Longitudinal hematologic decline during peginterferon alfa-2b therapy in chronic hepatitis B: clinical safety and exploratory whole-blood transcriptomics

This is the authors' abstract. We don't add key points for this paper.

Peginterferon-based therapy for chronic hepatitis B is known to cause leukopenia, neutropenia, and thrombocytopenia, but the relationship of early hematologic trajectories to serological response, supportive management, and whole-blood molecular signals remains uncertain. We conducted an exploratory hematologic safety-focused secondary analysis of a prospective, single-center cohort of 21 patients receiving peginterferon alfa-2b-based therapy (13 responders and 8 non-responders). WBC, ANC, and platelet counts were assessed at visits corresponding approximately to baseline, Week 5, and Week 10. Non-parametric comparisons were supplemented with effect sizes and 95% confidence intervals, and supportive-treatment timing was examined in locked sensitivity analyses. All paired Visit-1-to-Visit-3 changes were negative in both response groups, whereas responder-versus-non-responder estimates were imprecise and did not establish a response-group-specific toxicity pattern. Visit-2 WBC <4.0 × 10 9 /L occurred in 19/21 patients (90.5%; exact 95% CI: 69.6%-98.8%) and was treated as a pragmatic clinical-management context threshold rather than a discriminating transcriptomic endpoint. A post-documented oral-support-date indicator did not materially alter the fitted WBC or ANC pattern, but this non-randomized sensitivity analysis cannot estimate treatment efficacy. In the manuscript-consistent DESeq2 variance-stabilizing-transformation workflow, no unadjusted primary Hallmark or candidate-gene family met the prespecified exploratory BH-FDR q < 0.10 threshold. A reviewer-directed xCell2 sensitivity analysis showed longitudinal internal concordance between the prespecified neutrophil enrichment score and measured ANC; composition adjustment changed several pathway estimates, but no paired W0-W10 family reached q < 0.10 and unstable baseline-adjusted platelet signals did not support a specific pathway or mechanism. These small-cohort findings are exploratory, are not evidence of equivalence or mechanism, and require confirmation in larger independent cohorts.