Computed tomography-based radiomics predicts prognostic and treatment-related levels of immune infiltration in the immune microenvironment of clear cell renal cell carcinoma

Abstract

Objectives: The composition of the tumour microenvironment is very complex, and measuring the extent of immune cell infiltration can provide an important guide to clinically significant treatments for cancer, such as immune checkpoint inhibition therapy and targeted therapy. We used multiple machine learning (ML) models to predict differences in immune infiltration in clear cell renal cell carcinoma (ccRCC), with computed tomography (CT) imaging pictures serving as a model for machine learning. We also statistically analysed and compared the results of multiple typing models and explored an excellent non-invasive and convenient method for treatment of ccRCC patients and explored a better, non-invasive and convenient prediction method for ccRCC patients.

Methods: The study included 539 ccRCC samples with clinicopathological information and associated genetic information from The Cancer Genome Atlas (TCGA) database. The Single Sample Gene Set Enrichment Analysis (ssGSEA) algorithm was used to obtain the immune cell infiltration results as well as the cluster analysis results. ssGSEA-based analysis was used to obtain the immune cell infiltration levels, and the Boruta algorithm was further used to downscale the obtained positive/negative gene sets to obtain the immune infiltration level groupings. Multifactor Cox regression analysis was used to calculate the immunotherapy response of subgroups according to Tumor Immune Dysfunction and Exclusion (TIDE) algorithm and subgraph algorithm to detect the difference in survival time and immunotherapy response of ccRCC patients with immune infiltration. Radiomics features were screened using LASSO analysis. Eight ML algorithms were selected for diagnostic analysis of the test set. Receiver operating characteristic (ROC) curve was used to evaluate the performance of the model. Draw decision curve analysis (DCA) to evaluate the clinical personalized medical value of the predictive model.

Results: The high/low subtypes of immune infiltration levels obtained by optimisation based on the Boruta algorithm were statistically different in the survival analysis of ccRCC patients. Multifactorial immune infiltration level combined with clinical factors better predicted survival of ccRCC patients, and ccRCC with high immune infiltration may benefit more from anti-PD-1 therapy. Among the eight machine learning models, ExtraTrees had the highest test and training set ROC AUCs of 1.000 and 0.753; in the test set, LR and LightGBM had the highest sensitivity of 0.615; LR, SVM, ExtraTrees, LightGBM and MLP had higher specificities of 0.789, 1.000, 0.842, 0.789 and 0.789, respectively; and LR, ExtraTrees and LightGBM had the highest accuracy of 0. 719, 0.688 and 0.719 respectively. Therefore, the CT-based ML achieved good predictive results in predicting immune infiltration in ccRCC, with the ExtraTrees machine learning algorithm being optimal.

Conclusion: The use of radiomics model based on renal CT images can be noninvasively used to predict the immune infiltration level of ccRCC as well as combined with clinical information to create columnar plots predicting total survival in people with ccRCC and to predict responsiveness to ICI therapy, findings that may be useful in stratifying the prognosis of patients with ccRCC and guiding clinical practitioners to develop individualized regimens in the treatment of their patients.

Keywords: Clear cell renal cell carcinoma; Immune checkpoint inhibitors; Machine learning; Prognosis; Radiomics; Tumour immune infiltration.

Fig. 1

The ROIs were manually drawn layer by layer in ccRCC plain CT images using 3D Slicer 4.11 software

Fig. 2

Flow chart of the study

Fig. 3

Fitting scores of immune infiltration levels and clustering of samples. (A) Results of immune cell infiltration. The top panel of the heat map shows the clinical data of the samples, and the inner panel shows the infiltration of each immune cell; (B) Communication between immune cells

Fig. 4

Cluster analysis of the cohort based on gene expression levels of the fitted immune cells. (A is before fitting, B is after fitting); C. Survival curves of the two clusters before and after fitting; D. Dimensionality reduction based on PCA principal component analysis; E. Violin plots showing the ICI score values of the samples in the two clusters

Fig. 5

Multivariate Cox regression analysis calculated the survival time of ccRCC patients with immune infiltration. (A) Hazard ratio of ICI score as well as multiple clinical factors on the survival time of patients with renal cancer; (B) Prediction of patient’s survival probability based on ICI. score and multiple clinical factors including age, grade, stage, T, M) survival probability at 1,3,5 years; (C) ROC curve for single ICI-score predicting patient survival; (D) ROC curve of ICI combined with clinical information for predicting survival; (E) Correlation of our subgroups with immunotherapies

Fig. 6

(A) Distribution of features after feature extraction. A total of 1820 features, including 14 Shape features, 356 First order features, 440 GLCM features,320 GLRLM features,320 GLSZM features, 280 GLDM features and 100 NGTDM features; (B) Correlation coefficients between 31 relevant features after feature screening using Pearson’s correlation coefficient; (C) Cluster analysis of 31 features

Fig. 7

A, B: 7 highly weighted features selected for building our model using the LASSO method; C. Weights occupied by each feature

Fig. 8

(A) ROC plots of 8 models for predicting immune infiltration subgroups in the test set; (B) ROC plots of 8 models for predicting immune infiltration subgroups in the training set

Fig. 9

(A) ROC plots of the predictions of the ExtraTrees models in the training set and the test set; (B) Confusion matrix of the predictions of the ExtraTrees models in the test set; (C) Weights assigned to the 7 features of the ExtraTrees models; D, E: Histograms of the predictions of the samples in the training set (D) and the test set (E); F, G: DCA evaluation of the clinical usefulness of the ExtraTrees prediction model