Comparison of Tumor Response in Nude Mice and in Patients

Abstract

Four hundred and six human tumors of different origin were transplanted subcutaneously into nude mice. Histological evidence of viable tumor cells was obtained in 79%, rapid tumor growth after 3 months in 51%, and 49% of these tumors were transferred in serial passages. Most of the latter tumors were suitable for drug testing. In order to investigate whether tumor response in nude mice is consistent with patient response, 80 comparisons were performed comprising 55 tumors. Comparisons included carcinomas of the large bowel (27), lung (13), stomach (9), melanomas (8), breast (4) and 19 other tumors. Combination chemotherapy was more successful than single agent therapy (37 vs 14% remission rate). In 21 patients tumors went into remission, consistent with 19 cases in the nude mouse system. In 59 cases the patient did not respond to therapy, corresponding to 57 cases in the nude mouse system. Xenografts enabled correct predictions for resistance in 96% and for tumor response in 90% of cases. Despite great efforts to obtain a large number of comparisons, only 32 test results were available before the patients needed treatment. Therefore, our study has shown that the xenograft system will not have practical significance in determining individual patient treatment. Limitations are the duration of testing, the take rate of only 50% and the charges for nude mice. However, the highly correct prediction rates for tumor sensitivity and resistance validates human tumor xenografts as tumor models for testing new drugs and combinations. Furthermore, xenografts can be used as tumor source for experiments in cell culture.