Reply

We read with interest the letter from Vaz and colleagues and the comments on our recent study.1 As acknowledged in the letter, a major message of our paper is that transjugular intrahepatic shunt (TIPS) should be considered in carefully selected patients with cirrhosis older than 70 years. However, this implies the need for objective indicators to identify the patients more likely to have a good outcome. This is particularly relevant because these patients cannot access liver transplantation in case of further decompensation or progression of liver failure both facilitated by TIPS placement. In the derivation cohort of older adults, the AUC of ExPeCT was only slightly higher compared to that of Freibur index of post‐TIPS survival to predict liver‐related mortality. As suggested, this comparison was extended to the Model for End‐Stage Liver Disease (MELD) and MELD‐Na, and in this case the AUC was slightly, but not significantly, better at 1 year (ExPeCT 0.63, MELD 0.59, and MELD‐Na 0.60). Despite these similarities, it should be emphasized that the ExPeCT model was specifically derived in a population of patients who mainly received underdilated TIPS.2 This is in our opinion very relevant because we believe that TIPS undersizing represents the ideal treatment policy when features of frailty emerge, particularly in older adults where no “rescue strategy” exists in case of TIPS complications. Prediction of the prognosis of patients with decompensated cirrhosis, especially those treated with TIPS, remains difficult. Therefore, we agree that further studies with the incorporation of covariates related to nutritional status and sarcopenia3 are warranted. Unfortunately, data on sarcopenia were not available for the whole cohort included in the study. However, we are currently evaluating sarcopenia in a subgroup of our patients to possibly refine the performance of the ExPeCT model.