Author response to: Complete mesocolic excision for right colonic cancer: prospective multicentre study

Le résumé fourni par la source

Dear Editor We thank Akyol et al. for their interest in our study1. For its interpretation it is crucial to appreciate the quasi-experimental regression discontinuity design. This implies that the surgery is not standardized, but a continuous spectrum of operations is assessed, divided into two groups by a single variable (superior mesenteric vein dissection). The advantage is that the irreversibility of knowledge is avoided: a surgeon trained in complete mesocolic excision (CME) will inevitably perform an operation that adopts CME principles, even for the control group. Thus, only for non-trained surgeons does the control group reflect current routine care (high external validity). The drawback is that because standardization is impossible the CME group does not reach the maximum quality of specialized centres and the study does not test the pure concept of CME (reduced internal validity). So, the trade-off was to either optimize the CME group or the control group. We opted for the latter. In addition, we wanted to include a large number of centres, which was accomplished by a pragmatic registry-based design. The problem with this was that follow-up was very good for overall survival but much less so for all types of recurrences. With more than 1000 patients (18.3 per centre) included we do not agree that is a low rate. The definition of central nodes is given in the methods (central 4 cm of ileocolic pedicle). More specimen measures will be published in a separate article. We agree with the critique of the conclusions and think that our data are best interpreted as of likely benefit in stage III. Together with the current meta-analysis this would justify routine use of CME. Unfortunately, we did not predefine stage-specific survival as an endpoint. Therefore, the editor would not accept a positive conclusion even though evaluating stage III separately seems not an arbitrary post hoc analysis.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.