A predictive index of cure versus no cure in advanced ovarian carcinoma patients--replacement of second-look laparotomy as a diagnostic test

Three hundred forty-two Stage III and IV epithelial ovarian carcinoma patients received cytoreductive surgery followed by Adriamycin and cisplatin, 50 mg/m2 each, q 4 weeks for 9 courses. One hundred ninety-seven were clinically NED at completion of treatment and 173 of these 197 had a second-look laparotomy. One hundred twenty had persistent disease. Fifty-three were second-look negative and had no further treatment. Thirty of these latter patients relapsed--all (with one exception) within 2 years. Those not relapsing after negative second-look are considered "cured" (median follow-up 42 months, range 24-68 months) and all others "failures." Stage was a significant predictor of treatment failure--there were no Stage IV "cures." In Stage III patients, age and largest residual tumor diameter post initial surgery were significant predictors of failure. Performance status was marginally significant. In our series, any patient with Stage IV disease or Stage III disease with at least two of the following three poor prognostic factors had a chance of cure of 2.2% (2 "cures" out of 90 patients): age greater than 60 years, macroscopic residual initially, or initial performance status of 2 or 3. Under normal circumstances a second-look procedure to identify persistent disease in this group of patients does not appear justified.