Correlation of drug response with the ATP tumorchemosensitivity assay in primary FIGO stage III ovarian cancer

Objective. Our purpose was to: (a) study the in vitro chemosensitivity of p rimary epithelial ovarian cancer to drug combinations with cisplatin (CDDP) , carboplatin (CBDCA), paclitaxel (PTX), epirubicin (EPI), or cylophosphami de (CTX) utilizing the ATP tumorchemosensitivity assay (ATP-TCA); (b) corre late the test results with clinical response in patients with FIGO stage II I ovarian cancer; and (c) analyze the most useful parameters for interpreta tion of test results. Methods. CBDCA/CTX, CBDCA/PTX, CDDP/PTX, and EPI/PTX were tested in 93 fres h human primary epithelial ovarian cancer specimens. Correlations of in vit ro drug sensitivity/resistance and clinical response were performed in 38 p atients with FIGO stage III disease utilizing Fisher's exact test and by co mparison of progression-free (PFS) and overall survival (OS) between those testing as sensitive or resistant. A progression-free interval of more than 12 months following surgery was classified as clinical response. ATP-TCA r esults were analyzed using the median effective dose, area under the curve, or a defined sensitivity index. Results. Evaluable test results were achieved in 83 of 93 patients (89%). E PI/PTX had the highest in vitro activity (P < 0.001). In the clinical corre lation, 29 of 38 patients (76%) were classified as in vitro sensitive (sens itivity index [SI] <250) and 9 patients as in vitro resistant (SI > 250). T he SI was superior for interpretation of test results. Patients testing as chemosensitive had a significantly longer mean PFS (28.5 vs 12.6 months, P = 0.033) and OS (46.1 vs 17.6, P = 0.03) compared to those patients predict ed to be resistant. The assay demonstrated a sensitivity, specificity, and positive and negative predictive value of 95, 44, 66, and 89%, respectively (Fisher's exact test, P = 0.007). Conclusion. The observed in vitro efficacy of EPI/PTX in primary epithelial ovarian cancer specimens,warrants further clinical evaluation, The high ev aluability rate and the observed correlation with PFS and OS, within the li mitations of a nonrandomized study, support the use of the ATP chemosensiti vity assay in future prospective assay-directed trials. (C) 2000 Academic P ress.