ObjectiveTo establish an early predictive model using serological markers based on LASSO regression for predicting the possibility of HBsAg clearance in HBeAg-negative chronic hepatitis B （CHB） patients treated with pegylated interferon α-2b （PEG-IFNα-2b）， and to investigate the diagnostic value of the model. MethodsA total of 136 HBeAg-negative CHB patients who received PEG-IFNα-2b treatment in the Affiliated Hospital of Xuzhou Medical University from April 2020 to October 2021 were enrolled， among whom 47 received PEG-IFNα-2b for the first time （previously untreated） and 89 received PEG-IFNα-2b after 48 weeks of treatment with nucleos（t）ide analogues （treatment-experienced）. The patients were randomly assigned to a training set with 95 patients and a validation set with 41 patients at a ratio of 7∶3， and related data were collected for both groups， including virological markers， routine blood test results， and liver function at baseline and week 12 of treatment. According to HBsAg status at week 48 of treatment， the patients were divided into seroconversion group with 38 patients and non-seroconversion group with 98 patients. The independent-samples t test was used for comparison of normally distributed continuous data between two groups， and the Wilcoxon rank-sum test was used for comparison of non-normally distributed continuous data between two groups； the chi-square test was used for comparison of categorical variables between two groups. The LASSO regression analysis and univariate and multivariate logistic regression analyses were used to establish a nomogram model； the receiver operating characteristic （ROC） curve was used to assess its predictive ability， and the area under the ROC curve （AUC） was used for comparison of predictive value. ResultsIn the training set， 95 HBeAg-negative CHB patients were treated with PEG-IFNα-2b for 48 weeks， among whom there were 27 patients in the seroconversion group and 68 in the non-seroconversion group. The univariate Logistic regression analysis， with P<0.2 as the criterion for screening， showed that 9 indicators were included in the LASSO regression analysis， i.e.， sex， baseline HBV DNA level， the reduction in HBV DNA in 0 — 12 weeks， baseline HBsAg level， the reduction in HBsAg in 0 — 12 weeks， baseline aspartate aminotransferase （AST） level， the reduction in AST in 0 — 12 weeks， baseline alanine aminotransferase （ALT） level， and the reduction in ALT in 0 — 12 weeks. The LASSO regression analysis showed that sex， baseline HBsAg level， the reduction in HBsAg in 0 — 12 weeks， and the reduction in ALT in 0 — 12 weeks were non-zero variables and were included in the multivariate Logistic regression analysis. The multivariate Logistic regression analysis obtained 4 independent predictive factors， i.e.， sex （odds ratio ［OR］=5.38， 95% confidence interval ［CI］： 1.11 — 34.21， P=0.049）， baseline HBsAg level （OR=0.12， 95%CI： 0.04 — 0.26， P<0.001）， the reduction in HBsAg in 0 — 12 weeks （OR=5.54， 95%CI： 1.97 — 19.18， P=0.003）， and the reduction in ALT in 0 — 12 weeks （OR=0.99， 95%CI： 0.97 — 1.00， P=0.039）. A nomogram model was established based on the results of the multivariate Logistic regression analysis， and the ROC curve was used to assess the predictive value of this nomogram model. This nomogram model had an AUC of 0.934 （95%CI： 0.886 — 0.981） in the training set and an AUC of 0.921 （95%CI： 0.838 — 1.000） in the validation set. In addition， the results of calibration curve and decision curve analyses showed that the model had good consistency and accuracy. ConclusionBased on general information and serological markers， the LASSO regression analysis is used to establish a nomogram model using sex， baseline HBsAg level， the reduction in HBsAg in 0 — 12 weeks， and the reduction in ALT in 0 — 12 weeks， and this model can be used to predict the probability of achieving HBsAg clearance in HBeAg-negative CHB patients treated with PEG-IFNα-2b， which provides important reference and theoretical support for the clinical treatment of patients.