RESUMEN
BACKGROUND & AIMS: During treatment of chronic HBV infections, loss or seroconversion of the HBV surface antigen (HBsAg) is considered a functional cure. HBsAg consists of the large (LHBs), middle (MHBs), and small surface protein (SHBs) and their relative proportions correlate strongly with disease stage. Our aim was to assess the association between HBsAg composition and functional cure during treatment. METHODS: A total of 83 patients were retrospectively analyzed. HBsAg loss was achieved by 17/64 patients during nucleos(t)ide analogue (NA) treatment and 3/19 patients following treatment with pegylated interferon-alfa2a (PEG-IFN) for 48 weeks. Sixty-three patients without HBsAg loss were matched as controls. LHBs, MHBs and SHBs were quantified in sera collected before and during treatment. RESULTS: Before treatment, median MHBs levels were significantly lower in patients with subsequent HBsAg loss than in those without (p = 0.005). During treatment, MHBs and LHBs proportions showed a fast decline in patients with HBsAg loss, but not in patients with HBV e antigen seroconversion only or patients without serologic response. MHBs became undetectable by month 6 of NA treatment in all patients with HBsAg loss, which occurred on average 12.8 ± 8.7 (0-52) months before loss of total HBsAg. Receiver-operating characteristic analyses revealed that the proportion of MHBs was the best early predictor of HBsAg loss before NA treatment (AUC = 0.726, p = 0.019). In patients achieving HBsAg loss with PEG-IFN, the proportions of MHBs and LHBs showed similar kinetics. CONCLUSION: Quantification of HBsAg proteins shows promise as a novel tool to predict early treatment response. These assessments may help optimize individual antiviral treatment, increasing the rates of functional cure in chronically HBV-infected patients. LAY SUMMARY: The hepatitis B surface antigen (HBsAg) is a key serum marker for viral replication. Loss of HBsAg is considered stable remission, which can be achieved with antiviral treatments. We have investigated whether the ratios of the different components of HBsAg, namely the large (LHBs) and medium (MHBs) HBsAg during different treatments are associated with the occurrence of HBsAg loss. We found that LHBs and MHBs decrease earlier than total HBsAg before HBsAg loss and we propose LHBs and MHBs as promising novel biomarker candidates for predicting cure of HBV infection.