CT-based Radiomics-clinical Nomogram For Predicting Synchronous Distant Metastasis And Outcome In Clear Cell Renal Cell Carcinoma

Part Ⅰ A CT-based radiomics-clinical nomogram for prediction of synchronous distant metastasis in patients with clear cell renal cell carcinomaObjectiveRenal cell carcinoma(RCC)is the most common malignant tumor in adult kidneys,and the clear cell renal cell carcinoma(ccRCC)is the most common type.As an extremely aggressive urological neoplasm,ccRCC with synchronous distant metastases has a poorer prognosis.Early identification of distant metastases can ensure the effective treatments were timely offered and eventually improve patients survival.The aim of this study was to construct and verify a model combined CT-based radiomics with clinical factors for prediction of synchronous distant metastases in ccRCC patients.MethodsThis retrospective study recruited 578 patients(training cohort:n=325;internal validation cohort:n=138;external validation cohort:n=115)with ccRCC between January 2012 and December 2020.Demographics(including sex,age,weight,coronary heart disease,hypertension,diabetes,history of smoking),laboratory variables(including hemoglobin,calcium,creatinine,Red blood cell distribution width-to-Lymphocyte ratio[RLR],Neutrophil-to-Lymphocyte Ratio[NLR],Platelet-to-Lymphocyte Ratio[PLR],Albumin-to-Fibrinogen ratio[AFR])and CT findings(maximum diameter,tumor side,location,margin,shape,enhancement degree,necrosis)were analyzed to develop clinical score(Cli-score).Radiomics features were extracted from contrast enhanced CT images.A radiomics score(Rad-score)was developed based on reproducible features by means of the least absolute shrinkage and selection operator(LASSO)method.Integrating Rad-score and Cli-score,a Combined-Nomogram was developed.The performance of the Combined-Nomogram was determined by calibration,discrimination,and clinical usefulness.ResultsThe rates of ccRCC with synchronous distant metastases were 27.4%(89 of 325),27.5%(38 of 138),and 25.2%(29 of 115)in the training,internal validation,and external validation cohort,respectively,whereas no statistically significant difference was found(P=0.891).By incorporating the age,sex,maximum diameter,shape,margin,hemoglobin,calcium,and AFR,a Cli-score was developed,which yielded an area under the curve(AUC)of 0.924,0.896,and 0.920 in the training,internal validation,and external validation cohort,respectively.Ten features were used to build Rad-score,which yielded an AUC of 0.871,0.869,and 0.882 in the training,internal validation,and external validation cohort,respectively.The Combined-Nomogram showed better discrimination,with an AUC of 0.929,0.916,and 0.925 in the training,internal validation,and external validation cohort,respectively.The Combined-Nomogram showed good calibration.Decision curve analysis(DCA)demonstrated the clinical usefulness of the Cli-score and Combined-Nomogram.ConclusionThe Combined-Nomogram,integrating Rad-score and Cli-score,could help predict synchronous distant metastases in ccRCC,which might provide assistance in clinical decision-making process.Part Ⅱ A CT-based radiomics-clinical nomogram for prediction of recurrence in patients with localized clear cell renal cell carcinomaObjectiveApproximately 30%of patients with localized ccRCC will relapse after surgical excision.Identification of high-risk ccRCC is key for the selection of patients for adjuvant treatment who are at truly higher risk of recurrence.The aim of this study was to construct and verify a model combined CT-based radiomics with clinical factors for prediction recurrence after resection of localized ccRCC.MethodsThis retrospective study recruited 327 patients(training cohort:n=157;internal validation cohort:n=67;external validation cohort:n=103)with localized ccRCC between January 2012 and December 2018.After resection,patients were followed up by means of imaging examinations every 3 to 12 months during the first 2 years and then 12 months thereafter.Recurrence-free survival(RFS)was defined as the time from the date of surgery to the date of first recurrence,or last follow-up.Demographics,nephrectomy type,Fuhrman grade,laboratory variables and CT findings were analyzed to develop clinical factors model(Cli-score).Radiomics features were extracted from contrast enhanced CT images.A radiomics signature(Rad-score)was developed based on reproducible features by means of the LASSO Cox method.Integrating Radscore and Cliscore,a Combined-Nomogram was developed.The performance of the Combined-Nomogram was determined by C-index,calibration curve,and clinical usefulness.The proportional hazards assumption of the model was verified by examining the scaled Schoenfeld residual plots.RFS probabilities were estimated by using the Kaplan-Meier method and compared with the log-rank test.ResultsRFS was similar among the training,internal validation,and external validation cohorts(P=0.442,log-rank test).The multiple cox regression analysis showed that age(P=0.006),Fuhrman grade(P=0.001)and maximum diameter(P=0.016)remained as independent risk predictors associated with RFS.The above parameters were used for building Cli-score,which yielded C-index of 0.842,0.742,and 0.739 in the training,internal validation,and external validation cohort,respectively.With use of 0.799 as cutoff score of the training cohort,the Cli-score identified two risk categories of recurrence.There were significant differences between the low-and high-risk subgroup(P<0.001 for all cohorts,log-rank test).Five radiomics features were used to build Rad-score,which yielded C-index of 0.798,0.765,and 0.725 in the training,internal validation,and external validation cohort,respectively.With use of-0.103 as Rad-score cutoff score,there were significant differences between the low-and high-risk subgroup(P<0.001 for all cohorts,log-rank test).The Combined-Nomogram showed better predictive performance,with C-index of 0.847,0.765,and 0.743 in the training,internal validation,and external validation cohort,respectively.With use of 0.466 as Combined-Nomogram cutoff score,there were significant differences between the low-and high-risk subgroup(P<0.001 for all cohorts,log-rank test).The calibration curve showed good agreement between the RFS predicted by the Combined-Nomogram and observed outcomes.DCA demonstrated the clinical usefulness of the Combined-Nomogram.ConclusionThe Combined-Nomogram,integrating Rad-score and Cli-score,could be trained and validated in a two-center cohort with a good predictive performance of recurrence in ccRCC patients.