Introduction: Several non-invasive tests (NIT) have been reported for predicting liver fibrosis to avoid percutaneous liver biopsy (PLB).
Aim: To evaluate the performance of NIT in Tunisian patients with chronic hepatitis B (CHB).
Methods: We calculated the ASAT/platelet ratio index (APRI), GGT-to-platelet ratio (GPR), Fibrosis-4 score (FIB-4), and RDW/platelet ratio (RPR). The accuracy of NIT was compared with the Metavir score for the detection of liver fibrosis stage using the area under the ROC curves (AUROC).
Results: Seventy-seven CHB patients were included. For predicting significant fibrosis, the AUROC of GPR (0.81; CI95% [0.68-0.93]; P < 0.001) was significantly higher than that of RPR (0.67; CI95% [0.52-0.82]; P = 0.03) and FIB-4 (0.746; CI95% [0.61-0.88]; P = 0.002), but was similar to APRI (0.88; CI95% [0.79-0.97]; P < 0.001). For advanced fibrosis, the AUROC of GPR (0.93; CI95% [0.84-1]; P < 0.001) was higher than that of RPR (0.83; CI95% [0.69-0.97]; P < 0.001) and FIB-4 (0.88; CI95% [0.76-0.99]; P < 0.001), but similar to APRI (0.93; CI95% [0.87-0.99]; P < 0.001). For predicting cirrhosis, the AUROC of GPR (0.98; CI95% [0.95-1]; P < 0.001) was higher than that of APRI (0.95; CI95% [0.90-1]; P = 0.02), similar to RPR (0.99; CI95% [0.98-1]; P < 0.001) but lower than that of FIB-4 (1; CI95% [1-1]; P < 0.001). In multivariate analysis, APRI (OR = 3.78; P = 0.002) and FIB-4 (OR = 2.65; P = 0.01) were independent predictors of significant fibrosis. GPR was the only independent predictor of advanced fibrosis (OR = 4.64; P = 0.001) and FIB-4 was the independent predictor of cirrhosis (OR = 2.85; P < 0.001).
Conclusion: GPR does not demonstrate significant advantages over APRI, FIB-4, and RPR in identifying liver fibrosis in patients with chronic hepatitis B (CHB).
Introduction: Plusieurs tests non invasifs (TNI) ont été rapportés dans la prédiction de la fibrose hépatique afin d'éviter la ponction biopsie de foie (PBF). Objectif: Évaluer la performance des TNI chez les patients tunisiens atteints d'hépatite B chronique (HBC). Méthodes : Nous avons calculé les rapports suivants : ASAT/plaquettes (APRI), GGT/plaquettes (GPR), RDW/plaquettes (RPR) et le score de Fibrose-4 (FIB-4). La performance des TNI a été évaluée selon le score Métavir par la courbe ROC. Résultats : Un total de 77 patients atteints d’HBC ont été inclus. Au seuil d’une fibrose significative, l'air sous la courbe (AUROC) du GPR (0,81 ; IC95% [0,68-0,93] ; P < 0,001) était significativement plus élevé que celui du RPR (0,67 ; IC 95% [0,52-0,82] ; P = 0,03) et du FIB-4 (0,746 ; IC95% [0,61-0,88] ; P = 0,002), mais était similaire à l’APRI (0,88 ; IC95% [0,79-0,97] ; P < 0,001). Au seuil d’une fibrose avancée, l'AUC du GPR (0,93 ; IC95% [0,84–1] ; P < 0,001) était supérieur à celui du RPR (0,83 ; IC95% [0,69–0,97] ; P < 0,001) et du FIB-4 (0,88 ; IC95% [0,76–0,99] ; P < 0,001) mais similaire à l’APRI (0,93 ; IC95% [0,87–0,99] ; P < 0,001). Concernant la cirrhose, l'AUROC du GPR (0,98 ; IC95% [0,95-1] ; P < 0,001) était plus élevé que celui de l’APRI (0,95 ; IC95% [0,90-1] ; P = 0,02), similaire au RPR (0,99 ; IC95% [0,98-1] ; P < 0,001) mais inférieur à celui du FIB-4 (1 ; IC95% [1-1] ; P < 0,001). Conclusion: Le GPR ne présente pas d'avantages par rapport aux APRI, FIB-4 et RPR dans l'identification de la fibrose hépatique chez les patients atteints d’HBC.
Keywords: chronic hepatitis B; gamma-glutamyl; liver fibrosis; non-invasive biomarker; transpeptidase to-platelet ratio.