Effect of PNPLA3 polymorphism on diagnostic performance of various noninvasive markers for diagnosing and staging nonalcoholic fatty liver disease

Background and Aim Patatin‐like phospholipase domain‐containing protein 3 (PNPLA3) I148M (rs738409) genotype influences clinical/biochemical characteristics in patients with nonalcoholic fatty liver disease (NAFLD), but whether PNPLA3‐I148M (rs738409) genotype also influences the diagnostic performance of noninvasive diagnostic tests for NAFLD is uncertain. Our aim was to investigate the differences in diagnostic performance of noninvasive diagnostic tests for NAFLD according to PNPLA3‐I148M (rs738409) genotype. Methods Fifty‐eight healthy controls and 349 patients with biopsy‐proven NAFLD were included. Areas under the receiver operating characteristic curve (AUROCs) were calculated to predict hepatic steatosis (fatty liver index and hepatic steatosis index), nonalcoholic steatohepatitis (cytokeratin‐18 M30 and M65), and significant fibrosis (≥F2 fibrosis) (fibrosis‐4 and BARD), stratifying by rs738409 genotypes (CC and CG + GG groups). Results Fatty liver index and hepatic steatosis index showed good diagnostic performance for diagnosing steatosis only in the CG + GG group with AUROCs ranging from 0.819 to 0.832. Cytokeratin‐18 M30 (AUROC = 0.688) and M65 (AUROC = 0.678) had suboptimal performance for diagnosing nonalcoholic steatohepatitis in the CG + GG group, whereas both had good performance (AUROC = 0.814 and 0.813, respectively) in the CC group. BARD score showed good performance in the CG + GG group compared with the CC group (AUROC = 0.805 and 0.532, respectively). Fibrosis‐4 had suboptimal performance in the CG + GG group and good performance in the CC group (AUROC = 0.662 and 0.801, respectively). Conclusions Diagnostic performance of noninvasive tests for NAFLD varied markedly according to PNPLA3 genotypes. Clinicians should be aware that PNPLA3 genotype limits the clinical utility of noninvasive diagnostic tests for diagnosing NAFLD.