摘要
目的在慢性乙型肝炎病毒(HBV)感染患者中建立基于常规实验室指标的肝纤维化非创伤性诊断模型,为传统的肝穿刺活检提供简便的非创伤性替代手段。方法采用Logistic回归等方法分析386例慢性HBV感染患者的常规实验室指标与肝纤维化分期的关系,建立诊断模型。用受试者工作曲线(ROC)等方法验证和比较该模型与Forns指数、APRI指数、Hepascore及SLFG模型的诊断价值。结果各指标组合模型对肝纤维化分期的诊断价值优于单项常规实验室指标,其中SLFG模型、S指数和Hepascore均具有较好的表现。由γ-谷氨酰转肽酶(GGT)、血小板(PLT)和白蛋白(Alb)三个常规指标组成的S指数(S指数=1000×GGT/(PLT×Alb2))判断有无明显肝纤维化和有无早期肝硬化时的受试者工作曲线下面积(AUC)分别达到0.686和0.762。使用以下推荐界值,S指数<0.1预测无明显肝纤维化的灵敏度为90.4%,S指数≥0.5预测存在明显肝纤维化的特异度为86.2%;S指数<0.3预测无早期肝硬化的灵敏度为84.8%,S指数≥1.5预测存在早期肝硬化的特异度为97.7%。结论由常规实验室指标建立的简单组合S指数,能较准确而方便地区分存在明显肝纤维化或早期肝硬化的慢性HBV感染患者。
Objective To establish a noninvasive diagnostic model based on conventional laboratory markers for predicting liver fibrosis in patients with chronic hepatitis B virus（HBV） infection, and to provide a convenient noninvasive substitute for liver biopsies. Methods In 386 patients with chronic HBV infection, correlation between the conventional laboratory markers and fibrosis stages was assessed using statistical analysis such as Logistic regression and a diagnostic model was established. The diagnostic value of the novel model was assessed together with the Forns index, APRI index, Hepascore and SLFG model using the receiver operating characteristic curves （ROC）. Results The diagnostic value of each marker panel was better than simple conventional laboratory markers, especially the SLFG model, S index and Hepascore. The S index was consisted of GGT, PLT and albumin [S index = 1 000 × GGT / （PLT × Alb^2 ）1. The areas under the ROC curves （AUC） of S index were 0. 686 and 0. 762 for predicting significant fibrosis and cirrhosis, respectively. Using optimized cut off values, the sensitivity of predicting the absence of significant fibrosis （S index〈0.1 ） was 90. 4%, the specificity of predicting the presence （S index≥0.5） of significant fibrosis was 86.2%. The sensitivity of predicting the absence of cirrhosis （S index〈0.3） was 84.8%, the specificity of predicting the presence （S index≥1. 5） of cirrhosis was 97.7%. Conclusion The S index, a simple markers panel consisting of conventional laboratory markers, could easily predict significant fibrosis and cirrhosis with a high degree of accuracy.
出处
《肝脏》
2008年第5期362-367,共6页
Chinese Hepatology
基金
国家高技术研究发展(863)计划重大项目(2006AA02A411)
上海市科学技术委员会登山计划(064119519)
上海市医学发展基金重点研究项目(99ZDI001)
上海市重点学科建设项目(Y0205)
