Reply: Rethinking sustained recompensation in MASLD cirrhosis: The value of pharmacological maintenance

We appreciate the interest of Dr Toshikuni in our recently published study evaluating the rate and predictors of recompensation in patients with decompensated cirrhosis due to metabolic dysfunction–associated liver disease (MASLD),1 and for highlighting the clinical relevance of our findings, particularly the prognostic benefit of recompensation and the importance of targeting its underlying drivers. Several of the comments raised offer thoughtful and well-reasoned considerations. Hepatic recompensation was originally described as an observed clinical state in which some patients with decompensated cirrhosis achieve sustained improvement after control or removal of the etiological factor. While the Baveno criteria represent an important effort to standardize this concept, the current definition has notable limitations.2 First, the definition remains insufficiently precise, particularly regarding what constitutes a meaningful “improvement in liver function.” Second, it introduces subjectivity, as it relies on clinical resolution and treatment withdrawal without standardized criteria. Third, its applicability across etiologies is uncertain, being only feasible in diseases where etiological cure can be easily measured. Importantly, the natural history of cirrhosis is highly heterogeneous. The probability of recompensation is influenced not only by the underlying etiology but also by the type of prior decompensation event. Patients with ascites may follow a different course than those with variceal bleeding or encephalopathy, a variability not fully captured by current criteria. Moreover, emerging evidence questions the need for complete treatment withdrawal to achieve the prognostic benefit of recompensation3,4, suggesting that this concept may be better defined by clinical outcomes rather than therapeutic milestones. From this perspective, we propose that recompensation should be primarily understood as a state associated with improved prognosis. Identifying patients who reach clinically stable conditions with outcomes comparable to compensated stages—regardless of strict adherence to current criteria—may provide a more pragmatic and patient-centered framework. This distinction is especially relevant in MASLD and other noncurable etiologies. In these conditions, complete etiological control is often difficult to achieve or define, particularly in advanced disease stages where the primary driver becomes less clearly identifiable. Under the current definition, recompensation may therefore represent an elusive or even unattainable state. However, in our study, a subset of patients with MASLD achieved sustained clinical stability and functional improvement without fulfilling strict etiological control criteria or complete treatment withdrawal. Importantly, these patients showed improved survival and reduced liver-related mortality and transplantation, supporting a clinically meaningful benefit beyond current definitions. These observations highlight the need to refine recompensation criteria, especially in noncurable etiologies. Whether recompensation should be defined by etiological cure and fibrosis regression or by control of key drivers of decompensation, such as portal hypertension or systemic inflammation, remains an open question. In conclusion, the Baveno criteria provide a valuable foundation, but further refinement—focused on objective measures, prognostic relevance, and etiological diversity—is needed to better capture this clinically meaningful phenomenon.