Abstract:

Objective

To evaluate the diagnostic value of GGT-Plt ratio （GPR） combined with neutrophil-lymphocyte ratio （NLR），Plt-lymphocyte ratio （PLR） and lymphocyte-monocyte ratio （LMR） for HBV-related hepatocellular carcinoma （HBV-HCC）.

Methods

Clinical data of 1119 patients with chronic hepatitis B，post-hepatitis B cirrhosis and HBV-HCC admitted to the Third Affiliated Hospital of Sun Yat-sen University from January 2012 to September 2018 were retrospectively analyzed. Among them，784 patients were assigned in the non-HCC group，648 male and 136 female，aged 18-82 years，with a median age of 43 years. 335 cases were included in the HCC group，303 male and 32 female，aged 21-83 years，with a median age of 54 years. Univariate analysis of HCC diagnosis was conducted by Mann-Whitney U test or Chi-square test. In univariate analysis，the parameters with P＜0.1 were used for collinear diagnosis，and the parameters with variance inflation factor of ＜5 were included into multivariate Logistic regression analysis to obtain the independent influencing factors，which were used to establish Logistic regression model for HBV-HCC. The diagnostic value was evaluated by the area under the ROC curve （AUC）.

Results

Multivariate Logistic regression analysis showed that male，PLR，GPR and age were the independent risk factors for HBV-HCC （OR=2.876，1.004，1.151，1.070； P＜0.05）. The AUC of Logistic regression model based on the above factors was 0.782 with a sensitivity of 0.740 and a specificity of 0.719. The AUC of this model combined with AFP was 0.855，which enhanced the diagnostic efficiency of AFP （AUC=0.778）（Z=-5.400， P＜0.05）.

Conclusions

The Logistic regression model based on GPR combined with PLR has certain clinical value for early diagnosis of HBV-HCC. This model in combination with AFP can improve the diagnostic efficiency of AFP.

Key words: Carcinoma，hepatocellular, Hepatitis B virus （HBV）, Gamma-glutamyltransferase to platelet ratio （GPR）, Platelet to lymphocyte ratio （PLR）, Logistic regression model