Novel Clinical prognostic model to predict survival in hepatocellular carcinoma after LR or LT should be considered including biological characteristics and inflammation score

Abstract

Abstract Background: Since prognostic value of neutrophils-to-lymphocytes ratio (NLR) alone in hepatocellular carcinoma (HCC) patients is limited, incorporating tumor biological characteristics may further improve its prognostic ability. Methods: Cox regression analysis was conducted to identify risk factors affecting prognosis of 430 inlcuded HCC patients and the novel NLR (noNLR) model was established by taking correlation coefficient of Cox analysis results as the weight ratio. Then nomogram was used to verify the validity and reliability of this novel model. Patients were divided into high and low noNLR index group with the cutoff value of noNLR calculated by ROC curve. Finally, the predictive value of this novel model in predicting survival benefits was evaluated. Results: The noNLR model was established with independent risk factors for prognosis including NLR, AFP, tumor size and vascular invasion. The AUC of the training (0.7090 vs. 0.7599) and validation (0.6880 vs.0.7594) samples for the nomogram and noNLR models further demostrated the superiority of noNLR model. High noNLR was closely associated with shorter survival time with high recurrence risk and lower 5-year survival rate. Conclusion: noNLR model incorporating tumor biological characteristics can be qualified as a novel model to predict the prognosis of HCC patients after LR or LT.