• Potential conflict of interest: Nothing to report.

In light of our recently published multicenter European survey on the embolization of large spontaneous portosystemic shunts as a potential treatment option for patients with refractory hepatic encephalopathy,[1] we are highly pleased to read the confirmatory conclusions of Singh et al.

The authors confirm in their single-center case-series of 15 patients with refractory hepatic encephalopathy, including 10 with underlying cirrhosis, that embolization of spontaneous portosystemic shunts is efficient and safe. Of note in the cirrhosis group, the short- (at 2 months) and long-term (at on average 9 months) efficacy in the study of Singh et al. amounted to 90% and 60%, respectively, compared to 59.4 (at 3 months) and 49.6% (at on average 23 months postembolization) in our study. With regard to safety, incidence of complications was comparable. The authors suggest a Model for Endstage Liver Disease (MELD) of 15 as the cutoff to be used as a selection parameter for identification of potential candidates for this intervention, whereas in our series—using receiver operating characteristic (ROC) analysis and the Youden index—this cutoff was put at a maximal MELD score of 11.

Despite largely comparable findings, these two series differ in various aspects that are noteworthy in order to put the two datasets into the correct perspective.

The first distinction regards methodology. It is a well-known fact that case-series, which obviously lack comparison, can make things appear as if there is an association between an intervention and an outcome when, in fact, there is none. This is particularly the case when there is a small cohort of patients. This latter partially explains the heterogeneous results obtained in previously reported small single-center case reports or series.[2-6] In our series,[1] we wanted to overcome this particular problem by putting together the datasets of different European liver units leading to inclusion of 37 patients, resulting in the largest experience ever reported in this matter. As an additional advantage, this approach also omitted potential selection bias for this kind of procedure of an individual center

The second difference is that we opted to keep an open vizor with regard to selection of patients. More specifically, we opted to include cirrhosis patients with refractory hepatic encephalopathy without further restrictions (except for patients with iatrogenic shunts like transjugular intrahepatic portosystemic stent [TIPS] or conditions with known independent short-term ominous outcome like, for example, hepatocellular carcinoma). In this regard, we did not exclude patients with high MELD (range in our study 5 to 28) nor did we omit patients with renal insufficiency. This allowed a clear and unbiased evaluation of feasibility, efficacy, and safety of the intervention in a broad range of a homogenous group of patients. Based on this uncensored approach, we were able to perform uni- and multivariate analysis in a large enough population and make an estimate of the ideal cutoff point of the MELD score, which was retained as one of the predictive factors of recurrence of encephalopathy.

Third, our series also comprised sufficient follow-up to appreciate the long-term effect of intervention (overall follow-up in our series postembolization: 697 ± 157 days).

Overall, we appreciate that Singh et al. have substantiated our findings, namely, that embolization of large spontaneous portosystemic shunts for a refractory condition is feasible, safe, and effective. On the other hand, one should keep in mind that embolization of large spontaneous portosystemic shunts for a refractory condition should be reserved for a group of well-selected patients with sufficient critical functional liver mass (indirectly reflected by a MELD score ≤11) and should be done by a team with substantial experience to perform and manage these kinds of procedures.

  • Wim Laleman, M.D., Ph.D.1 Geert Maleux, M.D., Ph.D.2 Juan Cordoba, M.D., Ph.D.3 Frederik Nevens, M.D., Ph.D.1

  • 1Department of Liver and Biliopancreatic Disorders University Hospital Gasthuisberg KU Leuven, Leuven, Belgium

  • 2Department of Interventional Radiology University Hospital Gasthuisberg KU Leuven Leuven, Belgium

  • 3Department of Hepatology Hospital Vall d'Hebron Universitat Autònoma de Barcelona Barcelona, Spain