Meta-analysis of chemotherapy regimens for ovarian carcinoma: a reassessment of cisplatin, cyclophosphamide and doxorubicin versus cisplatin and cyclophosphamide.

OBJECTIVE: To compare three year survival, median survival and improved longevity with the addition of doxorubicin to a chemotherapy regimen of cisplatin and cyclophosphamide used in the treatment of ovarian cancer and to integrate this with a previous meta-analysis that compared three year survival. METHODS: Twenty-three studies that evaluated either the control or test arms were combined for meta-analysis. Five studies were randomized with both arms. Inclusion criteria consisted of median survival data, three year survival data, no previous chemotherapy or radiation and adequate follow-up. The data were analyzed with a twotailed t test, a fixed effects odds ratio, a random effects odds ratio, logistic regression modeling for three year survival and standard regression modeling for median survival. RESULTS: A statistically significant improvement in three year survival was demonstrated with the fixed effects odds ratio analysis combining the five prospective randomized studies and with logistic regression model of all the studies. Random effects odds ratio and the two-tailed t test failed to show statistical significance. Standard regression modeling demonstrated statistically significant improvement in median survival for a doxorubicin dose intensity of 40 mg/m2 and near significance for a doxorubicin dose intensity of 50 mg/m2. Median survival was improved by 1.91 months with the addition of doxorubicin to the cisplatin/cyclophosphamide regimen. CONCLUSION: Although there appears to be statistically significant improvement in three year survival and median survival with the addition of doxorubicin to the cisplatin/cyclophosphamide regimen for ovarian cancer, the actual improvement in median survival is less than two months and therefore, the added toxicity of doxorubicin may not be warranted.