Mechanistic modeling to understand variability in responses to chronic Hepatitis B treatment
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Chronic hepatitis B virus (HBV) remains the most common serious liver infection globally, accounting for an estimated 820,000 deaths each year. Patient responses to treatment vary widely, due to complex interplay between viral and immune system dynamics. As yet, there is no reliable way to predict response; this is one reason cure rates remain disappointingly low ( < 10%).
We developed a mechanistic model to simulate serum viral markers evolution during two HBV treatment mainstays - the nucleoside analog entecavir (ETV) and pegylated interferon alfa (IFN) - for a variety of patients, and identify and quantify the key processes driving variability in patient responses. Based on a detailed literature review, this model integrates key processes in chronic HBV pathophysiology and drug pharmacokinetics/dynamics and was calibrated on published data only.
Post simulation regression and classification analyses, including a global sensitivity analysis and a random forest, highlighted the importance of HBV replication cycle processes in explaining pre-treatment inter-patient variability in serum viral markers. Post-treatment with entecavir, most of the response variability could be attributed to interactions between the viral replication cycle and immune system processes. Response variability after IFN treatment, however, was more directly related to the drug mechanism of action, which includes direct antiviral effects and immune system modulation. Quantifying these measures may help to inform new drug development with identification of more direct tailored and effective HBV therapy.
Graphical Abstract
Highlights
-
A mechanistic model of chronic hepatitis B disease, accounting for intra-hepatocyte virus replication and an implemented immune response to entecavir and Peginterferon alfa-2a, allows for reproducing the observed variability between patients in terms of measured serum viral markers in response to treatments.
-
Statistical analysis of simulated virtual populations helps investigate the mechanisms involved in observed variability between patients for both baseline and responses to treatments.
-
Such a mechanistic model offers, via a QSP platform, new perspectives on the exploration of hepatitis B physiopathology, including treatment combinations or hepatitis D co-infection.
