A Comparative Study Of The Performance Of Three Risk Scores On The Risk Stratification Of Liver Cancer In Rural Residents Of Jilin Province With Chronic Hepatitis C

Objectives: Hepatocellular carcinoma(HCC)is a common cancer,and groups such as patients with hepatitis C virus(HCV)are at risk for HCC.Risk stratification is used to identify highrisk groups among the risk population,and screening and surveillance of HCC with different intensities and frequencies according to risk stratification is of public health importance.There are several models for HCC risk assessment,including the REVEAL-HCV(Risk Evaluation of Viral Load Elevation and Associated Liver Disease/Cancer-Hepatitis C Virus The a Map(age-Male-ALBI-Platelets)score and the THRI(Toronto HCC risk index)have not been validated in rural communities in northeastern China.Previous studies by our team found a high prevalence of HCV infection in the left bank area of the Lalin River in Fuyu,Songyuan City,Jilin Province,and there is an urgent need for risk stratification of local patients with chronic hepatitis C for hepatocellular carcinoma screening and surveillance.In this study,we evaluated the effectiveness of the REVEAL-HCV score,a Map score and THRI in stratifying the risk of hepatocellular carcinoma in patients with chronic hepatitis C through a prospective cohort study of chronic hepatitis C patients and explored the prospect of using the three scores in a liver cancer screening and surveillance program in a community-based risk population.Methods: A prospective cohort of patients with chronic hepatitis C was established based on a cross-sectional survey conducted in five townships on the left bank of the Lalin River in Fuyu City in September 2012,and the occurrence of hepatocellular carcinoma was confirmed by active follow-up and passive follow-up up to February 2022.The baseline survey data were used to calculate the REVEAL-HCV score,a Map score,and THRI.survival curves were plotted based on follow-up data,and one-way Cox regression was performed to estimate the HR(hazard ratio)and its 95% confidence interval(CI),and the three scores were analyzed for risk stratification The effect of the three scores for risk stratification was analyzed.The differences in the effectiveness of the three scores for risk stratification were compared by sensitivity,specificity,Yoden index,negative predictive value,negative likelihood ratio,Harrell’s C index,time-dependent Brier score,ROC(subject work characteristic)curve and time-dependent AUC(area under the ROC curve).Results: A total of 1059 patients with chronic hepatitis C were enrolled in the study,with a median follow-up time of 97.9 months.608,421,and 30 patients were in the low-,intermediate-,and high-risk groups distinguished by the REVEAL-HCV score,respectively,with a 5-year cumulative incidence of hepatocellular carcinoma of 1.1%,9.1%,and 31.8%,respectively;341,545,and 173 patients were in the low-,intermediate-,and high-risk groups distinguished by the a Map score,respectively.The 5-year cumulative incidence of liver cancer was 1.0%,3.1%,and 20.3% for the low,intermediate,and high-risk groups distinguished by a Map score of 341,545,and 173 patients,respectively;and the 5-year cumulative incidence of liver cancer was 0.7%,2.5%,and 9.8% for the low,intermediate,and high-risk groups distinguished by THRI of 165,474,and 420 patients,respectively.Relative to the low-risk group,the HRs were 6.15(95% CI: 3.61,10.46)and 24.98(95% CI: 12.46,50.09)for the intermediate and high-risk groups differentiated by REVEAL-HCV score,and 2.72(95% CI: 1.26,5.86)and 18.02(95% CI: 8.59,37.79),and the HR in the intermediate and high risk groups distinguished by THRI was 4.48(95% CI: 1.06,18.83)and 14.36(95% CI: 3.52,58.59).The sensitivity of the REVEAL-HCV score,a Map score and THRI at year 5 were 0.88,0.94 and 0.98,respectively,with specificities of 0.60,0.34 and 0.16,Youden indexes of 0.48,0.28 and 0.14,all with negative predictive values of 0.99 and negative likelihood ratios of 0.20,0.18 and 0.12.Except for the REVEAL-HCV score,the specificity was higher than that of the REVEAL-HCV score.HCV score had higher specificity than the a Map score and THRI,the differences between the three scores were not significant in any other index.The C-index for REVEAL-HCV score,a Map score and THRI was 0.741(95% CI: 0.700,0.782),0.761(95% CI: 0.716,0.806)and 0.686(95% CI: 0.645,0.727).Time-dependent AUC and ROC curves showed higher AUC for the a Map score than for the REVEAL-HCV score and higher REVEAL-HCV score than for the THRI.Conclusions: Patients with chronic hepatitis C in the low-,intermediate-,and high-risk groups distinguished by a Map score,REVEAL-HCV score,and THRI had sequentially higher cumulative incidence of hepatocellular carcinoma,and patients with chronic hepatitis C in the low-,intermediate-,and high-risk groups distinguished by a Map score and REVEAL-HCV score had sequentially higher risk of developing hepatocellular carcinoma.a Map score,REVEAL-HCV score,and THRI had sequentially higher sensitivity,Youden index,negative predictive value,negative likelihood ratio,C index and time-dependent Brier score of risk stratification of study subjects were not significantly different.a Map score had better specificity than a Map score and THRI.time-dependent AUC versus ROC curves showed that a Map score had better predictive performance than REVEAL-HCV score and THRI.The a Map score and REVEALHCV score can effectively stratify the risk of liver cancer in the study population.a Map score contains more readily available variables and is more suitable for use as a stratification tool for the risk of liver cancer in chronic hepatitis C patients in rural Jilin Province.