Exploratory associations of gut microbiota composition with HBeAg and HBsAg status in nucleos(t)ide analog-treated...

HBeAg and HBsAg reflect viral replication and host immune status and are important determinants of treatment responsiveness in chronic hepatitis B (CHB). The gut microbiota, a key component of the gut-liver axis, may influence immune regulation and antiviral outcomes. This study characterized gut microbiota profiles across different HBeAg and HBsAg statuses in CHB patients receiving nucleos(t)ide analog (NA) therapy. Fecal shotgun metagenomic sequencing was performed in 29 healthy controls (HC) and 101 NA-treated CHB patients without cirrhosis or hepatocellular carcinoma, including 30 HBeAg-positive (EP) and 71 HBeAg-negative (EN) individuals. The EN group was further stratified by serum HBsAg levels into an advantageous subgroup (<1,500 IU/mL, n = 57) and a non-advantageous subgroup (≥1,500 IU/mL, n = 14). Alpha and beta diversity did not differ significantly among EP, EN, and HC groups. Species-level analyses identified nominal group-related differences in several Bacteroidales- and Firmicutes-associated anaerobic commensals, including Bacteroides uniformis, Blautia obeum, Eubacterium ventriosum, Anaerobutyricum, and Parabacteroides timonensis; however, none remained significant after global FDR correction across all detected species. Within the EN group, selected taxa showed directionally consistent associations with HBsAg strata in downsampling, leave-one-out, and CLR-based sensitivity analyses, but these subgroup findings also did not remain significant after FDR correction. NA-treated CHB was associated with exploratory taxonomic differences without detectable loss of microbial diversity. The observed associations with HBeAg status and HBsAg burden were hypothesis-generating and require validation in larger longitudinal cohorts incorporating absolute microbial quantification, targeted metabolomics, host immune profiling, and more complete dietary, anthropometric, and other relevant clinical covariates.