Novel predictive models using serum ceruloplasmin levels for hepatic steatosis in patients with chronic hepatitis B infection.

Aims: To evaluate the significance of serum ceruloplasmin (CP) to diagnosis hepatic steatosis (HS) in Chronic hepatitis B (CHB) patients.

Methods: A total of 360 CHB patients with HS (n = 136) or without HS (n = 224) were included. Relationships between CP and HS degrees were analyzed by Spearman rank correlation. HS-predictive models including CP were constructed using multivariate logistic regression analysis and compared to other HS predicting indexes.

Results: Serum CP were significantly higher in CHB patients with HS than in patients without HS (P < 0.001) and were positively correlated with HS degree (r = 0.487, P < 0.001). The area under the receiver-operating characteristic curves (AUCs) of using CP to predict HS (S ≥ 1), moderate and severe steatosis (S ≥ 2) and severe steatosis (S = 3) were 0.758, 0.794 and 0.883, respectively. Multivariate analysis showed that CP, age, high density lipoprotein (HDL) and hemoglobin were independent predictors of HS, and CP, body mass index and HDL were independent predictors of moderate and severe HS. Two novel indexes for predicting HS of CHB patients were generated. The AUC of HSCHB-1 (for S ≥ 1) and HSCHB-2 (for S ≥ 2) were 0.881 and 0.916 in the training group, and 0.865 and 0.841 in the validation group, respectively. HSCHB-1 was superior to HS index (P < 0.001), fatty liver disease index (P = 0.0043) and steatosis index of patients with hepatitis B virus infection (P = 0.0029) in predicting HS in CHB patients.

Conclusions: HS of CHB patients was positively associated with serum CP. HSCHB-1 and HSCHB-2 with inclusion of CP are two novel models for predicting HS in CHB patients.