Letter: Trajectories of Hepatic Function Markers Enable Early Prediction of Hepatic Recompensation—Authors' Reply

We sincerely thank Hofer and colleagues for their insightful commentary 1 on our study describing the concept of ‘unstable recompensation’ in hepatitis B virus (HBV)-related decompensated cirrhosis 2. Their letter thoughtfully underscores the evolving paradigm of hepatic recompensation as a dynamic process, wherein serial assessment of laboratory markers provides superior prognostic information compared with static baseline evaluations. Importantly, we appreciate the insightful emphasising the monitoring of hepatic function trajectories rather than a baseline level. Whether MELD predicts hepatic recompensation appears to be highly dependent on the disease state at the time of assessment. In the previous studies by Hofer et al., patients were evaluated during a relatively stable phase after a period of effective etiological control, in which higher MELD values most likely reflected impaired and largely irreversible hepatic functional reserve, thereby being associated with a lower likelihood of recompensation 3. In contrast, in our study, higher baseline MELD levels may largely reflect acute necroinflammatory activity and disease instability, representing a potentially reversible process rather than fixed hepatic dysfunction. Accordingly, longitudinal changes in hepatic function markers provide more meaningful information for characterising recompensation trajectories. Moreover, predictors of recompensation vary across etiologies: albumin ≥ 34 g/L at 24 weeks in HBV cirrhosis 4, lymphocyte-to-monocyte ratio in primary biliary cholangitis 5, and IgG, bilirubin, and platelet levels in autoimmune hepatitis 6. Furthermore, regarding portal hypertension, our study found that patients whose initial decompensation subtype was variceal bleeding had lower recompensation rates and a worse prognosis compared to those presenting with ascites 7. This indicates that portal hypertension may persist despite clinical improvement. In terms of non-invasive monitoring, our data showed that baseline platelet counts correlate with stable recompensation, but 6-month trajectories do not provide additional prognostic value, which may reflect the delayed or incomplete reversibility of portal hypertension despite clinical recompensation as the authors mentioned. We also appreciate the authors' discussion of the concept of ‘unstable recompensation’. Our proposed concept of ‘unstable recompensation’ emerged from long-term patient monitoring and highlights that recompensation is neither binary nor unidirectional, but a changing state requiring serial evaluation. While Baveno VII and expansion of Baveno VII recompensation criteria recommend a one-year assessment window 8, 9, our previous data suggest that extending observation to 2 years may better capture long-term stability 10, indicating that optimal timing for evaluation deserves further refinement. In summary, we extend our gratitude to Hofer and colleagues for reinforcing the dynamic nature of hepatic recompensation. The serial assessment of hepatic function markers provides a more refined prognostic perspective than baseline evaluations alone. Future efforts should focus on validating aetiology-specific predictive models and optimising the timing of reassessment to improve early identification and management of patients with decompensated cirrhosis. We look forward to continued research that further elucidates the criteria and clinical pathways of recompensation across diverse etiologies. Shuai Xia: writing – original draft. Bingqiong Wang: writing – original draft. Xiaoning Wu: writing – review and editing. Hong You: writing – review and editing. The authors have nothing to report. The authors declare no conflicts of interest. This article is linked to Xia et al. papers. To view these articles, visit https://doi.org/10.1111/apt.70430 and https://doi.org/10.1111/apt.70521. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.