Letter: Concerns Regarding the Paradoxical Renal Outcomes in Patients Switching From Entecavir to Tenofovir Alafenamide in the Study by Ogawa et al.

We read with great interest the article by Ogawa et al. [1], “Five-Year Effectiveness and Renal Safety Following Switching to Tenofovir Alafenamide in Chronic Hepatitis B Patients With Renal Impairment.” The authors are to be commended for providing valuable real-world, 5-year data on the safety and efficacy of tenofovir alafenamide (TAF), particularly in a significant cohort of patients with pre-existing chronic kidney disease (CKD). The study robustly demonstrates the high virological efficacy and overall renal safety of sequential TAF therapy. However, we wish to highlight a paradoxical finding presented in Figure 4, which details the longitudinal changes in estimated glomerular filtration rate (eGFR). The data indicate that patients previously treated with entecavir (ETV) experienced a more pronounced eGFR decline over 5 years compared to those previously treated with tenofovir disoproxil fumarate (TDF) or a NUC combination. Specifically, in the CKD group (Figure 4A), the prior ETV group's eGFR declined by −1.8 mL/min/1.73 m2, whereas the prior TDF/Combo group's eGFR improved by +0.6 mL/min/1.73 m2. A similar trend was observed in the non-CKD group (Figure 4B), with the ETV group showing a greater decline (−3.9 mL/min/1.73 m2) than the TDF/Combo group (−3.0 mL/min/1.73 m2) [1]. This finding is counter-intuitive. TDF is well-known for its greater potential nephrotoxicity compared to ETV [2], and the observed “improvement” in the TDF/Combo group likely represents renal function recovery or “rebound” after cessation of TDF. Conversely, the ETV group, having switched from one renally-safe agent (ETV) to another (TAF) [3], had no such recovery potential. Their eGFR decline may simply reflect the natural, age-related decline, as the authors suggest. More critically, this discrepancy strongly suggests the presence of indication bias. It is highly plausible that clinicians, in the years prior to TAF availability, preferentially prescribed ETV (over TDF) to patients who were already perceived to have a higher baseline renal risk profile (e.g., older age or comorbidities like diabetes and hypertension). While the authors thoughtfully employed propensity score matching (PSM) to balance the CKD and non-CKD groups, this matching was limited to age, sex, and cirrhosis. It did not account for differences between the prior-NUC subgroups (ETV vs. TDF/Combo). The baseline characteristics in Table 1 already show that the CKD group was significantly older, and it is likely that other unmeasured confounders exist within the prior-ETV cohort. Therefore, to clarify this crucial point, we would like to ask the authors: Were there significant differences in the baseline prevalence of key comorbidities, such as hypertension and diabetes mellitus, between the prior ETV group and the prior TDF/Combo group? Did the authors collect data on the clinical reason for the initial NUC selection, or the reason for the switch to TAF? (e.g., was the TDF group switched primarily for renal/bone safety concerns, while the ETV group was switched for virological reasons, such as the low-level viremia mentioned in the methods?) Clarifying this paradox is essential. Without this context, the data in Figure 4 could be misinterpreted as suggesting that prior ETV treatment leads to poorer long-term renal outcomes after switching to TAF. We believe the more likely explanation lies in fundamental differences in the baseline patient characteristics established years ago by indication bias rather than a negative interaction between prior ETV and current TAF therapy. Yours faithfully The authors declare no conflicts of interest. The authors will be able to provide data as needed.