Relevance of ATM Status in Driving Sensitivity to DNA Damage Response Inhibitors in Patient-Derived Xenograft Models

SIMPLE SUMMARY: The inactivation of DNA damage response (DDR) pathways provides new opportunities to target cancers. One of the key DDR kinases is ATM which is reported to be mutated across a wide range of solid and haematological cancers. There is contrasting evidence on how ATM alterations in patients could enhance treatment responses to DDR inhibitors. In this pre-clinical study, we attempted to understand (1) how different types of ATM mutation correlate with protein expression/loss and (2) which ATM alterations could predict sensitivity to different DDR inhibitors both as monotherapy and combination therapy in a range of patient-derived xenograft models. ABSTRACT: Ataxia-telangiectasia mutated gene (ATM) is a key component of the DNA damage response (DDR) and double-strand break repair pathway. The functional loss of ATM (ATM deficiency) is hypothesised to enhance sensitivity to DDR inhibitors (DDRi). Whole-exome sequencing (WES), immunohistochemistry (IHC), and Western blotting (WB) were used to characterise the baseline ATM status across a panel of ATM mutated patient-derived xenograft (PDX) models from a range of tumour types. Antitumour efficacy was assessed with poly(ADP-ribose)polymerase (PARP, olaparib), ataxia- telangiectasia and rad3-related protein (ATR, AZD6738), and DNA-dependent protein kinase (DNA-PK, AZD7648) inhibitors as a monotherapy or in combination to associate responses with ATM status. Biallelic truncation/frameshift ATM mutations were linked to ATM protein loss while monoallelic or missense mutations, including the clinically relevant recurrent R3008H mutation, did not confer ATM protein loss by IHC. DDRi agents showed a mixed response across the PDX’s but with a general trend toward greater activity, particularly in combination in models with biallelic ATM mutation and protein loss. A PDX with an ATM splice-site mutation, 2127T > C, with a high relative baseline ATM expression and KAP1 phosphorylation responded to all DDRi treatments. These data highlight the heterogeneity and complexity in describing targetable ATM-deficiencies and the fact that current patient selection biomarker methods remain imperfect; although, complete ATM loss was best able to enrich for DDRi sensitivity.