Objective: A non-randomized comparison of outcome in women undergoing second-look laparotomy (SLL) or clinical follow-up, after receiving six cycles of combination chemotherapy with paclitaxel plus either cisplatin or carboplatin, for optimally resected stage III ovarian cancer.
Methods: Prior to chemotherapy randomization, patients chose whether or not to undergo SLL; this was a stratification factor to insure balance of treatment assignment. Any subsequent therapy was physician-directed. Explanatory analysis replaced intent-to-treat because of a higher likelihood of detecting SLL effect in the presence of noncompliance.
Results: There were 393 patients (median age: 54) who Elected SLL and 399 (median age: 59) who Elected No SLL. The former group was more likely to have gross residual disease at initial surgery than the latter group (69% versus 60%, respectively). In the Elected SLL group, 59 (15%) patients subsequently refused surgery, in nine (2%) surgery was contraindicated, and 31 (8%) relapsed or died prior to the procedure. Cancer was found in 46% of 294 (75%) patients undergoing SLL. Since early failures (prior to SLL) do not address benefit, such patients (SLL: 32; No SLL: 33), defined as progression-free survival (PFS) < 6 months, were excluded from analysis. The adjusted relative risk of progression is 0.89 (95% confidence interval: 0.75, 1.07); the difference in median PFS is 1.0 month (SLL: 23.9 months; No SLL: 22.9 months). The survival rate curves are superimposable.
Conclusion: In the context of a non-randomized comparison, the performance of a SLL was not associated with longer survival.