The prognostic significance of peritoneal seeding and size of postsurgical residual in patients with stage III epithelial ovarian cancer treated with surgery, chemotherapy, and high-dose radiotherapy.

OBJECTIVE: The aim of this study was to retrospectively analyze the prognostic importance of age, histologic type and grade, ascites, lymph node status, size and type of postoperative residual disease, and radiation dose on disease-specific (DSS) and progression-free survival (PFS) in stage III epithelial ovarian cancer patients who had been treated with radical surgery, postoperative chemotherapy, and high-dose radiotherapy. METHODS: Consolidation radiotherapy including whole abdominal radiation, pelvic, and upper abdominal boosts was employed in 46 patients who showed no evidence of residual or progressive disease after completion of multiagent chemotherapy. The median follow-up for all patients was 36 months and 103 months for patients at risk. The prognostic impact of pretreatment and treatment parameters on DSS and PFS was tested in univariate and multivariate analyses. RESULTS: The 5-year DSS and PFS rates for all patients were 38 and 33%, and for patients with 0-< or =2 cm residual tumor 65 and 61%, respectively. In univariate analysis, initial peritoneal seeding (both: P = 0.02), ascites (P = 0.03; 0.01), size of residual (0-< or =2 cm vs >2 cm), and residual miliary subdiaphragmatic (MDS) and localized peritoneal seeding (LPS) in the upper abdomen (P = 0.0002; 0.0003) were significantly correlated with DSS and PFS. Dose of radiation (< or =30 vs >30 Gy) correlated with DSS only (P = 0.02). In multivariate analysis size of residual disease (0-< or =2 cm vs >2 cm and/or MDS or LPS) remained the only independent prognostic factor for DSS and PFS (both; P = 0. 001). CONCLUSION: Patients with localized peritoneal seeding who were rendered free of disease elsewhere had an outcome equally poor as that of patients with gross residuals (>2 cm) in the upper abdomen. If our findings can be confirmed, attempted resection of all localized seeding in patients who are otherwise cytoreducible to no or minimal residual disease may be considered in combination with Taxol-containing regimens as are now being utilized for patients with gross disease.