Letter: should colectomy be the end-point to evaluate the effectiveness of drug therapies in severe ulcerative colitis? Authors' reply



We thank Drs Domènech and Gisbert for their comments on our article.[1, 2] They note that some methodological issues might limit the application of our proposed score and decision tree.

First, we recruited patients with severe active ulcerative colitis (UC) (Lichtiger score ≥10) based on the criteria of the first double-blind RCT of continuous intravenous infusion of ciclosporin for severe active UC.[3] We consider that this is the usual clinical scenario when ciclosporin is used for patients with UC, whereas infliximab can be used for patients with both moderate and severe refractory UC.[4] In our article, neither cytomegalovirus reactivation, Clostridium difficile infection nor previous leucocytoapheresis was shown to affect the outcome of ciclosporin treatment.

Second, we consider that intravenous corticosteroids for 5 days is an acceptable option for treating severe UC based on the recent guidelines. The BSG guideline states that intravenous steroids are generally given for up to 5 days for acute severe UC.[5] The ECCO guideline mentions that extending corticosteroid therapy beyond 7–10 days carries no additional benefit for severe UC.[6]

Thirdly, we agree with their comments that it is not the need of colectomy, but rather the need for rescue therapies that should be the end-point for predictive studies in steroid refractory UC. It was not until 2010 that infliximab was approved for the treatment of moderate-to-severe active UC in Japan. Therefore, the need of colectomy should be the only appropriate end-point in our study that was performed from 2004 to 2011. If infliximab is considered as the third line of treatment in the usual clinical scenario, the need for rescue therapies could be the end-point for predictive studies.

The message we attempted to deliver through this article is that predictive formulae or decision trees can be constructed by the combination of simple clinical markers. Predictive formulae or decision trees for early decision-making should be modified by each institution, according to the differences in clinical background, the dosage and period of treatment and the differences in clinical end-points.


The authors' declarations of personal and financial interests are unchanged from those in the original article.2