Reply to “Methodological Considerations on Early Hemostatic Treatment for Spontaneous Rupture of <scp>HCC</scp> ”

We thank Professor Wen X [1] for his methodological comments on our clinical trial, which highlighted that early hemostatic treatment could improve survival in cases of spontaneous rupture of hepatocellular carcinoma (HCC) [2]. We appreciate the opportunity to address the concerns raised. First, we would agree that an immortal time bias could have influenced our results regarding the benefit of subsequent HCC treatments if the overall survival analysis had included only patients who survived long enough to receive them. However, all patients with a ruptured HCC—whether they survived the haemorrhage or not and whether they received subsequent treatment or not—were included in the analysis. We now provide two additional survival analyses: one including patients who survived more than 30 days (first quartile of time to first subsequent treatment: 39 days after bleeding) and another including patients who survived more than 84 days (median time to first subsequent treatment: 84 days). At 30 days or later after bleeding, 38 patients had received a subsequent treatment and 40 had not; at 84 days and beyond, 35 patients had received treatment and 25 had not. In both analyses, subsequent treatment remained a protective factor (Table 1). Second, we acknowledge the limitation related to the imaging diagnosis of HCC rupture. The multicenter retrospective design did not allow centralised rereading, but we applied clear diagnostic criteria to define rupture. Our study aimed to evaluate the real-life management of HCC diagnosed as ruptured, rather than to assess the accuracy of imaging interpretation in guiding hemostatic treatment. A centralised review could have excluded some cases, but doing so would have introduced selection bias for the real-life approach. Finally, as already mentioned, this was a retrospective study with inherent limitations, particularly variability in management across centres. Patients who received hemostatic treatment were more likely to have BCLC stage A/B HCC and higher transfusion rates. However, transfusion was not associated with prognosis, and transfusion rates did not differ between BCLC A/B (22%) and C/D (35%) (p = 0.105). Due to the small sample size and limited number of deaths, neither propensity score matching nor inverse probability weighting was feasible. In conclusion, we thank the reviewer for the constructive feedback and agree that our findings should be interpreted with caution given the retrospective nature of the study. Considering the low incidence of HCC rupture, conducting prospective and randomised studies in this context remains a major challenge. Apolline Commin and Isabelle Ollivier: writing and editing. Chloé Metivier and Remy Morello: statistical analysis and review. We thank all the reviewers for their assistance and support. The authors have nothing to report. The authors have nothing to report. The authors declare no conflicts of interest. This article is linked to https://doi.org/10.1111/liv.70396. All information required is given in the text.