| Objective: The purpose is to study the application value of combined detection of preoperative system inflammation indexes and multiple serum tumor markers in predicting colorectal cancer liver metastasis.Methods: The study selected the clinical data and follow-up data of colorectal cancer patients who underwent surgical treatment of colorectal cancer in the First Affiliated Hospital of Guangxi Medical University from January 2016 to June2018.Finally,702 complete medical records were included.The training set selected 123 patients with liver metastases from colorectal cancer from January2016 to December 2017 as the experimental group.Based on gender,age,BMI,and primary tumor location,123 patients were selected as control group from 478 patients with colorectal cancer without liver metastasis by PSM analysis.The study selected 101 patients with colorectal cancer from January 2018 to June 2018 as the validation set.The study records all patients’ preoperative blood tests and serum carcinoembryonic antigen(CEA),carbohydrate antigen l9-9(CAl9-9),carbohydrate antigen 12-5(CA12-5),carbohydrate antigen 15-3(CA15-3),Carbohydrate antigen 72-4(CA72-4),carbohydrate antigen 24-2(CA24-2)and alpha-fetoprotein(AFP)test results.The study relied on whether colorectal cancer patients had liver metastasis as the dependent variable.By LASSO regression analysis,preoperative system inflammation indexes and the above tumor markers.The variables screened out by the study are predictive variables.The research establishes a Logistic prediction model and conducts a preliminary test on the prediction efficiency of the model.And draw visualization tool nomogram and web tools to show the prediction model.Results:1.Among clinicopathological factors,patients with liver metastases from colorectal cancer have a significantly higher N stage after surgery(P<0.05).Postoperative tumor N stage is an independent risk factor for liver metastasis in colorectal cancer patients(P<0.05).2.Among the preoperative system inflammation indicators,SII,NLR and PLR of patients with colorectal cancer liver metastasis were significantly increased(P<0.05),PNI was significantly decreased(P<0.05),and m GPS score was significantly increased(P<0.05).SII is an independent risk factor for liver metastasis in colorectal cancer patients(P<0.05).3.Among the preoperative tumor markers,the levels of CA19-9,CA12-5and CA72-4 in patients with liver metastases of colorectal cancer were significantly increased(P<0.05).The levels of CEA,CA12-5 and CA72-4 were not independent risk factors for liver metastasis in patients with colorectal cancer(P>0.05).4.In other blood tests before surgery,the AST level of patients with colorectal cancer liver metastases increased significantly(P<0.05).5.In the construction and verification of the prediction model,the Logistic prediction model is constructed by the 6 factors obtained after the LASSO regression variable screening.Among them,PLR takes the best cutoff value of118.96.Based on the principle that the higher the PLR is,the worse the prognosis is,the PLR greater than 118.96 is counted as 1 and the less than 118.96 is counted as 0.The results showed that N staging and SII were independent factors affecting postoperative liver metastasis.The prediction model has an AUC value of 0.709,95% CI: 0.645-0.773,a sensitivity of 62.6%,a specificity of 72.4%,a cutoff value of 0.49,and a Brier score of 0.213.The AUC value verified by the validation set was 0.694,95% CI: 0.572-0.817,sensitivity was 69.0%,specificity was 65.3%,cut-off value was 0.41,Brier score was 0.197,calibration intercept was-0.629,and calibration slope was 1.19.Conclusions:1.SII and PLR in preoperative system inflammation indexes,CEA,CA12-5and CA72-4 levels in preoperative tumor markers,and postoperative tumor N stage among clinicopathological factors are all risk factors affecting liver metastasis in patients with colorectal cancer.SII and postoperative tumor N stage are independent risk factors for liver metastasis in patients with colorectal cancer.2.ROC curve test results and internal data verification results show that the prediction model constructed by preoperative system inflammation indexes,serum tumor markers and clinicopathological factors has a good predictive ability for colorectal cancer patients with liver metastasis.The constructed nomogram prediction model and web tool(https://jin183176989.shinyapps.io/prediction＿model＿of＿crclm/)help to facilitate the clinical application of the model. |
PDF Full Text Request