitivity was 66.7%, the specificity 70.7%, and the efficiency 70.0%. Thirty-one of the patients had both prospective and retrospective assays. There were 59 patients for whom the clinical response to chemotherapy could not be determined. It is believed that the clinical utility of the SRC assay has been validated by the good prospective sensitivity of the assay.