Importance

Immune checkpoint inhibitors (ICIs) can elicit durable antitumor responses in patients with non-small cell lung cancer (NSCLC), but only 20% to 25% of patients respond to treatment. As important genes in the DNA damage response pathway, comutation in the tumor protein p53 (TP53) and ataxia-telangiectasia mutated (ATM) genes may be associated with genomic instability and hypermutation. However, the prevalence of TP53 and ATM comutation and its association with response to ICIs are not fully understood.

Objective

To examine the prevalence of the TP53 and ATM comutation, the potential mechanism, and its association with response to ICIs among patients with NSCLC.

Design, Setting, and Participants

This multiple-cohort study included patients with NSCLC from the Geneplus Institute, the Cancer Genome Atlas (TCGA), and the Memorial Sloan Kettering Cancer Center (MSKCC) databases and from the POPLAR and OAK randomized controlled trials. Samples in the Geneplus cohort were collected and analyzed from April 30, 2015, through February 28, 2019. Data from TCGA, the MSKCC, and the POPLAR and OAK cohorts were obtained on January 1, 2019, and analyzed from January 1 to April 10, 2019. Next-generation sequencing assays were performed on tumor samples by the Geneplus Institute. Genomic, transcriptomic, and clinical data were obtained from TCGA and MSKCC databases.

Exposures

Comprehensive genetic profiling was performed to determine the prevalence of TP53 and ATM comutation and its association with prognosis and response to ICIs.

Main Outcomes and Measures

The main outcomes were TP53 and ATM comutation frequency, overall survival (OS), progression-free survival, gene set enrichment analysis, and immune profile in NSCLC.

Results

Patients with NSCLC analyzed in this study included 2020 patients in the Geneplus cohort (mean [SD] age, 59.5 [10.5] years; 1168 [57.8%] men), 1031 patients in TCGA cohort (mean [SD] age, 66.2 [9.5] years; 579 [56.2%] men), 1527 patients in the MSKCC cohort (662 [43.4%] men), 350 patients in the MSKCC cohort who were treated with ICIs (mean [SD] age, 61.4 [13.8] years; 170 [48.6%] men), and 853 patients in the POPLAR and OAK cohort (mean [SD] age, 63.0 [9.1] years; 527 [61.8%] men). Sites of TP53 and ATM comutation were found scattered throughout the genes, and no significant difference was observed in the frequency of TP53 and ATM comutation within the histologic subtypes and driver genes. In 5 independent cohorts of patients with NSCLC, TP53 and ATM comutation was associated with a significantly higher tumor mutation burden compared with the sole mutation and with no mutation (TCGA, MSKCC, Geneplus, and POPLAR and OAK cohort). Among patients treated with ICIs in the MSKCC cohort, TP53 and ATM comutation was associated with better OS than a single mutation and no mutation among patients with any cancer (median OS: TP53 and ATM comutation, not reached; TP53 mutation alone, 14.0 months; ATM mutation alone, 40.0 months; no mutation, 22.0 months; P = .001; NSCLC median OS: TP53 and ATM comutation, not reached; TP53 mutation alone, 11.0 months; ATM mutation alone, 16.0 months; no mutation, 14.0 months; P = .24). Similar results were found in the POPLAR and OAK cohort in which the disease control benefit rate, progression-free survival, and OS were all greater in patients with the TP53 and ATM comutation compared with the other 3 groups (median progression-free survival: TP53 and ATM comutation, 10.4 months; TP53 mutation, 1.6 months; ATM mutation, 3.5 months; no mutation, 2.8 months; P = .01; median OS: TP53 and ATM comutation, 22.1 months; TP53 mutation, 8.3 months; ATM mutation, 15.8 months; no mutation, 15.3 months; P = .002).

Conclusions and Relevance

This study's findings suggest that the TP53 and ATM comutation occurs in a subgroup of patients with NSCLC and is associated with an increased tumor mutation burden and response to ICIs. This suggests that TP53 and ATM comutation may have implications as a biomarker for guiding ICI treatment.