Nonalcoholic Fatty Liver Disease: A Special Consideration in Transjugular Intrahepatic Portosystemic Shunts?

The benefits of transjugular intrahepatic portosystemic shunts (TIPS) in decompensated cirrhosis have been widely demonstrated in randomized trials for the treatment of variceal bleeding and ascites. Although these results have greatly advanced the management of portal hypertension, pivotal trials to support these indications included mainly patients with alcohol-related liver disease,^{(1, 2)} an etiology of cirrhosis with an important degree of reversibility upon abstinence. In contrast, nonalcoholic fatty liver disease (NAFLD), a much less reversible entity and frequently associated with other comorbidities, has been grossly underrepresented in randomized trials for TIPS. Thus, whether patients with NAFLD can garner the same outcomes from TIPS, both in efficacy and safety, compared with other etiologies (alcohol and viral hepatitis) remains to be established.

In the current issue of Liver Transplantation, the Advancing Liver Therapeutic Approaches (ALTA) Study group presents data on the risk of renal dysfunction during the first month after TIPS and furthers the understanding of the impact of TIPS according to disease etiology.⁽³⁾ They studied 673 patients who were alive without a liver transplant for at least 1 month after TIPS insertion. Of these patients, 10% developed kidney dysfunction, with 3% requiring renal replacement therapy. NAFLD and the presence of diabetes mellitus were independently associated with post-TIPS renal dysfunction.

The first consideration of these results is that this incidence rate does not reflect the overall risk of post-TIPS acute kidney injury (AKI). There was a shift in TIPS indications from the initial full cohort to the cohort of 30-day survivors, with much less variceal bleeding and more refractory ascites in the latter. This likely reflects that many critically ill patients undergoing salvage TIPS who subsequently died or received liver transplantation were excluded from the cohort. Nevertheless, the validity of the data remains unchanged because it is in more stable patients under non–life threatening circumstances (with available alternatives) where understanding the balance in the risks and harms of TIPS becomes most relevant for therapeutic decisions. The second consideration is that being an observational study without a randomized control arm, it is difficult to infer the specific impact of TIPS on renal dysfunction, since the counterfactual outcomes if these patients had not received a TIPS cannot be estimated. Regardless, the study serves as a step forward in appreciating the specificities of TIPS use within the growing population of patients with NAFLD cirrhosis.

The granularity of the data was insufficient to achieve a full understanding of the increased susceptibility to renal dysfunction of patients with NAFLD and/or diabetes mellitus. Patients with NAFLD have increased risk factors for underlying parenchymal chronic kidney disease that might render them more susceptible to develop periprocedural AKI. In addition, the authors discussed the interesting hypothesis that patients with NAFLD could have higher chances of underlying silent heart dysfunction that could result in larger increases in right heart pressures and therefore in renal venous pressures. This may be worse in patients with obesity, with an already increased inferior vena cava pressure as a result of increased intra-abdominal pressure.⁽⁴⁾ Although body mass index (BMI) was not specifically reported in the current study, patients with NAFLD are expected to have higher BMI than other etiologies. Increased renal venous pressure is an increasingly recognized risk factor for kidney dysfunction⁽⁵⁾ and, as recently shown in patients with refractory ascites,⁽⁶⁾ can substantially contribute to decreased renal perfusion pressure. Although the case number was low, the authors showed that post-TIPS inferior vena cava pressures were, on average, 4 mm Hg higher in patients developing kidney dysfunction after TIPS. This might suggest that, different from other cirrhosis contexts, peri-TIPS albumin infusion might be deleterious because TIPS in itself already results in an immediate approximately 2-fold increase in central venous pressures.⁽⁷⁾ Albumin infusion could therefore contribute to worse renal venous congestion and an increased risk of pulmonary edema.⁽⁸⁾

The other main finding shown by the authors was that, among 30-day survivors without transplants, kidney dysfunction in the early post-TIPS period was associated with increased mortality. Even if this result should be taken with caution from a causal perspective, as only a limited adjustment for potential confounders was provided, it emphasizes the need for peri-TIPS kidney protective strategies. This approach may be especially important in patients with NAFLD and/or diabetes mellitus, with explicit orders to avoid nonsteroidal anti-inflammatory drugs for postprocedure pain, discontinue angiotensin receptor blockers or angiotensin-converting enzyme inhibitors, and hold diuretics in preparation for TIPS. Lastly, refinements in the TIPS technique, for example, the use of endovascular ultrasound to guide TIPS creation,⁽⁹⁾ might decrease the use of contrast and could be preferred in patients at risk of kidney dysfunction.

Finally, this study raises the question if other complications from TIPS may be more prominent in patients with NAFLD. Yin et al. demonstrated that diabetes mellitus was independently associated with the risk of hepatic encephalopathy in patients following TIPS creation.⁽¹⁰⁾ Although the mechanism behind this relationship is not completely clear, it can also translate to higher risk of post-TIPS encephalopathy in patients with NAFLD.

In summary, this study highlights the need for a better understanding of the tradeoffs between the benefits and risks of TIPS in patients with NAFLD cirrhosis. This is especially relevant in situations in which TIPS is not a last resort, life-saving procedure. Results of large cohorts such as the ALTA Study group are 1 step forward in this direction. However, there is a need for new randomized trials to define the role of TIPS in recurrent ascites and as a preemptive treatment in acute variceal bleeding, specifically in patients with NAFLD where clinical equipoise of TIPS compared with alternative standards of care still remains.