Integration of IgG and IgA autoantibodies for early diagnosis of hepatocellular carcinoma.

BACKGROUND

Autoantibodes against tumor-associated antigens (TAAs) have been recommended for the early diagnosis of malignancies. In this study, we intend to comprehensively evaluate the performances of four autoantibodies including anti-p53, CTAG1A, TIF1γ-IgG and anti-TIF1γ-IgA for the early diagnosis of hepatocellular carcinoma (HCC), and then determine an optimal panel of autoantibodies for early HCC diagnosis.

METHODS

The performances of four autoantibodies were evaluated by enzyme-linked immunosorbent assay (ELISA) for the early diagnosis of HCC with 380 retrospective serum samples. A training set comprised of 92 patients with early HCC, 72 patients with hepatic benign lesions (HBL), and 86 healthy controls (HC) was used to develop the predictive model for early HCC. And then, data obtained from an independent validation set was applied to evaluate and validate the predictive model to distinguish the early HCC from the controls (HBL + HC).

RESULTS

The results of the training set showed the levels and positive rates of four autoantibodies in early HCC group were significantly higher than that in HBL group/HC group (P < 0.01), of which anti-p53-IgG exhibited the highest AUC of 0.679, with 33.7% sensitivity at 93.7% specificity; the panel comprised of four autoantibodies showed the highest AUC for the patients with early HCC, up to 0.814 (95%CI 0.760-0.860), with 72.8% sensitivity at 84.2% specificity among all possible combinations of four autoantibodies. Additionally, this four-autoantibody panel showed the AUC of 0.824, 70.8% sensitivity at 84.2% specificity in the validation set.

CONCLUSIONS

Serum IgG autoantibodies against p53, CTAG1A and TIF1γ, and IgA autoantibody against TIF1γ present the diagnostic value for early HCC, of which anti-p53-IgG is a preferable biomarker. The panel comprised of four autoantibodies might contribute to early HCC diagnosis.