Reply.

Abstract

We sincerely thank Marot and coworkers for their letter. We agree with them and their meta‐analysis concluding that self‐expanding metal stents provide an effective alternative to balloon tamponade in selected patients with acute variceal bleeding (AVB), in whom esophageal stent can be used as a bridge to rescue therapies (mainly transjugular intrahepatic portosystemic shunt). Our primary endpoint was to compare the efficacy and safety of the two devices in AVB from esophageal varices refractory to medical and endoscopic therapy, which is the only accepted setting for the use of balloon tamponade (massive AVB or AVB uncontrolled by standard treatment).1 Being an accepted treatment, for ethical reasons we could not omit the use of esophageal balloon tamponade as a comparator, just to have a perfect control with regard to efficacy. We differ from Marot et al. on their appreciation of the applicability of our results. In real life, we will also treat with stents patients with previous/malpositioned esophageal balloon tamponade (who were not allowed in the trial as this would have precluded obtaining conclusive results on adverse events). In fact, if we had included the 23 patients excluded due to previous balloon tamponade, our initially planned sample size would have been achieved and even exceeded. Moreover, patients with known hepatocellular carcinoma were not excluded from receiving esophageal stents in our current practice but were excluded in the randomized study because of its potential impact on mortality. However, the remaining exclusion criteria were unavoidable when using balloon tamponade. As our study confirms the data from observational studies suggesting that esophageal stents are effective and acceptably safe in refractory AVB, it may be proposed that these might be used in other settings in the context of clinical trials, for instance, in the initial therapy of AVB, where esophageal stents may be compared to standard accepted therapy (combination of vasoactive drugs and endoscopic therapy); however, this would be a completely different study. We do not think that retrospective observational studies from small series with different and heterogeneous entry criteria may provide valid data with regard to comparative efficacy and side effects of these two treatments in this setting either.