The magnitude of early hepatitis B RNA and DNA declines directly inform capsid assembly modulator effectiveness

Capsid assembly modulators (CAMs) are a promising class of antiviral treatments for hepatitis B virus (HBV) infection. Several CAMs have been evaluated in clinical trials but there is no simple method to estimate their in vivo antiviral effectiveness. We performed viral dynamics modeling of the intracellular and extracellular dynamics of HBV RNA, HBV DNA, and ALT during phase I trials of two CAMs, vebicorvir and ABI-H2158, which inhibit the encapsidation of pgRNA. Fitting our model to the data, we quantify the drug-induced percent inhibition of encapsidated pgRNA production, which we term their in vivo antiviral effectiveness. In both trials, the HBV RNA and HBV DNA declined in a biphasic manner during therapy. The model described these decays well and, by fitting the model to the data, we estimated the CAM effectiveness in each trial participant. Mathematical analysis of the model showed that the magnitude of the first phase of decline of HBV RNA and HBV DNA is explicitly related to CAM effectiveness. However, in the clinic, the end of the first phase may not be known due to sparse sampling. Using clinical trial simulations, we show that the HBV RNA and HBV DNA declines between baseline and day 14 of CAM monotherapy can be used to predict CAM effectiveness. We show that HBV RNA is a clinically relevant biomarker and that very short-term phase I clinical trials can be used to evaluate the in vivo effectiveness of new CAMs, thus reducing the danger of drug resistance developing in trial participants.