CHB has emerged as a significant global public health concern. Peg-IFN α enhances innate immunity, triggers T-cell-mediated immune responses, suppressing HBV protein synthesis, and reducing the levels of covalently closed circular DNA (cccDNA). This leads to a greater clearance of HBsAg than that achieved with NAs [11]. The clearance of HBsAg in serum is considered indicative of clinical recovery from hepatitis B infection, thus serving as the preferred endpoint for treatment [12]. In this study, the clearance of HBsAg at 48 weeks was used as the endpoint measure. By performing multivariate analysis, we systematically evaluated the factors affecting HBsAg clearance and observed three predictive variables: gender, baseline HBsAg levels, and liver cirrhosis.

This study found that, compared to male CHB patients, female patients have a higher chance of HBsAg clearance after 48 weeks of interferon therapy (OR = 1.879, 95% CI 1.2–2.943, p = 0.006). This finding is consistent with previous multiple research results. A large cohort study involving 518 HBeAg-negative CHB patients demonstrated that female patients have a 1.93 times higher chance of achieving virologic response at 24 weeks after interferon therapy compared to male patients [13]. Research conducted by Chann et al. on HBeAg-positive CHB patients also demonstrates that females are a favorable population for interferon therapy, exhibiting a higher interferon response rate [14]. However, the precise mechanisms behind the higher interferon response in female patients still require further investigation.

It has been confirmed that baseline HBsAg levels are a highly influential predictor of HBsAg clearance [15]. A retrospective cohort study demonstrated that lower baseline HBsAg levels are associated with a higher probability of HBsAg clearance after 48 weeks of interferon therapy. Furthermore, baseline HBsAg levels < 100 lU/mL are considered an independent predictive indicator of clinical cure [16]. In long-term NAs therapy, the combination treatment strategy of Peg-IFNα−2b in HBeAg-negative CHB with HBsAg ≤ 1500 IU/mL yields a higher HBsAg clearance rate compared to monotherapy with NAs. Furthermore, lower baseline HBsAg levels, lower HBsAg levels at 12 and 24 weeks of follow-up, and a rapid decline of HBsAg in the early stage of treatment (weeks 12 and 24) are independent predictive factors for HBsAg clearance in the Peg-IFNα−2a combination therapy [17]. Our research findings are consistent with the above statement, showing that the lower the baseline HBsAg level, the higher the probability of HBsAg clearance at 48 weeks (OR = 0.371, 95% CI 0.307–0.448, p < 0.001).

This study found that the presence of cirrhosis also influences the probability of HBsAg clearance in interferon-treated patients at 48 weeks. The probability of HBsAg clearance at 48 weeks is significantly lower in patients with baseline cirrhosis (OR = 0.438, 95% CI 0.221–0.868, p = 0.018). Prior studies have also indicated that patients with mild liver fibrosis are generally more tolerant to treatment and exhibit a higher response to Peg-IFN therapy compared to those with advanced liver fibrosis or cirrhosis [18]. Due to the numerous adverse reactions associated with interferon itself, patients with cirrhosis or decompensated cirrhosis are usually excluded from study cohorts. Therefore, further research is required to confirm the impact of cirrhosis on interferon response rates.

A number of studies have demonstrated that age is a significant factor influencing the clearance of HBsAg in patients with hepatitis B treated with interferon. These studies have indicated that the likelihood of HBsAg clearance decreases with age [13,14,15, 19]. Furthermore, patients with low baseline HBV DNA levels [13, 18], HBeAg negativity [20], and ALT elevations exceeding five times the upper limit of normal (5 × ULN) [14] demonstrate enhanced responsiveness to interferon therapy. However, multivariate analyses in this study revealed that age, baseline HBV DNA level, baseline HBeAg and ALT levels were not significantly associated with HBsAg clearance after 48 weeks of interferon therapy. This may be attributed to the characteristics of the study population, differences in follow-up time, sample size limitations, and differences in study endpoints.

Following this, we developed a rigorous probability prediction model for seroconversion through multivariate logistic regression analysis. We identified gender, baseline HBsAg levels, and cirrhosis as key predictors of HBsAg clearance. The model exhibited strong discriminative ability and good fit within both the training and validation datasets. These findings align with those reported by Jiang et al [21], who also highlighted the importance of these factors in predicting HBsAg seroclearance. Moreover, our model’s predictive performance was validated using ROC curve analysis, demonstrating high discriminative ability with an AUC of 0.802 in the training set and 0.816 in the validation set. This strong performance indicates that our model can reliably predict HBsAg clearance in CHB patients undergoing Peg-IFNα−2b therapy. Zhang et al [22] have successfully developed a predictive model based on baseline HBsAg levels to forecast the potential for functional cure in CHB patients treated with PEG-IFNα, further underscoring the significance of baseline HBsAg levels. However, our study diverges from Zhang et al by incorporating additional factors—gender and the presence of cirrhosis—into the model, alongside baseline HBsAg levels. In comparison to previous studies [13, 14, 21, 23, 24], our study has the following advantages: Firstly, the prediction model utilises nomograms, a non-invasive risk prediction model that is crucial for screening and clinical practice [25], and has been extensively employed for risk assessment of numerous diseases [26, 27]. Secondly, our study had a broader coverage, including all patients with CHB aged 18–80 years who received 48 weeks of Peg-IFNα−2b therapy. In conclusion, the sample size of our study is more appropriate and the training and validation sets are more adequate. However, This study did not include certain important influencing factors, such as HBV genotypes. Previous research has demonstrated the association between interferon response and HBV genotypes [13, 28, 29]. But, the absence of these tests improves their applicability to clinical practice, as these tests are often not conducted in clinical settings due to limitations in economic resources and testing capabilities. Moreover, this study was exclusively conducted in China, and the geographical limitations may affect the generalizability of the results to other populations with different genetic backgrounds, healthcare systems, and environmental factors. Nonetheless, being a multicenter study, it still demonstrates strong representativeness among Asian populations.