肝脏 ›› 2016, Vol. 21 ›› Issue (10): 810-818.

• 论著 • 上一篇    下一篇

血清HBsAg、HBcrAg和HBV DNA预测慢性乙型肝炎肝组织病理状态的性能评价

张占卿, 陆伟, 丁荣蓉, 翁齐铖, 张智勇, 王雁冰, 周新兰, 黄丹, 李秀芬   

  1. 201508 上海市公共卫生临床中心肝炎二科(张占卿,陆伟,丁荣蓉,王雁冰,周新兰,黄丹,李秀芬);日本富士瑞必欧株式会社(翁齐铖,张智勇)
  • 收稿日期:2016-05-05 发布日期:2020-06-18
  • 通讯作者: 张占卿,Email:doctorzzqsphc@163.com
  • 基金资助:
    上海市卫生和计划生育委员会重点科研项目(20134032);国家“十二五”科技重大专项(2013ZX10002005)

Evaluation on predictive value of serum HBsAg, HBcrAg and HBV DNA for liver pathology in chronic hepatitis B patients

ZHANG Zhan-qing, LU Wei, DING Rong-rong, WENG Qi-cheng, ZHANG Zhi-yong, WANG Yan-bing, ZHOU Xin-lan, HUANG Dan, LI Xiu-fen   

  1. Division Two of Hepatitis Department, Shanghai Public Health Clinical Center, Shanghai 201508, China
  • Received:2016-05-05 Published:2020-06-18
  • Contact: ZHANG Zhan-Qing, Email: doctorzzqsphc@163.com

摘要: 目的 评价血清HBsAg、HBcrAg、HBV DNA预测慢性乙型肝炎肝组织病理状态的性能。方法 将324例HBeAg阳性和255例HBeAg阴性慢性乙型肝炎患者随机分为3组匹配的训练集和验证集。血清HBsAg和HBcrAg分别采用Abbott Architect I2000和Fujirebio Lumipulse G1200全自动化学发光免疫系统检测,血清HBV DNA采用Bio-Rad Icycler荧光定量PCR仪检测。肝组织病理学诊断采用Scheuer评分系统,其中病理学分级包括G0~G4五级,分期包括S0~S4五期。结果 无论HBeAg阳性或阴性患者,3个训练集与3个匹配的验证集性别比例和平均年龄、病理学分级和分期构成比之间的差异均无统计学意义(P>0.05)。HBeAg阳性患者,血清HBsAg、HBcrAg、HBV DNA预测全集病理学分级≥G3和分期≥S4的ROC曲线下面积最大;参照预测3个训练集病理学分级≥G3和分期≥S4的最佳截断值,血清HBsAg、HBcrAg、HBV DNA预测3个匹配的验证集病理学分级≥G3和分期≥S4的灵敏度极差分别为37%和9%、30%和16%、17%和14%,特异度极差分别为12%和5%、13%和3%、15%和6%。HBeAg阴性患者,血清HBcrAg、HBV DNA预测全集病理学分级≥G2和分期≥S2的ROC曲线下面积最大;参照预测3个训练集病理学分级≥G2和分期≥S2的最佳截断值,血清HBcrAg、HBV DNA预测3个匹配的验证集病理学分级≥G2和分期≥S2的灵敏度极差分别为11%和20%、46%和19%,特异度极差分别为15%和2%、38%和16%。结论 HBeAg阳性患者,血清HBsAg、HBcrAg、HBV DNA可预测的最佳病理状态为病理学分级≥G3和分期≥S4,其预测理学分期≥S4的稳定性高于预测病理学分级≥G3;HBeAg阴性患者,血清HBcrAg和HBV DNA可预测的最佳病理状态为病理学分级≥G2和分期≥S2,血清HBcrAg预测病理学分级≥G2和分期≥S2的稳定性高于HBV DNA。

关键词: 乙型肝炎表面抗原, 乙型肝炎核心相关抗原, 乙型肝炎病毒DNA, 肝组织, 病理学, 无创诊断

Abstract: Objective To evaluate the predictive value of serum hepatitis B surface antigen (HBsAg), hepatitis B core-related antigen (HBcrAg) and hepatitis B virus (HBV) DNA for liver pathology in chronic hepatitis B (CHB) patients. Methods CHB patients, containing 324 HBeAg-positive and 255 HBeAg-negative cases, were randomly divided into 3 paired train and validation sets. Serum HBsAg and HBcrAg were measured by Lumipulse G1200 and Abbott Architect I2000 automatic chemiluminescence immunoassay analyzers, respectively. Serum HBV DNA was detected by Bio-Rad Icycler fluorescence quantitative PCR system. Scheuer scoring system was applied for pathological evaluation of liver tissues, containing 5 grades from G0 to G4 and 5 stages from S0 to S4. Results Gender ratio, average age and proportion of pathological grades and stages in 3 paired train and validation sets showed no statistically significant differences (P>0.05) in both HBeAg positive and negative cases. In HBeAg-positive patients, the areas under receiver operating characteristic curve (ROC) of serum HBsAg, HBcrAg and HBV DNA for predicting ≥ G3 and ≥ S4 in complete set were the largest. Referring to the optimal cutoffs for predicting ≥ G3 and ≥ S4 in 3 train sets, the ranges of sensitivities of serum HBsAg, HBcrAg and HBV DNA for predicting ≥ G3 and ≥ S4 in the 3 matched validation sets were 37%, 30%, 17% and 9%, 16%, 14%, respectively. Meanwhile, the ranges of specificity were 12%, 13%, 15% and 5%, 3%, 6%, respectively. In HBeAg-negative patients, the areas under ROC of serum HBcrAg and HBV DNA for predicting ≥ G2 and ≥ S2 in complete set were the largest. Based on the optimal cutoffs in 3 train sets, the ranges of sensitivity of serum HBcrAg and HBV DNA for predicting ≥ G2 and ≥ S2 in the 3 matched validation sets were 11%, 46% and 19%, 20%, respectively, and the ranges of specificities were 15%, 38% and 2%, 16%, respectively. Conclusion In HBeAg-positive patients, serum HBsAg, HBcrAg and HBV DNA were suitable for predicting pathologic states of ≥ G3 and ≥ S4, and the stabilities for predicting ≥ S4 were better than those for predicting ≥ G3. In HBeAg-negative patients with pathologic states of ≥ G2 and ≥ S2, serum HBcrAg and HBV DNA were the best predictable indicators, and serum HBcrAg showed better stabilities than serum HBV DNA.

Key words: Hepatitis B surface antigen; Hepatitis B core-related antigen; Hepatitis B virus DNA; Liver tissue; Pathology; Noninvasive diagnosis