超声实时组织弹性成像无创预测慢性乙型肝炎肝纤维化的前瞻性研究

Real-time tissue elastography-based noninvasive prediction model for liver fibrosis in patients with chronic hepatitis B: a prospective study

目的进一步提高超声实时组织弥散定量分析技术中的肝纤维化指数(LF指数)评估肝纤维化的临床应用价值。方法选取2015年1月到12月在温州医科大学附属第一医院就诊的116例明确诊断为慢性乙型肝炎的患者,均行实时组织弹性成像(RTE)检查及肝组织活检。所有患者根据是否为重度肝纤维化,以及是否为肝硬化分别分组,采用Logistic回归单因素及多因素分析筛选出重度肝纤维化与肝硬化的独立预测因子,并由此构建预测模型。结果 (1)多因素分析确认脾门指数(OR=13.956,P=0.002)、LF指数(OR=6.283,P=0.023)为重度肝纤维化的独立危险因素;γ-谷氨酸转肽酶(OR=1.012,P=0.049)、脾门指数(OR=5.676,P=0.002)、LF指数(OR=14.102,P=0.001)为肝硬化的独立危险因素。由此新建了重度肝纤维化与肝硬化的预测模型:LFI-SPI评分(LSP评分)与LFI-SPI-GGT评分(LSPG评分)。(2)LSP评分的受试者工作特征曲线下面积(AUROC)为0.87,相对于现有的纤维化预测模型:LF指数(AUROC=0.76,P=0.0109),天冬氨酸转氨酶/血小板比值(APRI)评分(AUROC=0.64,P=0.0031),FIB-4评分(AUROC=0.67,P=0.0044)及Fibro Scan(AUROC=0.68,P=0.0021),LSP评分诊断准确度最高。(3)LSPG评分预测肝硬化的准确度较高,ROC曲线下面积0.93,除Fibro Scan(AUROC=0.85,P=0.134)外,LF指数(AUROC=0.81,P=0.0113)、APRI评分(AUROC=0.67,P<0.0001)、FIB-4评分(AUROC=0.72,P=0.0005)的诊断准确性均低于LSPG评分。结论预测重度肝纤维化与肝硬化的新模型,LSP评分主要用于筛查重度肝纤维化,而LSPG评分主要用于排除肝硬化。

Objective To further improve the value of clinical application of a quantitative analysis method called LF-index（LFI） which based on real-time tissue elastography（RTE）. Methods We prospectively enrolled 116 consecutive patients with chronic hepatitis B（CHB） and all patients underwent a liver biopsy and RTE between January 2015 and December 2015 at the First affiliated hospital of Wenzhou Medical University. Univariate and multivariate analyses were performed, and the prediction models for predicting significant fibrosis and cirrhosis were derived from independent predictors. Results（1） In multivariate analyses, spleno-portal index（SPI）（OR=13.956, P=0.002） and LFI（OR=6.283, P=0.023） were confirmed as independent predictors of significant fibrosis. In multivariate analyses of patients with and without cirrhosis, we found significant differences in the γ-Glutamyl transferase（GGT）（OR=1.012, P=0.049）, SPI（OR=5.676, P=0.002） and LFI（OR=14.102, P=0.001）.（2） A novel model called LFI-SPI score（LSPS） for prediction of significant fibrosis was developed（area under receiver operating characteristic curve [AUROC]=0.87）, showing the superiority of diagnostic accuracy than LFI（AUROC=0.76, P=0.0109）, aspartate aminotransferase to platelet ratio index（APRI）（AUROC=0.64, P=0.0031）, fibrosis-4 index（FIB-4）（AUROC=0.67, P= 0.0044） and Fibroscan（AUROC=0.68, P=0.0021）.（3） We also developed a LFI-SPI-GGT score（LSPGS） for predicting cirrhosis, with an AUROC of 0.93. The diagnostic accuracy of LSPGS was similar to that of Fibroscan（AUROC=0.85, P=0.134）, and was superior to LFI（AUROC=0.81, P=0.0113）, APRI（AUROC=0.67, P〈0.0001）, and FIB-4（AUROC=0.72, P=0.0005）. Conclusions We developed new formulas, LSPS and LSPGS for predicting significant fibrosis and cirrhosis in this prospective study. LSP score was mainly used for screening of significant liver fibrosis, and LSPG score was mainly used to exclude cirrhosis.