BACKGROUND AND AIMS: We identified patients with nonalcoholic fatty liver disease (NAFLD) to determine the predictive value of serum markers to diagnose histological steatohepatitis (NASH). METHODS: Demographic, serological, radiological and histological variables on 95 consecutive patients with NAFLD were recorded. The serum markers studied were CK18, Hyaluronic acid, TIMP 1 and YKL 40. The NAS score and the metavir score were the histological scoring systems used. RESULTS: CK18 levels were higher in the NASH group compared to the simple steatosis group (394 +/- 53 micro/L vs 194 +/- 26 micro/L; P < 0.05). In assessing clinical effectiveness, CK18 yielded an AUC of 0.8 for NASH (cut-off value 300 micro/L gives PPV 81% and NPV 85%).The fibrosis markers showed no differences between groups. We stratified the same cohort according to liver fibrosis (F0 vs F1-F4). Fibrosis was associated with advanced age, high body mass index and type 2 diabetes. The biomarkers performed relatively poorly at identifying liver fibrosis (F1-F4), with HA performing the best (AUC 0.73); performance improved for advanced fibrosis (F3/F4) - (HA: AUC 0.77). The NAS score performed the best overall at identifying liver fibrosis (AUC 0.79). DISCUSSION: CK18 is the only biomarker studied that can identify NASH. Additionally, liver biopsy should be performed in all high risk patients to determine the standardised NAS score to identify patients at high risk of disease progression.
BACKGROUND AND AIMS: We identified patients with nonalcoholic fatty liver disease (NAFLD) to determine the predictive value of serum markers to diagnose histological steatohepatitis (NASH). METHODS: Demographic, serological, radiological and histological variables on 95 consecutive patients with NAFLD were recorded. The serum markers studied were CK18, Hyaluronic acid, TIMP 1 and YKL 40. The NAS score and the metavir score were the histological scoring systems used. RESULTS:CK18 levels were higher in the NASH group compared to the simple steatosis group (394 +/- 53 micro/L vs 194 +/- 26 micro/L; P < 0.05). In assessing clinical effectiveness, CK18 yielded an AUC of 0.8 for NASH (cut-off value 300 micro/L gives PPV 81% and NPV 85%).The fibrosis markers showed no differences between groups. We stratified the same cohort according to liver fibrosis (F0 vs F1-F4). Fibrosis was associated with advanced age, high body mass index and type 2 diabetes. The biomarkers performed relatively poorly at identifying liver fibrosis (F1-F4), with HA performing the best (AUC 0.73); performance improved for advanced fibrosis (F3/F4) - (HA: AUC 0.77). The NAS score performed the best overall at identifying liver fibrosis (AUC 0.79). DISCUSSION: CK18 is the only biomarker studied that can identify NASH. Additionally, liver biopsy should be performed in all high risk patients to determine the standardised NAS score to identify patients at high risk of disease progression.