The predictability of clinical antitumor effects using two distinctive in vitro chemosensitivity tests: an analysis of true positive cases.

The results of two types of in vitro chemosensitivity tests, namely, the human tumor clonogenic assay (HTCA) and the succinic dehydrogenase inhibition assay (SDIA), for solid tumors, including stomach, colorectal and lung cancers, were analyzed and their correlation with clinical effects evaluated. The anticancer agents employed were mitomycin C (MMC), 5-fluorouracil (5-FU), adriamycin (ADM) and cisplatin (DDP). The evaluability rates of the assays were 54.5% for HTCA and 89.0% for SDIA. Among the 29 cases with evaluable lesions subjected to HTCA, there were 4 true positives, 9 false positives, and 16 true negatives, whereas among the 32 cases subjected to SDIA, the corresponding numbers were 2, 6, and 24, respectively. There were no false negatives for either assay, the accuracy of prediction for HTCA being 69.0% and for SDIA, 81.3%. The true positives of both assays included one complete response (CR) and five partial responses (PR), although the eventual outcome was cancer death in all cases. Interestingly, in five out of the six true positive cases, the agent involved was either ADM or DDP, both the which are usually regarded as "second line" anticancer agents for gastrointestinal carcinomas.