Electronic medical records of CHB patients with ETV resistance in addition to prior LAM resistance were reviewed. Patients were treated with sequential ETV monotherapy due to genotypic LAM resistance. A total of 73 patients were treated with TDF alone or combination (with LAM or ETV) by clinician or patient choice from December 2012 to November 2015. All patients had hepatitis B surface antigen (HBsAg). Serum HBV DNA level was documented for at least 6 months before the start of LAM therapy. Patients with impaired renal function (serum creatinine > 1.5 mg/dL), autoimmune hepatitis, or antibodies against hepatitis C virus or human immunodeficiency virus (HIV) were excluded from this study. Additional criteria for exclusion were: pregnancy, lactation, and alcohol abuse (> 40 g/d ethanol). Diagnoses of chronic hepatitis and liver cirrhosis were based on liver biopsy features or on clinical, laboratory, and ultrasound data if liver biopsy data were unavailable. Written informed consents were obtained from all of patients participating in this study. This research was approved by the Institutional Review Board at the Ulsan University Hospital (UUH 2016-09-002, approved on Sep. 20, 2016).

Liver and kidney function tests were performed every three months during TDF based rescue treatment. HBV DNA levels were quantified using COBAS TaqMan HBV test (Roche, Branchburg, NJ, USA) with a lower detection limit of 12 IU/mL (60 copies/mL). Genotypic resistance to ETV was defined as the emergence of ETV resistance mutations in the presence of lamivudine resistance substitutions M204I/V ± L180M. When a viral breakthrough had developed before TDF-based rescue therapy, we tested a restriction fragment mass polymorphism (RFMP; Genematrix, Youngin, Korea) to identify LAM (rt180, rt204), ADV (rt181, rt236), and ETV (rt173, rt184, rt202, rt250). The specific HBV genotypes were identified using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) of the surface gene of HBV genome. The two fragments of the HBV genome between the nucleotide positions 2823 and 2845 and 61 and 80 were amplified by PCR. PCR products were treated with restriction enzymes. Patients underwent surveillance for HCC every 6 months with serial abdominal ultrasound and serum α-fetoprotein measurements.

A virologic response (VR) was defined as the absence of serum HBV DNA by PCR assay (< 12 IU/mL) on two consecutive measurements during TDF treatment. Hepatitis B envelop antigen (HBeAg) seroconversion was defined as the loss of HBeAg accompanied by the detection of anti-HBe. Partial virologic response (PVR) was defined as a decrease in HBV DNA of more than 1 log₁₀ IU/mL but with detectable HBV DNA after 6 months of antiviral therapy. Virologic breakthrough (VBT) was defined as a > 1 log₁₀ IU/mL increase in serum HBV DNA from the nadir on two consecutive measurements or in the last available measurement. Safety and tolerability were evaluated by the occurrence of adverse events (AEs), serious AEs (SAEs), laboratory abnormalities, discontinuation of the study drug due to AEs, or death. Specific markers of renal abnormalities included confirmed (defined as two consecutive visits) increase in serum creatinine of at least 0.5 mg/dL above the baseline value, serum phosphorus values < 2 mg/dL, and creatinine clearance < 50 mL/min.

Serum HBV DNA (IU/mL) levels were logarithmically transformed for statistical analysis. Continuous variables were compared using the one-way analysis of variance (ANOVA). Categorical variables were compared using the χ² test. The cumulative probability rates of clinical outcomes were calculated using Kaplan-Meier method. To identify the factors predictive of outcome among the baseline variables, the variables for clinical outcomes were compared using the χ² test or univariate logistic regression. Variables with a P-value of 0.1 or less in the univariate analyses were included in a Cox step-wise multivariate regression analysis using forward selection. To determine predictive value of baseline serum HBV DNA level, the hazard ratio (HR) and receiver operating characteristic (ROC) curves with the overall accuracy assessed by the area under the curve (AUROC) were used for the prediction of VR for TDF treatment. The AUROCs and their corresponding 95% confidence intervals (CIs) were measured to assess the degree of discrimination of the variables. All data were analyzed using Statistical Package for the Social Sciences software for Windows (version 21.0; SPSS Inc., Chicago, IL, USA). In all cases, a two-tailed P-value less than 0.05 was considered statistically significant.

The baseline characteristics of the 73 CHB patients are summarized in Table 1. All patients were genotype C2. All patients had been treated with LAM as a first line oral antiviral agent followed by ETV treatment for LAM resistance. The median duration of LAM therapy was 32 months (range, 15-144 months). The median duration of ETV treatment prior to TDF based rescue therapy was 26.5 months (range, 20-70 months). The known genotypic ETV mutations, 169SMT, 184IALT, 202G, 250LV, and 184LV 202G in the presence of LAM resistance substitutions M204I/V ± L180M were detected in 8, 28, 21, 7, and 9 patients, respectively. Patients were treated with TDF based rescue therapy for more than 6 months. Of the 73 patients, 42 received TDF+ETV combination therapy, 19 received TDF+LAM combination therapy, and 12 had TDF monotherapy. The median duration of TDF based rescue therapy was 37 months (range, 6 to 45 months). Of the 73 patients, 36 (49.5%) had liver cirrhosis, 63 (86.3%) were HBeAg positive. Twenty-three (31.5%) patients had abnormal alanine transaminase (ALT) levels. All patients had persistent HBV viremia prior to TDF based rescue therapy. The median HBV DNA level prior to TDF based treatment was 4.11 log₁₀ IU/mL (range, 1.42-8.23 log₁₀ IU/mL). Among the 23 patients with elevated ALT levels at baseline, 14 (60.9%) patients had ALT normalization during TDF based rescue treatment. During TDF based rescue therapy, VR was achieved in 63 patients (86.3%). The cumulative rates of VR at 12 and 24 months were 63.3% and 87.1%, respectively. During TDF-based rescue therapy, 30 (41.1%) had PVR. Among the 63 HBeAg-positive patients, 6 (9.5%) achieved HBeAg seroconversion during TDF based treatment. However, no patient lost serum HBsAg during treatment. During the follow-up visits, 7 patients (9.6%) were found to be non-adherent to TDF therapy (determined by review of medical and pharmacy records). Among these patients, 2 (2.7%) experienced VBT (increase in HBV DNA by 1 log₁₀ IU/mL at 9-12 months of treatment). These 2 patients with VBT responded well to TDF rescue therapy after continuation of treatment.

To define whether there was any difference in the rate of VR in TDF alone, TDF+LAM, or TDF+ETV therapy, VR rates were compared according to the variables using a log-rank test. The baseline characteristics of the TDF monotherapy (n=12), TDF+LAM (n=19), and TDF+ETV (n=42) treatment groups are shown in Table 2. At baseline, there was no significant difference among the three groups except HBV DNA levels at the start of TDF based rescue therapy (4.10 ± 2.80 in TDF alone, 3.20 ± 1.03 in TDF+LAM, and 4.69 ± 1.89 in TDF+ETV therapy, P=0.021) and the duration of TDF rescue treatment (25.8 ± 14.4 months in TDF alone, 38.2 ± 13.0 months in TDF+LAM, 33.6 ± 9.8 months in TDF+ETV therapy, P=0.003). The mean reduction rates of HBV DNA level during 12 months from baseline are shown in Fig. 1. The mean reduction in serum HBV DNA concentration was not significantly different among the three groups. Overall clinical outcomes among the three treatment groups are summarized in Table 3. The rates of VR among the three groups (TDF monotherapy, TDF+LAM, and TDF+ETV) were not statistically different (Fig. 2; log-rank P=0.200) at 12 months (59.3%, 78.9%, and 51.8 %, respectively) or at 24 months (88.4%, 94.7%, and 84.2 %, respectively). Additionally, the proportion of patients that achieved a normalized ALT level was also similar between the groups.

Of the 73 patients, 36 (the presence of ETV+ADV salvage treatment) received combination therapy of ETV+ADV as salvage therapy for LAM and ETV resistance prior to TDF based rescue therapy. The remaining 37 patients (the absence of ETV+ADV salvage treatment) only received ETV monotherapy prior to TDF based rescue therapy. The presence of ETV+ADV salvage treatment group had significantly lower baseline AST, ALT (P< 0.05, respectively), and HBV DNA (P< 0.001) levels than the absence of ETV+ADV salvage treatment group. VR was observed in 94.4% (34/36) of patients in the presence of ETV+ADV salvage treatment group and in 78.4% (29/37) of the patients in the absence of ETV+ADV salvage treatment group. A higher proportion of patients in the presence of ETV+ADV salvage treatment group achieved a VR at 12 months (73.3% vs. 51.6 %) and 24 months (93.8% vs. 80.4%) than that of patients in the absence of ETV+ADV salvage treatment group (log-rank P=0.019); however, the former subgroup had lower baseline HBV DNA values (mean, 3.21 ± 1.22 vs. mean, 5.08 ± 2.14 log₁₀ IU/mL for the presence and absence of ETV+ADV salvage treatment groups, respectively).

To define whether there were any differences in the rates of VR according to the clinical and virological factors (such as genotypic resistance profile, the presence of ETV+ADV salvage treatment or TDF monotherapy vs. combination therapy with LAM or ETV), VR rates according to these variables were compared using a logistic regression analysis (Table 4). Univariate analysis showed that lower baseline HBV DNA levels and the presence of ETV+ADV salvage treatment prior to TDF based rescue therapy were significantly associated with the VR rates. In multivariate analysis, absolute HBV DNA level at the start of TDF rescue treatment was the only significant predictor for VR (HR, 0.723; 95% CI, 0.627-0.834; P<0.001).

To determine whether an optimal cut-off level of HBV DNA at baseline of TDF treatment could predict VR, we analyzed the HBV DNA cut-off values could provide maximal predictive efficacy. Our results revealed that the best cut-off value of HBV DNA level at baseline for the prediction of VR was 5.61 log₁₀ IU/mL, with a sensitivity and specificity of 80.6% and 70.0%, respectively. Using cut-off value of HBV DNA at baseline, the cumulative VR rates in patients with HBV DNA levels ≥ 5.61 log₁₀ IU/mL (n=19) and those with HBV DNA levels < 5.61 log₁₀ IU/mL (n=54) were 22.7% and 77.9% at 12 months and 73.2% and 98.2% at 24 months, respectively (Fig. 3; log-rank P<0.001).

No clinically significant AEs were observed during the TDF-based treatment. Mean creatinine level and estimated glomerular filtration rate did not change during the treatment period. Two patients (2.7%) had an increase in serum creatinine. However, no patient had an increase in serum creatinine of > 0.5 mg/dL above the baseline value. Mild hypophosphatemia (a serum phosphorus level of < 2.7 mg/dL) was detected in one patient without change in serum creatinine. HCC occurrence was not detected in any patient during the follow-up period.