摘要
A model was constructed consisting of clinical and serum variables to discriminate between hepatitis B e antigen (HBeAg) positive chronic hepatitis B (CHB) patients with and without significant fibrosis (stages 2- 4 vs. stages 0- 1). Consecutive treatment-naive CHB patients who underwent liver biopsy were divided into 2 sequential groups: a training group (n = 200) and a validation group (n = 172). Multivariate analysis identified α 2- macroglobulin, age, gamma glutamyl transpeptidase, and hyaluronic acid as independent predictors of fibrosis. The area under the receiver operating characteristic curve was 0.84 for the training group and 0.77 for the validation group. Using a cutoff score of< 3.0, the presence of significant fibrosis (F2 to F4) could be excluded with high accuracy (86.1% negative predictive value [NPV], 70.1% positive predictive value [PPV], and 94.8% sensitivity) in 43 (21.5% ) of 200 patients in the training group, and with the same certainty (90.9% NPV, 64.7% PPV, and 98.0% sensitivity) in 22 (12.8% ) of 172 patients in the validation group. Similarly, applying a cutoff score of >8.7, the presence of significant fibrosis could be correctly identified with high accuracy (91.1% PPV, 51.6% NPV, and 95.2% specificity) in 41 (20.5% ) of 200 patients in the training group, and with the same certainty (84.8% PPV, 52.4% NPV, and 90.4% specificity) in 39 (22.7% ) of 172 patients of the validation group. In conclusion, a predictive model with a combination of easily accessible variables identified HBeAg-positive CHB patients with and without significant fibrosis with a high degree of accuracy. Application of this model may decrease the need for liver biopsy in staging of 35.5% CHB.
A model was constructed consisting of clinical and serum variables to discriminate between hepatitis B e antigen （HBeAg） positive chronic hepatitis B （CHB） patients with and without significant fibrosis （stages 2 -4 vs. stages 0 - 1 ）. Consecutive treatmentnaive CHB patients who underwent liver biopsy were divided into 2 sequential groups： a training group （n = 200） and a validation group （n = 172）. Multivariate analysis identified α 2 -macroglobulin, age, gamma glutamyl transpeptidase, and hyaluronic acid as independent predictors of fibrosis. The area under the receiver operating characteristic curve was 0. 84 for the training group and 0. 77 for the validation group. Using a cutoff score of 〈 3.0, the presence of significant fibrosis （F2 to F4） could be excluded with high accuracy （86. 1% negative predictive value [ NPV ], 70. 1% positive predictive value [PPV], and 94.8% sensitivity） in 43 （21.5%） of 200 patients in the training group, and with the same certainty （90.9% NPV, 64. 7% PPV, and 98.0% sensitivity） in 22 （12. 8% ） of 172 patients in the validation group. Similarly, applying a cutoff score of 〉 8.7, the presence of significant fibrosis could be correctly identified withhigh accuracy （91.1% PPV, 51.6% NPV, and95.2% specificity） in 41 （20. 5% ） of 200 patients in the training group, and with the same certainty （84.8% PPV, 52.4% NPV, and 90.4% specificity） in 39 （22.7%） of 172 patients of the validation group. In conclusion, a predictive model with a combination of easily accessible variables identified HBeAg-positive CHB patients with and without significant fibrosis with a high degree of accuracy.
