O117 : Predictors of clinical response: Results from a large, randomized controlled study with tenofovir disoproxil fumarate (TDF) plus peginterferon alfa-2A (PEG) combination for chronic hepatitis B (CHB)

Abstract

Background: There are scarce on-treatment data on HBsAg and HBV DNA kinetics with concomitant nucleos(t)ide analog and immunomodulator HBV therapy. Predictors of clinical response may help manage patients on combination therapy. Methods: From study GS-US-174-0149, 740 CHB patients without advanced disease were randomized 1:1:1:1 to receive (TDF+PEG) x48 weeks (arm A); (TDF+PEG) x16 weeks followed by TDFx32 weeks (arm B); continuousTDF(arm C); PEGx48 weeks (arm D).Associations between baseline and on-treatment variables with change in HBV DNA (log10 IU/ ml) or HBsAg (log10 IU/ml) levels from baseline to week 48 or HBsAg loss atWeek 72 were examined via linear regression or Cox regression analyses, respectively, in univariate models as well as multivariate models. Results: In a multivariate analysis, treatment arms A, B, and C compared to arm D, baseline HBeAg-negative status, higher baseline HBV DNA, and lower baseline HBsAg levels were associated with greater HBV DNA decline at Week 48. For every 1 log10 IU/ml increase in baseline HBV DNA or 1 log10 IU/ml decrease in baseline HBsAg, there is an expected greater HBV DNA decline by 0.85 or 0.31 log10 IU/mL, respectively, at Week 48. On multivariate analysis, achieving HBsAg loss was associated with treatment Arm A, GT A, early ALT flare on treatment, and, at Week 12, with HBsAg decline from baseline >1 log10 IU/ml (table). The positive predictive values for achieving HBsAg loss at week 72 with (TDF+PEG)x 48 weeks with either 1 log10 IU/ml decline or HBsAg < 100 IU/ml at week 12 were 43% and 50%, respectively, and the negative predictive values (NPV) were 97% and 95%, respectively. TDF+ PEG combination for 48 weeks exhibited the largest HBsAg decline (log10 IU/ml) as compared to other treatment arms (Arm A: -1.1, B: -0.5, C: -0.3, D: -0.8; p < 0.01 for all comparisons of arm A versus others). Similar levels of HBsAg decline were observed in genotypes (GT)Aand B (mean log10 IU/ ml +/- SD: -1.2 +/- 2.0 and -1.1 +/- 1.2, respectively), both of which exhibited a greater decline than GT C and D (-0.5 +/- 0.9 and -0.4 +/- 1.0, respectively) (p < 0.05). Conclusions: Higher baseline HBsAg levels appear to impact negatively on-treatment HBV DNA response. HBsAg decline on TDF plus PEG combination therapy x48 weeks was synergistically greater than on either TDF or PEG monotherapy. HBsAg decline appears to favor HBV GT A and B. Moreover, HBsAg decline at week 12 shows high NPV for week 72 HBsAg loss and may provide a valuable tool for response-guided therapy in PEG+TDF combination treatment.