LCK and CD3E Orchestrate the Tumor Microenvironment and Promote Immunotherapy Response and Survival of Muscle-Invasive Bladder Cancer Patients

Abstract

Background: Studies have demonstrated the significance of multiple biomarkers for bladder cancer. Here, we attempt to present biomarkers potentially predictive of the prognosis and immunotherapy response of muscle-invasive bladder cancer (MIBC). Method: Immune and stromal scores were calculated for MIBC patients from The Cancer Genome Atlas (TCGA). Core differential expression genes (DEGs) with prognostic value were identified and validated using an independent dataset GSE31684. The clinical implications of prognostic genes and the inter-gene correlation were presented. The distribution of tumor-infiltrating immune cells (TICs), the correlation with tumor mutation burden (TMB), and the expression of eight immune checkpoint-relevant genes and CD39 were accordingly compared. Two bladder cancer cohorts (GSE176307 and IMvigor210) receiving immunotherapy were recruited to validate the prognostic value of LCK and CD3E for immunotherapy. Results: 361 MIBC samples from TCGA revealed a worse overall survival for higher stromal infiltration (p = 0.009) but a better overall survival for higher immune infiltration (p = 0.042). CD3E and LCK were independently validated by TCGA and GSE31684 to be prognostic for MIBC. CD3E was the most correlative gene of LCK, with a coefficient of r = 0.86 (p < 0.001). CD8⁺ T cells and macrophage M1 are more abundant in favor of a higher expression of CD3E and LCK in MIBC and across pan-cancers. Immune checkpoints like CTLA4, CD274 (PD-1), and PDCD1 (PD-L1) were highly expressed in high-CD3E and high-LCK groups for MIBC and also for pan-cancers, except for thymoma. LCK and CD3E had a moderate positive correlation with CD39 expression. Importantly, high-LCK and high-CD3E groups had a higher percentage of responders than the low-expression groups both in GSE176307 (LCK: 22.73vs. 13.64%, CD3E: 22.00 vs. 13.16%) and IMvigor210 cohorts (LCK: 28.19 vs. 17.45%, CD3E: 25.50 vs. 20.13%). Conclusion: CD3E and LCK were potential biomarkers of MIBC. CD3E and LCK were positively correlated with several regular immunotherapy biomarkers, which is supported by real-world outcomes from two immunotherapy cohorts.

Keywords: CD3e; LCK; immunotherapy; muscle-invasive bladder cancer; tumor microenvironment.

FIGURE 1

Prediction of the level of tumor-infiltrating immune cells and stromal cells. (A–B) The optimal cutoff value of the immune score and stromal score was calculated. (C–D) Prognostic value of the immune score and stromal score. (E–F) Identification of differential expressed genes according to the immune and stromal scores. A p value < 0.05 indicates statistical significance.

FIGURE 2

Identification and validation of prognostic genes of muscle-invasive bladder cancer. (A) Intersection of differential expressed genes. (B–C) Enrichment analyses and the protein–protein network construction of the intersected differential expressed genes. (D–E) Partial presentation of the prognostic genes with the core module of the protein–protein network found through TCGA samples. (F–G) External validation with the GSE31684 dataset identifies that LCK and CD3E were the prognostic genes of muscle-invasive bladder cancer. A p value < 0.05 indicates statistical significance.

FIGURE 3

Correlation between LCK and CD3E. (A–B) co-expressed genes of LCK and CD3E. (C–D) Spearman correlation and co-expression analysis between LCK and CD3E. (E) Intersection of the first 100 genes co-expressed with LCK and CD3E. (F–G) Enrichment analyses and the network pathway of the 85 intersected genes.

FIGURE 4

Clinical implication of LCK and CD3E. (A) Comparison of the expression value of LCK between normal tissue and bladder cancer tissue from TCGA; immunohistochemical staining of LCK in normal tissue (C)and bladder tumor tissue (D). (B) Comparison of the expression value of CD3E between normal tissue and bladder cancer tissue from TCGA. Immunohistochemical staining of LCK in normal tissue (E) and bladder tumor tissue (F). (G–H) Correlation of LCK and CD3E with clinical information of muscle-invasive bladder cancer.

FIGURE 5

Association of LCK and CD3E with the distribution of tumor-infiltrating immune cells and with the expression of immune checkpoint genes in muscle-invasive bladder cancer. (A–B) Distribution of tumor-infiltrating immune cells based on the expression level of LCK and CD3E. (C–D) Expression level of immune checkpoint genes based on the expression level of LCK and CD3E. * p < 0.05, ** p < 0.01, *** p < 0.001.

FIGURE 6

Correlation between tumor-infiltrating immune cells with the expression of LCK (A) and CD3E (B) across pan-cancers.

FIGURE 7

Correlation of LCK/CD3E with immune checkpoint genes and tumor mutation burden across pan-cancers. (A–B) Correlation of LCK/CD3E with immune checkpoint genes across pan-cancers. (C–D) Correlation of LCK/CD3E with tumor mutation burden across pan-cancers. (E–F) Spearman correlation between LCK/CD3E and tumor mutation burden in bladder cancer. (G–H) Spearman correlation between ENTPD1 (CD39) with LCK/CD3E in bladder cancer.

FIGURE 8

Prognostic value of LCK and CD3E for MIBC in predicting immune response and survival after immunotherapy among two independent validation cohorts. (A–D) Percentage of high/low LCK/CD3E expression in responders and non-responders. (E–H) Proportion of responder and non-responder in high/low-LCK/CD3E groups. (I–J) Percentage of high/low-LCK/CD3E expression in different immune phenotypes. (K–L) Proportion of different immune phenotypes in high/low-LCK/CD3E groups. (M–P) Kaplan–Meier curves showing the prognostic value of LCK and CD3E for MIBC overall survival in GSE176306 and IMvior210.