Letter: Impact of HCV Eradication on Recurrence Pattern and Long‐Term Outcomes in Patients With HCV‐Related Hepatocellular Carcinoma Undergoing Radiofrequency Ablation. Authors' Reply

Editors, We thank Drs. Xu and Pang for their interest in our study [1] and appreciate their valuable comments regarding the statistical methods and time-related bias in our previous analysis [2]. We acknowledge the importance of addressing these concerns to ensure the robustness of our findings. Our cohort comprised patients with varying statuses of sustained viral response (SVR), achieved at different time points relative to radiofrequency ablation (RFA) of hepatocellular carcinoma (HCC). In our recurrence analysis, we excluded patients with very early local tumour progression (LTP) or distant recurrence within 6 months of complete ablation, aiming to avoid interference from incomplete ablation or synchronous tumours not detectable in early imaging [3]. Current guidelines recommend that patients with HCC eligible for curative treatment, such as resection or ablation, defer direct-acting antiviral (DAA) therapy for 4–6 months post-treatment, but delaying DAA therapy beyond 1 year is generally discouraged [4, 5]. Thus, we excluded patients with a prolonged delay in SVR achievement beyond 1 year, which aligned with global clinical practice and ensures a more relevant real-world analysis. Additionally, patients who developed LTP (n = 0) or distant recurrence (n = 1) before SVR achievement were excluded. Within the selected timeframe, the relatively small delay in SVR achievement after complete ablation further suggests that the immortal time bias was likely minimal. Nevertheless, we agree that immortal time bias is an inherent challenge in retrospective studies [6], and time-dependent covariate Cox analysis may offer an alternative approach to address the potential bias by time-dependent covariate [7]. In response to the concerns raised, we performed a time-dependent covariate Cox regression analysis, including all patients regardless of the timing of SVR achievement. This analysis treated SVR status as a time-varying covariate, ensuring that patients remained in the non-SVR group until the initiation of antiviral therapy, after which they transitioned to the SVR group. All other exclusion criteria and covariates remained consistent with our previous model. In line with our earlier findings, the time-dependent analysis confirmed that SVR achieved through DAA therapy was independently associated with improved distant recurrence (adjusted hazard ratio [AHR] = 0.493, p = 0.006), overall survival (AHR = 0.312, p < 0.001), and decompensation-free survival (AHR = 0.466, p = 0.014) (Table 1). It is worth noting that the risk of HCC recurrence typically decreases over time after the first year [8-10]. Including patients who achieved SVR later might have inadvertently included a group of patients who experienced a longer period of recurrence-free survival before receiving DAA treatment. This could potentially overestimate the benefits observed in the DAA-treated cohort, introducing another form of immortal time bias. In summary, our study provides compelling evidence supporting the benefits of timely antiviral therapy in HCC patients cured by RFA. While immortal time bias remains a potential limitation of retrospective studies, the consistent results across both our initial and time-dependent Cox regression analyses reinforce the validity of our findings. Kuo-Cheng Wu: writing – original draft, investigation, formal analysis, methodology. I-Cheng Lee: writing – review and editing, conceptualization, methodology, investigation, formal analysis, supervision, resources. The authors' declarations of personal and financial interests are unchanged from those in the original article.8 This article is linked to Wu et al papers. To view these articles, visit https://doi.org/10.1111/apt.18199 and https://doi.org/10.1111/apt.18345. 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.