Relative impact of T4 and N2 on the efficacy of 3 versus 6 months of adjuvant CAPOX for high-risk stage II and stage III colon cancer: ACHIEVE and ACHIEVE-2 trials

Background Results in the IDEA Collaboration have allowed clinical practice guidelines to prioritise 3-months (3m) of CAPOX for the adjuvant chemotherapy of low-risk (T1-3 and N1) stage III colon cancer, whereas 6-months (6m) of oxaliplatin-based treatment is recommended for high-risk stage III (T4 and/or N2) cancer. As for high-risk stage II, 3m CAPOX is not the most preferred regimen in present guidelines. The purpose of this study was to clarify the relative impact of T4 and N2 on the efficacy of two durations (3m vs 6m) of CAPOX. Methods In the combined data from ACHIEVE (n=1,291) for stage III and ACHIEVE-2 (n=514) for high-risk stage II, patients with CAPOX were analysed. We firstly conducted two interaction tests: “duration by T-stage (T4 vs non-T4)” and “duration by N-stage (N2 vs non-N2)”. Conventionally an interaction test was statistically significant if an interaction P was < 0.10. Then we compared the disease-free survival (DFS) of 3m vs 6m CAPOX according to T- or N-stage. Results In the two trials combined, 1,401 patients were treated with CAPOX (3m, 702; 6m, 699). Interaction P -values for duration by T-stage and by N-stage were 0.083 (significant) and 0.152 (not significant), respectively, suggesting that T-stage was associated with a stronger interaction than N-stage. This led to the comparison of 3m vs 6m CAPOX according to T-stage. Of 1,401, T4 was observed in 423 patients (i.e., T4N any ; referred to as “T4 tumours”), where 3m CAPOX was worse than 6m with a 3-year DFS of 67% vs 73% and a hazard ratio (HR) of 1.19 (3m/6m; 95% CI, 0.85 to 1.67). In the remaining 978 patients with non-T4 (i.e., T3N0 and T1-3N1-2; “non-T4 tumours”), 3m CAPOX was not worse than 6m with a 3-year DFS of 91% vs 87% and an HR of 0.78 (0.56 to 1.10); of note, even in 151 patients with non-T4 and N2 (i.e., T1-3N2; “T1-3N2 tumours”), 3m CAPOX was not worse than 6m with a 3-year DFS of 76% vs 71% and an HR of 0.88 (0.50 to 1.56). There were 126 and 116 patients with a documented initial relapse in the T4 and non-T4 tumours, respectively. Of those, metastasis of the peritoneum was significantly more frequent in the T4 tumours (n=43; 34%) than in the non-T4 tumours (n=14; 12%) (P =0.001). The same results held in the comparison between the T4 and T1-3N2 tumours. Conclusion The T4 tumours had a different metastatic spread from the non-T4 tumours. Despite the study being underpowered, our results suggest that 3m CAPOX may be sufficient for patients with a non-T4 tumour; in particular, 3m CAPOX may be sufficient for patients with a T1-3N2 tumour although 6m oxaliplatin-based therapy is currently recommended for this group. Similarly, 3m CAPOX may be adequate for T3N0 patients with a high-risk feature. Further investigations with a larger number of CAPOX patients will be warranted. Legal entity responsible for the study Japanese Foundation for Multidisciplinary Treatment of Cancer. Funding The two phase III studies of ACHIEVE and ACHIEVE-2 were funded by Yakult Honsha Co., Ltd. Disclosure Takeharu Yamanaka has an affiliation with Grant/Research Support: Chugai Pharma, Daiichi-Sankyo, Bayer.