Editorial: Caught in the Middle—Exploring the Grey Zone Between Decompensated and Recompensated <scp>HBV</scp> ‐Related Cirrhosis. Authors' Reply

We sincerely thank Professors Tsai and Fallahzadeh for their thoughtful editorial [1], ‘Caught in the Middle – Exploring the Grey Zone between Decompensated and Recompensated HBV-Related Cirrhosis’ on our recent publications [2]. Their insightful comments highlight the evolving understanding of hepatic recompensation and the complexity of the transition between decompensated and recompensated stages in chronic liver disease. Despite robust evidence that recompensation confers a survival benefit [3], its reported prevalence varies markedly among studies, reflecting the diverse and dynamic nature of the recompensation process [4, 5]. Recompensation should be viewed as an evolving clinical state shaped by ongoing changes in hepatic reserve, portal haemodynamics and aetiological control, rather than a one-time milestone. Most importantly, our data emphasise that the definition of recompensation needs further refinement. We would also like to clarify the distinctions within the unstable recompensation subgroup, which comprises two patient trajectories. Interestingly, the long-term prognosis appeared better in patients who experienced multiple episodes of decompensation before achieving recompensation than in those who decompensated after initial recompensation (Figure 1A–C). Regarding clinical features, baseline characteristics at the first episode of decompensation did not differ significantly between the two groups. Patients achieving recompensation after multiple decompensations tended to maintain slightly higher platelet counts and serum albumin levels, and lower Model for End-Stage Liver Disease (MELD) scores over time (Figure 1D–F), although these differences were not statistically significant. This might be explained by the possibility that patients achieving recompensation after multiple decompensations represent individuals with gradually restored hepatic reserve after repeated insults, whereas the latter may reflect patients with incomplete resolution of portal hypertension or ongoing aetiological activity leading to recurrent instability. Nevertheless, given the limited sample size, this observation requires validation in larger cohorts. Additionally, this pattern should be interpreted with caution, as survivorship bias could contribute: patients with recurrent decompensations are more likely (69.4% in our study) to progress to persistent decompensation or death. Another urgent unmet clinical need is to establish a precise and reliable model to predict recompensation, which could guide patient management and therapeutic decision-making. However, current models of recompensation mainly rely on baseline characteristics, such as baseline MELD, albumin, total bilirubin and alpha-fetoprotein [5-7]. Indeed, unstable recompensation may reflect a paradox between functional recovery and persistent structural or hemodynamic derangements [8]. Incorporating dynamic on-treatment variables and hemodynamic recovery indicators into future predictive frameworks may substantially improve the accuracy of identifying patients who will achieve recompensation and maintain stability [9]. Future work should integrate longitudinal clinical trajectories with haemodynamic assessment (HVPG or elastography), dynamic biomarkers such as changes in platelet count. Validation of dynamic thresholds and development of a composite, time-dependent risk score could enable early identification of patients at high risk for recurrence and guide intensified surveillance or preemptive interventions. Finally, we thank Drs Tsai and Fallahzadeh once again for their constructive perspective and for underscoring the clinical relevance of our findings. We hope that continued collaborative efforts will lead to a unified and biologically informed framework for assessing hepatic recompensation and its stability in patients with advanced liver disease. Bingqiong Wang: writing – original draft, writing – review and editing, validation, investigation. Shuai Xia: writing – original draft, software, visualization, investigation. Xiaoning Wu: writing – review and editing, supervision. Hong You: supervision, writing – review and editing, funding acquisition, conceptualization, project administration, resources. 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.70456. 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.