Lonafarnib (LNF) is an investigational agent that specifically targets hepatitis D virus (HDV) but does not act directly on hepatitis B virus (HBV). This selective antiviral activity creates an opportunity to study HDV kinetic responses and to model how reductions in HDV burden affect HBV markers in HBV/HDV coinfected patients. The LOWR HDV-1 study applied detailed kinetic analysis and mathematical modeling to serial measurements of serum HDV RNA, HBV DNA, and HBsAg during LNF-based therapy.
The analysis included 15 patients with documented HBV/HDV coinfection who received LNF-based treatment regimens. Regimens in the study included LNF monotherapy and LNF combined with ritonavir or pegylated interferon-α (PEG-IFN α). The abstract does not provide a full breakdown of patient numbers per regimen or other enrollment details; those specifics were not reported in the source provided.
Across patients there was an initial delay of 0–2 days after treatment start, followed by a rapid first-phase decline in serum HDV RNA. After the initial decline, three distinct patterns were observed in individual patients: a viral plateau, a slower second-phase decline, or a viral breakthrough (VB).
LNF monotherapy typically produced a flat partial response and was often followed by VB. In contrast, combination therapy with ritonavir or PEG-IFN α was associated with a biphasic HDV decline and did not show VB in the reported observations.
Changes in serum HBV DNA occurred while patients received LNF-based therapy despite LNF not directly targeting HBV. All treatment groups except the LNF + PEG-IFN α cohort included at least one patient who experienced an increase in serum HBV DNA during treatment. The model and data analysis sought to explain these on-treatment HBV DNA increases in relation to HDV suppression.
Serum HBsAg levels remained stable during treatment. In the modeling framework this stability was explained by a constant number of cells producing HBsAg, implying that short-term changes in HDV and HBV replication did not measurably change the HBsAg-producing cell pool over the study period.
The authors developed a mathematical model to reproduce the observed kinetics of HDV RNA, HBV DNA, and HBsAg in treated patients. Key model components described in the abstract include:
The abstract does not report full mathematical equations, parameter-fitting methodology, or model diagnostics; those details were not provided in the source text available here.
Using the model, the investigators estimated the serum half-life of HDV RNA at 1.26 days (95% confidence interval, CI 1.05–1.47). Across all treatment regimens the estimated efficacy in inhibiting HDV RNA production during the initial phase was 94% (95% CI 89–97%), which corresponds to the observed first-phase decline.
To account for the observed second-phase HDV decline in some patients, the model included a time-dependent increase in efficacy that reached a maximum estimated value of 98.9%. These parameter estimates underpin the interpretation of the biphasic declines seen with combination therapies and the partial responses or breakthroughs seen with monotherapy.
The modeling framework explained on-treatment increases in serum HBV DNA by an increase in the HBV DNA production rate when HDV levels fell below an inhibitory threshold. The median modeled increase in HBV DNA production rate was four-fold, with an interquartile range (IQR) of 1–28, corresponding to the observed HBV DNA rises in some patients. This mechanism links HDV suppression to a release of inhibition on HBV replication within the modeling assumptions.
The combined experimental and modeling results support a rapid HDV decline with LNF-based therapy and quantify the dynamics of that decline. The estimated short serum half-life for HDV RNA and high estimated treatment efficacy explain the rapid first-phase decreases; a further time-dependent increase in efficacy can account for biphasic declines observed with combination regimens.
The model provides a plausible mechanism for transient on-treatment increases in HBV DNA—namely, loss of HDV-mediated inhibition of HBV production—while explaining stable HBsAg by an unchanged number of HBsAg-producing cells during the observation window.
Specific methodological details such as full cohort breakdown by regimen, raw kinetic plots, parameter-fitting algorithms, sensitivity analyses, and other model equations or diagnostics are not reported in the abstract and therefore are not included here.
In the LOWR HDV-1 cohort, LNF-based therapy produced rapid reductions in serum HDV RNA with variable later-phase behavior depending on regimen. A mathematical model reproduced these patterns and yielded quantitative estimates: an HDV RNA half-life of 1.26 days, initial treatment efficacy of 94%, and a maximal time-dependent efficacy up to 98.9%. The model attributes observed on-treatment HBV DNA increases to a modeled increase in HBV production rate when HDV falls below an inhibitory threshold and explains stable HBsAg by a constant number of producing cells. Further methodological and cohort details are not available in the source abstract.