Shi, Qian

3501 Background: 6m of oxaliplatin-based treatment is an option as adj chemotherapy for patients with high risk stage II CC (T4, inadequate nodal harvest, poorly differentiated, obstruction, perforation or vascular/perineural invasion). The IDEA collaboration showed shorter treatment duration to be appropriate for most pts with stage III colon cancer. The results of the 4 IDEA studies with stage II pts are presented here. Methods: A prospective, pre-planned pooled analysis of high-risk stage II patients from 4 concurrently conducted randomized phase III trials (SCOT, TOSCA, ACHIEVE-2, HORG) was performed to evaluate non-inferiority (NI) of 3m compared with 6m (ref) of adj FOLFOX/CAPOX (regimen preselected, not randomized). The primary endpoint was disease-free survival (DFS), NI was to be declared if the 2-sided 80% confidence interval (CI) for DFS hazard ratio (HR 3m v 6m) estimated by a stratified Cox model was below 1.2. 542 DFS events were required to provide 80% power to declare NI. NI was also examined within regimen, T4 (Yes v No) and inadequate nodal harvest (Yes v No) as pre-planned subgroups. Results: The primary analysis included 3273 randomised pts of which 1254 had FOLFOX and 2019 had CAPOX. There were 552 events and the median follow-up was 60.2 m. There was significantly less grade 3-5 toxicity with 3m treatment (p < .0001). The 5-year DFS rate was 80.7% and 84.0% for 3m and 6m treatment with an estimated DFS HR of 1.18 (80% CI:1.05-1.31, p for NI = 0.404). For CAPOX the estimated HR was 1.02 (80% CI: 0.88-1.17, p for NI = 0.087) and for FOLFOX the estimated HR was 1.42 (80% CI: 1.19-1.70, p for NI = 0.894). The test for interaction between duration and regimen was not statistically significant (p = .174 adjusted for multiple testing) but was stronger than that for the other subgroups examined. Conclusions: In the overall population non-inferiority for 3m adj treatment in pts with high-risk stage II CC was not shown. As with the stage III population the choice of adj regimen appears important (although this did not reach statistical significance) with a small difference in DFS between 3 and 6 m treatment if CAPOX is used. Clinical trial information: ISRCTN59757862.

Predicting the times of milestone events, ie, interim and final analyses in clinical trials, helps resource planning. This manuscript presents and compares several easily implemented methods for predicting when a milestone event is achieved. We show that it is beneficial to combine the predictions from different models to craft a better predictor through prediction synthesis. Furthermore, a Bayesian approach provides a better measure of the uncertainty involved in prediction of milestone events. We compare the methods through two simulations where the model has been correctly specified and where the models are a mixture of three incorrectly specified model classes. We then apply the methods on two real clinical trial data, North Central Cancer Treatment Group (NCCTG) N0147 and N9841. In summary, the Bayesian prediction synthesis methods automatically perform well even when the data collection is far from homogeneous. AnRshiny app is under development to carry out the prediction in a user‐friendly fashion.