Author’s reply to the Letters to the Editor discussing main outcomes of the PRODIGE 23 study.

We thank Drs Moretto, Dr Li, Dr Socha and Dr Shubhra and their colleagues for their experts' comments of the PRODIGE 23 trial1,2. Their requests for clarification cover inclusion criteria, statistical methods, trial's interpretation and applicability of our results to lower risk patients. Concerning inclusion criteria, we agree that the age restriction to patients younger than 76 years is a limitation and we have discussed it in our publication. However, at the time we launched the PRODIGE 23 trial, there was no evidence indicating that adding oxaliplatin to adjuvant therapy would impact survival of elderly colorectal cancer patients and this was recently confirmed by a meta-analysis of 1,985 patients2. Moreover, there was no information on the use of mFOLFIRINOX in patients over 75 years old. There is now some data suggesting that mFOLFIRINOX is feasible in selected elderly patients through multidimensional geriatric assessment, probably with an initial dose adaptation and optimizing supportive care3. Beside the question of age, we agree that rare tumor histology subtypes might be considered and we previously answered to this comment4. PRODIGE 23 was open to all subtypes of rectal adenocarcinoma, yet, no patients with signet ring cell carcinoma and only 21 (5%) patients with mucinous carcinoma (11 in the standard-of-care group vs 10 in the TNT group) were included. Thus, we did not performed analysis on these small subgroups of patients. Finally, can we expand TNT with mFOLFIRINOX beyond traditionally high-risk patients? The randomized GRECCAR 12 phase III trial investigated this question in patients with cT2T3, N0-1, a tumor size ≤4cm, and low or mid rectal cancer to increase rectal preservation. Good responders to TNT or preoperative chemoradiation were offered a local excision, with mandatory TME in case of ypT3 or ypT2cN1. Preliminary results suggest that TNT with FOLFIRINOX increased 1-year organ preservation from 62.7% with chemoradiation to 71.6% with TNT (P=0.056)5.