The association between alterations in the ataxia telangiectasia mutated (ATM) gene and response to immune checkpoint inhibitors

Background Molecular alterations in ataxia telangiectasia mutated (ATM) gene, a DNA damage repair (DDR) gene, leads to DNA damage and a predisposition to cancer. ATM and other DDR gene alterations are emerging as potential biomarkers that predict response to immune checkpoint inhibitors (ICIs). Here, we explored the association between ATM alterations and response to ICIsMethods We used the cBioCancer Genomics Portal (cBioportal, http://www.cbioportal.org) to obtain genomic data. The prevalence of ATM alterations was calculated in the TMB and immunotherapy cohort of 1661 patients. This cohort included patients with solid tumors, including colon, melanoma, non-melanoma skin, gastroesophageal, lung, unknown primary, bladder, head and neck, breast, brain, and kidney who had received at least one dose of ICIs. Kaplan-Meier survival curves were generated and compared using a log-rank test, setting statistical significance at P < 0.05 for all comparisons.Results The prevalence of ATM alterations was 3%. Other DDR alterations, including CHEK1, CHEK2, RAD51, BRCA1, BRCA2, MLH1, MSH2, ATM, ATR, MDC1, PARP1, and FANCF were reported at <1%, <1%, <1%, 1%, 2%, 1%, 1%, 3% 1%, 0%, 1% and NP (not profiled) respectively. The median overall survival (mOS) in patients with ATM alterations (N=44) compared to patients without ATM alterations (1617) was 40 m vs. 18 m (p=0.0394). The mOS of patients with ATM alterations only (no other co-occurring DDR alterations) (N=36), compared to patients without ATM or other DDR alterations (N=1519) was 40 m vs. 17 m (p=0.0341). The mOS of patients with ATM alterations only (no other co-occurring DDR alterations) (N=36) compared to patients with any DDR but without ATM alterations (N=106) was 40 m vs. 36 m (p=0.629). When KRAS mutation, ICI type, sex and age were adjusted for a multivariable Cox regression analysis, ATM alterations were an independent predictive biomarker of response to ICIs (hazard ratio: 1.69; (1.019-2.82, P: 0.046)).Conclusions In this cohort, the prevalence of ATM alterations was 3% and it was associated with better mOS compared to patients without ATM alterations. The mOS was numerically higher in patients with ATM alterations compared to patients with other DDR alterations. If confirmed/validated in larger datasets it would highlight the importance of the distinction between ATM and other DDR alterations. It also would support considering the novel combination of ICIs and an ATM inhibitor such as AZD1390 in patients with ATM alterations.