Inflammation and Immunity Gene Expression Patterns and Machine Learning Approaches in Association with Response to Immune-Checkpoint Inhibitors-Based Treatments in Clear-Cell Renal Carcinoma

Abstract

Clear cell renal cell carcinoma (ccRCC) is the most common renal cancer. Despite the rapid evolution of targeted therapies, immunotherapy with checkpoint inhibition (ICI) as well as combination therapies, the cure of metastatic ccRCC (mccRCC) is infrequent, while the optimal use of the various novel agents has not been fully clarified. With the different treatment options, there is an essential need to identify biomarkers to predict therapeutic efficacy and thus optimize therapeutic approaches. This study seeks to explore the diversity in mRNA expression profiles of inflammation and immunity-related circulating genes for the development of biomarkers that could predict the effectiveness of immunotherapy-based treatments using ICIs for individuals with mccRCC. Gene mRNA expression was tested by the RT2 profiler PCR Array on a human cancer inflammation and immunity crosstalk kit and analyzed for differential gene expression along with a machine learning approach for sample classification. A number of mRNAs were found to be differentially expressed in mccRCC with a clinical benefit from treatment compared to those who progressed. Our results indicate that gene expression can classify these samples with high accuracy and specificity.

Keywords: TKIs; cancer; ccRCC; immunotherapy; machine learning.

Figure 1

(A) Differential gene expression results at baseline (pre-therapy) for patients with clinical benefits (CB) and progressive disease (PD) (subgroup A). Bar charts showcase differences in expression based on log₂ΔCt values for each group. Fold regulations for each gene highlight their differences after applying the 2^−ΔΔCt method. Negative fold regulation signifies downregulation in the progressive disease group. (B) Top 5 ML models that best classify the samples to either group. For each one of them, the most important features as well as their name, AUC (Area Under the Curve), and RMSE (Root Mean Square Error) values are displayed.

Figure 2

(A) Differential gene expression results for patients with Clinical Benefits (CB) before and after therapy (subgroup B). Bar charts showcase differences in expression based on log₂ΔCt values for each group. Fold regulations for each gene highlight their differences after applying the 2^−ΔΔCt method. Negative fold regulation signifies downregulation in the patients after therapy. (B) Top 5 ML models that best classify the samples into either group. For each one of them the most important features as well as their name, AUC (Area Under the Curve), and RMSE (Root Mean Square Error) values are displayed.

Figure 3

(A) Differential gene expression results for patients with progressive disease (PD) before and after therapy (subgroup C). Bar charts showcase differences in expression based on log₂ΔCt values for each group. Fold regulations for each gene highlight their differences after applying the 2^−ΔΔCt method. Negative fold regulation signifies downregulation in the patients after therapy. (B) Top 5 ML models that best classify the samples into either group. For each one of them, the most important features as well as their name, AUC (area under the curve), and RMSE (root mean square error) values are displayed.

Figure 4

Differential gene expression results after therapy for patients with clinical benefits (CB) and progressive disease (PD) (subgroup D). Bar charts showcase differences in expression based on log₂ΔCt values for each group. Fold regulations for each gene highlight their differences after applying the 2^−ΔΔCt method. Negative fold regulation signifies downregulation in the progressive ccRCC group.

Figure 5

Differential gene expression results after therapy for patients under combination of ICI/ICI or ICI/TKI treatments. Bar charts showcase differences in expression based on log₂ΔCt values for each group. Fold regulations for each gene highlight their differences after applying the 2^−ΔΔCt method. Negative fold regulation signifies downregulation in the ICI/ICI group.