Renal function and urinary tubular biomarker abnormalities in untreated and entecavir- or tenofovir disoproxil fumarate–treated patients with chronic hepatitis B: a retrospective cross-sectional study

Abstract

Background Renal safety is an important consideration during nucleos(t)ide analogue (NA) therapy for chronic hepatitis B (CHB). Tenofovir disoproxil fumarate (TDF) may affect proximal tubular function, while serum creatinine and estimated glomerular filtration rate (eGFR) primarily reflect glomerular filtration. This study compared glomerular-function measures and urinary tubular biomarker patterns among NA-untreated, entecavir (ETV)-treated, and TDF-treated patients with CHB. Methods This single-center retrospective cross-sectional study included 242 adults with CHB assessed between June 2025 and June 2026: 113 NA-untreated, 64 ETV-treated, and 65 TDF-treated participants. For treated participants, the single index assessment was obtained after at least 12 months of continuous corresponding monotherapy. Serum creatinine, eGFR, urinary α1-microglobulin (A1M), urinary β2-microglobulin (B2M), and urinary microalbumin (U-mAlb) were compared among groups. eGFR was calculated using the 2009 CKD-EPI creatinine equation and analyzed primarily as a continuous measure; eGFR <90 mL/min/1.73 m2 was retained as a secondary screening outcome. Urinary biomarkers were measured as spot-urine concentrations without urinary creatinine normalization. Results Urinary A1M ≥ 12.0 mg/L occurred in 13.3%, 42.2%, and 67.7% of NA-untreated, ETV-treated, and TDF-treated participants, respectively; corresponding frequencies of urinary B2M > 0.30 mg/L were 25.7%, 48.4%, and 75.4%. Both A1M and B2M concentration-threshold exceedance were more frequent in TDF-treated than ETV-treated participants (Bonferroni-adjusted P = 0.0108 and P = 0.0048, respectively), and continuous B2M was also higher with TDF (adjusted P = 0.0074). Median eGFR was 112.98 (102.47–119.36), 104.79 (92.08–112.39), and 104.47 (91.24–116.58) mL/min/1.73 m2, respectively (overall P = 0.0008). Both treated groups had lower continuous eGFR than the NA-untreated group, whereas serum creatinine, continuous eGFR, and the frequency of eGFR <90 mL/min/1.73 m2 did not differ between ETV and TDF (all Bonferroni-adjusted P = 1.0000). Conclusion TDF-treated participants showed a more prominent urinary A1M/B2M concentration-based biomarker pattern than ETV-treated participants, whereas conventional glomerular-function measures did not differ between the two treated groups. These findings represent cross-sectional associations rather than evidence of treatment-induced renal injury. The absence of pretreatment and serial renal measurements, urinary creatinine normalization, and a complete proximal tubular dysfunction assessment requires cautious interpretation and prospective longitudinal confirmation.