Second Surgery Slows Recurrent Ovarian Cancer

CHICAGO -- The duration of progression-free survival (PFS) in relapsed ovarian cancer increased by more than a third in patients who had repeat surgery at relapse, a randomized trial showed.

Patients who underwent a second cytoreductive procedure before chemotherapy had a median PFS of 19.6 months as compared with 14.0 months who received chemotherapy without surgery. However, the difference met the prespecified level of statistical significance only in the subgroup of patients who had complete resection (no residual disease), according to Andreas Du Bois, MD, of Center for Gynecology and Gynecologic Oncology (ASCO) in Essen, Germany, and colleagues.

A preliminary analysis of overall survival survival (OS) -- the primary endpoint -- showed a higher-than-expected survival in all patients at 2 years and no significant difference between treatment groups, the authors reported here at the American Society of Clinical Oncology annual meeting.

"A benefit of surgery was exclusively seen in patients with complete resection, indicating the importance of selecting both the right center, with the capability to achieve a complete resection in a majority of patients, and the right patients," Du Bois said. "Hopefully, further follow-up will show that this benefit translates into an overall survival advantage."

The results are exciting but not definitive, said ASCO invited discussant Ritu Salani, MD, of the Ohio State University in Columbus. She pointed out that two other randomized trials evaluating secondary cytoreduction are ongoing, and that the current study raised some questions that require answers.

"The results complement the findings of retrospective studies and emphasize the importance of complete resection," said Salani. "There were 67 patients who were not completely resected. Are there any identifying factors for these patients? Should we consider using minimally invasive or less invasive approaches to surgery? Should patients receive maintenance therapy?"

Although surgery and chemotherapy form the basis for initial treatment of ovarian cancer, the role of surgery for recurrent disease remains controversial. Several retrospective and nonrandomized studies suggested a potential benefit of secondary cytoreductive surgery, leading to two prospective, randomized phase III trials by cooperative clinical research groups in the U.S. and Europe.

Du Bois reported initial findings from the DESKTOP III trial, sponsored by the German Society of Gynecological Oncology (AGO) and conducted with other collaborative groups in Europe and Asia. The trial followed two previous AGO studies designed to develop and evaluate a scoring system to identify patients who achieved complete resection with secondary cytoreductive surgery.

DESKTOP I provided data for the AGO score, which comprises good performance status, complete resection during initial surgery, and ascites <500 mL. DESKTOP II validated the AGO score, showing that it could predict complete resection with 95% probability in more than two-thirds of patients.

DESKTOP III applied the AGO score in a randomized trial, assigning patients with recurrent, platinum-sensitive ovarian cancer to receive immediate platinum-based chemotherapy or secondary cytoreduction followed by chemotherapy. The primary endpoint of OS was assessed after 244 OS events. A planned interim analysis of OS occurred after 122 OS events. Secondary endpoints included PFS and resection rate.

Data analysis included 407 patients, who had a median age of about 61. Three-fourths of patients had FIGO stage IIIB/IV disease, 80% had grade 2/3 serous histology, 90% had received prior platinum-taxane combination chemotherapy, and three-fourths had a platinum-free interval (PFI) exceeding 12 months. The median PFI was about 20 months.

Pooled data at the interim analysis of OS showed a 2-year postrandomization OS of 83%, far better than the expected 55% to 66% and no difference between the two groups. No patient in either arm died within 30 days, and a total of six patients had died at 6 months (five in the chemotherapy group).

Overall, the analysis of median PFS showed a 34% reduction in the hazard ratio in favor of the surgery group, but difference that did not meet the prespecified level of significance.

A planned analysis of PFS by surgical outcome showed that patients with a complete resection (71% of patients randomized to surgery) had a median PFS of 21.2 months, which was significantly better than the median for patients randomized to immediate chemotherapy (HR 0.56, 95% CI 0.43-0.72, P<0.0001). In contrast, patients with residual disease after surgery had a median PFS of 13.7 months, no different from the chemotherapy-alone arm.

The prespecified endpoint of time to first subsequent therapy (third-line therapy) also favored the surgery arm with a median of 21.0 months versus 13.9 months for the chemotherapy group (HR 0.61, 95% CI 0.48-0.77, P<0.001).

Comment