Thandassery Ragesh B, Al Kaabi Saad, Soofi Madiha E, Mohiuddin Syed A, John Anil K, Al Mohannadi Muneera, Al Ejji Khalid, Yakoob Rafie, Derbala Moutaz F, Wani Hamidullah, Sharma Manik, Al Dweik Nazeeh, Butt Mohammed T, Kamel Yasser M, Sultan Khaleel, Pasic Fuad, Singh Rajvir
Departments of *Medicine, Division of Gastroenterology and Hepatology †Laboratory Medicine and Pathology ‡Biostatistics, Hamad General Hospital, Doha, Qatar.
J Clin Gastroenterol. 2016 Jul;50(6):518-23. doi: 10.1097/MCG.0000000000000489.
Many indirect noninvasive scores to predict liver fibrosis are calculated from routine blood investigations. Only limited studies have compared their efficacy head to head. We aimed to compare these scores with liver biopsy fibrosis stages in patients with chronic hepatitis C.
From blood investigations of 1602 patients with chronic hepatitis C who underwent a liver biopsy before initiation of antiviral treatment, 19 simple noninvasive scores were calculated. The area under the receiver operating characteristic curves and diagnostic accuracy of each of these scores were calculated (with reference to the Scheuer staging) and compared.
The mean age of the patients was 41.8±9.6 years (1365 men). The most common genotype was genotype 4 (65.6%). Significant fibrosis, advanced fibrosis, and cirrhosis were seen in 65.1%, 25.6, and 6.6% of patients, respectively. All the scores except the aspartate transaminase (AST) alanine transaminase ratio, Pohl score, mean platelet volume, fibro-alpha, and red cell distribution width to platelet count ratio index showed high predictive accuracy for the stages of fibrosis. King's score (cutoff, 17.5) showed the highest predictive accuracy for significant and advanced fibrosis. King's score, Göteborg university cirrhosis index, APRI (the AST/platelet count ratio index), and Fibrosis-4 (FIB-4) had the highest predictive accuracy for cirrhosis, with the APRI (cutoff, 2) and FIB-4 (cutoff, 3.25) showing the highest diagnostic accuracy.We derived the study score 8.5 - 0.2(albumin, g/dL) +0.01(AST, IU/L) -0.02(platelet count, 10(9)/L), which at a cutoff of >4.7 had a predictive accuracy of 0.868 (95% confidence interval, 0.833-0.904) for cirrhosis.
King's score for significant and advanced fibrosis and the APRI or FIB-4 score for cirrhosis could be the best simple indirect noninvasive scores.
许多用于预测肝纤维化的间接非侵入性评分是通过常规血液检查计算得出的。仅有有限的研究对它们的效能进行过直接比较。我们旨在比较这些评分与慢性丙型肝炎患者肝活检纤维化分期的情况。
从1602例在开始抗病毒治疗前接受肝活检的慢性丙型肝炎患者的血液检查结果中,计算出19种简单的非侵入性评分。计算这些评分各自的受试者操作特征曲线下面积及诊断准确性(参照Scheuer分期)并进行比较。
患者的平均年龄为41.8±9.6岁(1365例男性)。最常见的基因型是4型(65.6%)。分别有65.1%、25.6%和6.6%的患者出现显著纤维化、进展性纤维化和肝硬化。除天冬氨酸转氨酶(AST)与丙氨酸转氨酶比值、Pohl评分、平均血小板体积、纤维α以及红细胞分布宽度与血小板计数比值指数外,所有评分对纤维化分期均显示出较高的预测准确性。King评分(临界值为17.5)对显著和进展性纤维化显示出最高的预测准确性。King评分、哥德堡大学肝硬化指数、APRI(AST/血小板计数比值指数)和Fibrosis-4(FIB-4)对肝硬化具有最高的预测准确性,其中APRI(临界值为2)和FIB-4(临界值为3.25)显示出最高的诊断准确性。我们得出研究评分8.5 - 0.2(白蛋白,g/dL)+0.01(AST,IU/L) - 0.02(血小板计数,10⁹/L),该评分在临界值>4.7时对肝硬化的预测准确性为0.868(95%置信区间,为0.833 - 0.904)。
用于显著和进展性纤维化的King评分以及用于肝硬化的APRI或FIB-4评分可能是最佳的简单间接非侵入性评分。