Purpose: To examine the prognostic factors associated with prolonged progression-free survival (PFS) and overall survival (OS) in 100 consecutively treated women undergoing autologous stem-cell transplant for advanced ovarian cancer.
Patients and methods: From October 1989 to February 1996, we transplanted 100 patients with ovarian cancer following chemotherapy with high-dose carboplatin, mitoxantrone, and cyclophosphamide with or without cyclosporine (n = 70); melphalan and mitoxantrone with or without paclitaxel (n = 25); or other regimens (n = 5). Their median age was 48 years (range, 23 to 65), 70% had papillary serous histology, 72% had grade III tumors, 66% were platinum-resistant, and 61% had > or = 1 cm bulk. The median number of prior regimens was two (range, one to six). Univariate and multivariate analyses were performed to examine age (< v > or = mean), stage, initial bulk, histology, grade, response to initial therapy, number of prior regimens, time from diagnosis to transplant, transplant regimen, platinum sensitivity, and bulk (< v > or = 1 cm) at transplant.
Results: The median PFS and OS times for the 100 patients were 7 and 13 months. A stepwise Cox proportional hazards model identified tumor bulk (P = .0001), and cisplatin sensitivity (P = .0249) as the best predictors of PFS. Age (P = .0017), bulk at transplant (P = .0175), and platinum sensitivity (P = .0330) provided the best prediction of OS. The median PFS and OS times for the 20 patients with platinum-sensitive, < or = 1-cm disease were 19 and 30 months. No differences in OS were seen when chemotherapy or surgery was used to achieve a minimal disease state.
Conclusion: Before consideration of high-dose therapy for recurrent/persistent advanced ovarian cancer, patients should undergo debulking surgery or chemotherapy to achieve a minimal disease state. Patients with platinum-resistant, bulky disease should not be transplanted. The optimal patients for this therapy may be those with minimal disease responsive to initial chemotherapy.