Jamali Raika, Arj Abbas, Razavizade Mohsen, Aarabi Mohammad Hossein
From the Research Development Center, Sina Hospital/Digestive Disease Research Institute, Non-Alcoholic Fatty Liver Disease Center, Tehran University of Medical Sciences, Tehran (RJ); Department of Internal Medicine, Shahid Beheshti Hospital (AA, MR); and Research Center for Biochemistry and Nutrition in Metabolic Disease, Faculty of Medicine, Kashan University of Medical Sciences, Kashan, Iran (MHA).
Medicine (Baltimore). 2016 Feb;95(5):e2630. doi: 10.1097/MD.0000000000002630.
Considering limitations of liver biopsy for diagnosis of nonalcoholic liver disease (NAFLD), biomarkers' panels were proposed. The aims of this study were to establish models based on serum adipokines for discriminating NAFLD from healthy individuals and nonalcoholic steatohepatitis (NASH) from simple steatosis.This case-control study was conducted in patients with persistent elevated serum aminotransferase levels and fatty liver on ultrasound. Individuals with evidence of alcohol consumption, hepatotoxic medication, viral hepatitis, and known liver disease were excluded. Liver biopsy was performed in the remaining patients to distinguish NAFLD/NASH. Histologic findings were interpreted using "nonalcoholic fatty liver activity score." Control group consisted of healthy volunteers with normal physical examination, liver function tests, and liver ultrasound. Binary logistic regression analysis was applied to ascertain the effects of independent variables on the likelihood that participants have NAFLD/NASH.Decreased serum adiponectin and elevated serum visfatin, IL-6, TNF-a were associated with an increased likelihood of exhibiting NAFLD. NAFLD discriminant score was developed as the following: [(-0.298 × adiponectin) + (0.022 × TNF-a) + (1.021 × Log visfatin) + (0.709 × Log IL-6) + 1.154]. In NAFLD discriminant score, 86.4% of original grouped cases were correctly classified. Discriminant score threshold value of (-0.29) yielded a sensitivity and specificity of 91% and 83% respectively, for discriminating NAFLD from healthy controls. Decreased serum adiponectin and elevated serum visfatin, IL-8, TNF-a were correlated with an increased probability of NASH. NASH discriminant score was proposed as the following: [(-0.091 × adiponectin) + (0.044 × TNF-a) + (1.017 × Log visfatin) + (0.028 × Log IL-8) - 1.787] In NASH model, 84% of original cases were correctly classified. Discriminant score threshold value of (-0.22) yielded a sensitivity and specificity of 90% and 66% respectively, for separating NASH from simple steatosis.New discriminant scores were introduced for differentiating NAFLD/NASH patients with a high accuracy. If verified by future studies, application of suggested models for screening of NAFLD/NASH seems reasonable.
考虑到肝活检在非酒精性肝病（NAFLD）诊断中的局限性，人们提出了生物标志物组合。本研究的目的是建立基于血清脂肪因子的模型，以区分NAFLD患者与健康个体，以及非酒精性脂肪性肝炎（NASH）患者与单纯性脂肪肝患者。本病例对照研究纳入了血清转氨酶持续升高且超声检查显示有脂肪肝的患者。排除有饮酒证据、使用肝毒性药物、患有病毒性肝炎及已知肝脏疾病的个体。对其余患者进行肝活检以区分NAFLD/NASH。组织学结果采用“非酒精性脂肪性肝病活动评分”进行解读。对照组由体格检查、肝功能检查及肝脏超声均正常的健康志愿者组成。应用二元逻辑回归分析来确定自变量对参与者患NAFLD/NASH可能性的影响。血清脂联素降低以及血清内脂素、IL-6、TNF-α升高与患NAFLD的可能性增加相关。NAFLD判别评分如下：[（-0.298×脂联素）+（0.022×TNF-α）+（1.021×log内脂素）+（0.709×log IL-6）+1.154]。在NAFLD判别评分中，86.4%的原始分组病例被正确分类。判别评分阈值（-0.29）在区分NAFLD与健康对照时，敏感性和特异性分别为91%和83%。血清脂联素降低以及血清内脂素、IL-8、TNF-α升高与患NASH的概率增加相关。NASH判别评分如下：[（-0.091×脂联素）+（0.044×TNF-α）+（1.017×log内脂素）+（0.028×log IL-8）-1.787]。在NASH模型中，84%的原始病例被正确分类。判别评分阈值（-0.22）在区分NASH与单纯性脂肪肝时，敏感性和特异性分别为90%和66%。引入了新的判别评分以高精度区分NAFLD/NASH患者。如果未来研究证实，应用所建议的模型筛查NAFLD/NASH似乎是合理的。
