M2 subtype tumor associated macrophages (M2-TAMs) infiltration predicts poor response rate of immune checkpoint inhibitors treatment for prostate cancer

Abstract

Background: Prostate cancer (PCa) is poor response to the immunotherapy for its high heterogeneity of immune microenvironment. In this study, we aim to introduce a new immune subtype for PCa involving M2 tumour associated macrophages (M2-TAMs). Methods: Three hundred and sixty-two PCa patients and matched normal prostate tissues were selected from the Cancer Genome Atlas and Gene Expression Omnibus databases. Patients’ immune infiltration characters were then analyzed based on the gene expressions. The immune subtypes were identified by the method of unsupervised hierarchical clustering. Finally, the relationship between the M2-TAMs infiltration and anti-programmed death-ligand-1 (PD-L1) therapy was investigated in the IMvigor210 cohort. Results: PCa expressed lower immune-related genes levels compared with the adjacent normal tissues. Based on the proved immunosuppressive mechanisms in PCa, tumour patients were classified into three independent subclasses with high infiltrated cytolytic activity (CYT), M2-TAMs and regulatory T cell (Tregs), respectively. Among these subtypes, M2-TAMs infiltration subtype showed the worst clinicopathological features and prognosis compared with the other two subtypes. The results of the IMvigor210 cohort demonstrated poor response of anti-PD-L1 therapy for patients with high M2-TAMs infiltration. Conclusion: Prostate tumours involved in significant immunosuppression, and high infiltration of M2-TAMs can be applied to predict the effect of anti-PD-L1 therapy.Key Messages PCa patients can be classified into three immunotypes of high infiltrated CYT, M2-TAMS, and Tregs according to the immunosuppressive mechanisms. High M2-TAMs infiltration subtype reflected the worst clinical characters, immune infiltration, and lowest expression of immune checkpoint inhibitors among the three subclasses in PCa. High M2-TAMs infiltration predicts the low response rate of anti-PD-L1 therapy.