Predictive Value of the TP53/PIK3CA/ATM Mutation Classifier for Patients With Bladder Cancer Responding to Immune Checkpoint Inhibitor Therapy

BACKGROUND: Only a proportion of patients with bladder cancer may benefit from durable response to immune checkpoint inhibitor (ICI) therapy. More precise indicators of response to immunotherapy are warranted. Our study aimed to construct a more precise classifier for predicting the benefit of immune checkpoint inhibitor therapy. METHODS: This multi-cohort study examined the top 20 frequently mutated genes in five cohorts of patients with bladder cancer and developed the TP53/PIK3CA/ATM mutation classifier based on the MSKCC ICI cohort. The classifier was then validated in a validation set consisting of IMvigor210 cohort and Broad/Dana-Farber cohort. The molecular profile and immune infiltration characteristics in each subgroup as defined by this classifier were explored. RESULTS: Among all 881 patients with bladder cancer, the mutation frequency of TP53, PIK3CA, and ATM ranked in the top 20 mutated genes. The TP53/PIK3CA/ATM mutation classifier was constructed based on the Memorial Sloan Kettering Cancer Center (MSKCC) ICI cohort and only showed predictive value for patients with bladder cancer who received ICI therapy (median overall survival: low-risk group, not reached; moderate-risk group, 13.0 months; high-risk group, 8.0 months; P<0.0001). Similar results were found in subgroups of MSKCC ICI cohort defined by tumor mutation burden. Multivariate Cox analysis revealed that the risk group defined by the classifier served as an independent prognostic factor for overall survival in patients with bladder cancer. Efficacy of the classifier was verified in a validation set consisting of IMvigor210 cohort and Broad/Dana-Farber cohort. Lower expression of PD-1/PD-L1 and less tumor immune infiltration were observed in the high-risk group than the other two groups of the TCGA cohort and the IMvigor210 cohort. CONCLUSION: Our study constructed a TP53/PIK3CA/ATM mutation classifier to predict the benefit of immune checkpoint inhibitor therapy for patients with bladder cancer. This classifier can potentially complement the tumor mutation burden and guide clinical ICI treatment decisions according to distinct risk levels.