A Predictive Model Of Early Recurrence Of Primary Hepatocellular Carcinoma And Construction And Verification Of Nomogram

Objectives:Hepatocellular Carcinoma(HCC)is a malignant tumor with a high recurrence rate.However,the current staging system lacks the ability to predict early recurrence of HCC.We aimed to explore the risk factors for early postoperative recurrence of primary liver cancer(within 2 years postoperatively)and to construct a model for predicting early recurrence of liver cancer in order to facilitate the development of clinical individualized treatment options.Methods:Retrospective analysis of clinical data of 137 patients with primary hepatocellular carcinoma who underwent surgery from April 2017 to April 2018,the First Affiliated Hospital of Kunming Medical University,Collect the patient’s age,body mass index(BMI),gender,drinking history,diabetes history(Diabetes,DM),liver cirrhosis background,hepatitis B disease history,tumor maximum diameter(cm),tumor number,tumor location,hemoglobin,Neutrophil to lymphocyte ratio(Neutrophil to lymphocyte ratio,NLR),serum albumin,total bilirubin,direct bilirubin,indirect bilirubin,serum alpha-fetoprotein(AFP),Child-Pugh classification,albumin bilirubin score(Albumin-Bilirubin,ALBI),tumor differentiation,macrovascular invasion,microvascular invasion(MVI)and other data.The cases bounded within 2 years were the early relapse group and Late relapse group.According to the ratio of 8:7,137 patients were randomly divided into training set and verification set to construct and verify the predictive model.COX regression analysis was used to screen the independent risk factors of early recurrence of liver cancer,and then COX regression was used to construct liver cancer.After the early recurrence of the nomogram model,the receiver operating characteristic curve(ROC)was used to evaluate the discrimination degree of the prediction model,and the calibration curve was used to evaluate the calibration degree of the prediction model.Finally,the clinical decision curve was used to evaluate the prediction model.Clinical value and the model has a high degree of calibration(P>0.05).P value was obtained by Kruskal Wallis rank sum test.If the count variable has a theoretical number<10,it is obtained by Fisher’s exact probability test.Results:Follow-up was performed on 137 cases of liver cancer after radical operation.The total number of cases that relapsed within 2 years was 43(31.4%),and the number of cases that did not relapse within 2 years was 94(68.6%).The shortest recurrence time during the follow-up was 2 months.There was a statistically significant difference in the history of hepatitis B,the largest tumor diameter,and the number of tumors between the early relapse group and the early non-relapse group(P<0.05).There was no statistical difference between the training group and the verification group(P>0.05),single factor COX regression analysis showed that history of hepatitis B,neutrophil lymphocyte ratio,tumor number,and tumor maximal were the potential risk factors for early recurrence of liver cancer P<0.05.Multivariate COX regression analysis showed history of hepatitis B and neutrophil lymphocytes.The independent risk factors for the early postoperative recurrence of liver cancer were P<0.05.The above four risk factors were included in the prediction model constructed by COX.The area under the ROC curve of the nomogram is 0.7279,the sensitivity is 0.7500,the specificity is 0.6667,and the accuracy is 0.6875.The area under the ROC curve of the verification set is 0.7053,the sensitivity is 0.4815,the specificity is 0.8478,and the accuracy is 0.7123.The model has a high degree of calibration(P>0.05).The decision curve proves that the nomogram has high clinical application value.Conclusions:The history of hepatitis B,the ratio of neutrophils to lymphocytes,the number of tumors,and the maximum tumor length are independent risk factors for early recurrence after radical resection of liver cancer.The alignment chart composed of these four factors has high predictive power and clinical practical value.For patients with a history of hepatitis B,low lymphocytes,large liver cancer,and multiple lesions,early personalized adjuvant therapy should be used to reduce early recurrence.