E. Trépo

Self-expanding metal stent should be considered in patients with cirrhosis and uncontrolled variceal bleeding.

We read the article by Escorsell et al. with great interest. This is the first randomized, controlled trial comparing self-expanding metal stent (SEMS) and balloon tamponade in acute refractory variceal bleeding in patients with cirrhosis. Although this study showed that SEMSs have greater efficacy in the control of bleeding and present less serious adverse events than does balloon tamponade, it failed to demonstrate a survival benefit. Even if randomized, controlled trials are considered the best way to assess the impact of an intervention, this is probably not the case when SEMSs are used to control acute refractory variceal bleeding for the following reasons. First, blinding the therapeutic intervention was not possible. Second, the investigators used patients treated with balloon tamponade as controls despite high rates of serious adverse events and rebleeding after balloon deflation. Thus, patients treated with balloon tamponade were not perfect controls. Third, patients enrolled in this study likely differed from the average patient seen in daily practice. This statement is supported by the 11 exclusion criteria mentioned in the study design and by the low number of patients included during a 3-year period, despite the many centers participating in the study. This is a clear limitation of this study that reduces the robustness of the conclusions. In the specific setting of acute refractory variceal bleeding, the results from observational studies may seem more relevant to clinical practice. In line with this comment, a recent meta-analysis (which already included Escorsell’s results) assessed the usefulness of SEMS in patients with cirrhosis and severe or refractory variceal bleeding. This metaanalysis showed that failure to control bleeding occurred in 18% of cases. Fewer than 40% of patients treated with SEMS died after 30 days and only 12% died from recurrent bleeding. Even if these results are not a proof that SEMS reduce mortality, this percentage compares favorably to the mortality rates reported in previous studies. Furthermore, a significant percentage of patients had access to transjugular intrahepatic portosystemic shunt (26%) or to liver transplantation (10%), which underlines that SEMS can also serve as a bridge to a more definitive treatment. Overall, we believe that SEMS should be considered in patients with cirrhosis and uncontrolled variceal bleeding. How early SEMS should be placed during the course of acute variceal bleeding to achieve better control of bleeding and prevention of rebleeding before the development of more-severe liver dysfunction remains to be assessed.