To appreciate why the findings matter, it helps to understand what ATM actually does. Every day, human DNA is battered by damage from both inside and outside the cell: reactive oxygen species generated by metabolism, replication errors, hydrolysis, aging, ultraviolet radiation, and chemical exposures. Among the repair systems that respond, homologous recombination is the primary mechanism for fixing double-strand breaks, the most dangerous form of DNA injury. When both strands of the double helix snap, unrepaired lesions can trigger chromosomal abnormalities and mutagenic cascades that push cells toward malignancy. ATM sits at the command center of this response, and when it malfunctions, the consequences ripple through the entire genome.

The gene itself is a molecular heavyweight: it spans 66 exons and encodes a serine/threonine kinase whose active monomeric form weighs roughly 370 kilodaltons. When a double-strand break occurs, a sensor complex called MRN, composed of the MRE11, RAD50, and NBS1 proteins, recognizes the damage and recruits ATM to the site. Once there, ATM undergoes autophosphorylation and begins phosphorylating a battery of downstream effectors, including BRCA1, BRCA2, and CHEK2. CHEK2 acts to halt the cell cycle by inhibiting CDC25A, buying the cell time to repair. ATM also stabilizes p53, both directly and by degrading its suppressor MDMX, allowing p53 to upregulate p21 and enforce arrest at the G1/S or G2/M checkpoints. If damage proves irreparable, p53 flips the switch toward programmed cell death by activating genes such as BID, BAX, and PUMA.

Germline mutations in ATM have long been associated with ataxia-telangiectasia, an autosomal recessive syndrome marked by genomic instability, immunodeficiency, and heightened cancer predisposition. Beyond the syndrome, inherited ATM variants have been linked to pancreatic, esophageal, lung, melanoma, breast, ovarian, prostate, and bladder cancers, as well as certain lymphoma subtypes. Somatic ATM mutations appear most frequently in hematological malignancies, particularly lymphoid cancers, where frequencies range from 5 to 45 percent. Among myeloid clonal disorders, the numbers are far lower: under 5 percent in AML, about 0.4 percent in MDS, and rare in chronic myeloproliferative subsets. That relative rarity, the review argues, has led researchers to underestimate ATM’s clinical significance in these diseases.

To address that gap, the Brazilian team conducted a PROSPERO-registered systematic review, following the PRISMA 2020 guidelines and searching PubMed, SciELO, and LILACS for peer-reviewed studies published between January 2010 and August 2025. The initial search returned 80 records. After removing duplicates, screening titles and abstracts, and excluding preclinical murine and in vitro studies, case reports, letters, reviews, and unrelated diseases, only ten original studies survived the filter. Methodological quality assessment classified two studies as low risk of bias, thanks to large samples and robust observational designs, while the remainder carried moderate risk due to small cohorts, single-center recruitment, missing validation data, and heterogeneous methods for assessing ATM alterations.

Five of the ten studies focused on MDS, the most common bone marrow cancer and the disorder where ATM’s footprint is most visible. In the largest investigation, covering 232 MDS patients across three cohorts analyzed with pyrosequencing, real-time PCR, immunohistochemistry, and next-generation sequencing, ATM showed the highest methylation levels among patients who progressed to acute leukemia. Those who transformed to AML exhibited roughly double the methylation rate of those who did not, with a median methylation score of 4.97 versus 2.45, a statistically significant difference. The same team also mined 7,583 MDS cases from the cBioportal database and found ATM frequently altered by frameshift and missense mutations, most classified as likely oncogenic, reinforcing the picture of a gene silenced or downregulated during disease evolution.

Other MDS studies added nuance. A prospective analysis of 31 young, low-risk MDS patients with neutropenia and thrombocytopenia identified only heterozygous ATM variants of undetermined significance, with no clear impact on survival or transfusion needs. Two studies examining ATM polymorphisms and expression found that the rs228593 single nucleotide polymorphism was associated with increased ATM expression, particularly among patients at high or very high risk of leukemic progression, while ATM expression was reduced in hypocellular MDS but showed no survival correlation. A fifth study measuring DNA damage response proteins by immunofluorescence and immunohistochemistry in 74 patients found no significant relationship between phosphorylated ATM and clinical outcomes, leaving the prognostic value of ATM in MDS genuinely contested.

Three studies addressed AML, and here the mechanisms came into sharper focus. In two pediatric cohorts, researchers showed that ATM expression was suppressed by microRNAs: miR-100 in 30 children aged up to 14 years, and miR-181a in 57 children with a mean age of seven. When ATM was knocked down, cells accelerated through the G1/S checkpoint and myeloblast proliferation increased, offering a direct mechanistic link between ATM silencing and leukemic growth. MicroRNAs of the 181a/b cluster are known to be triggered by the PML/RARα fusion in acute promyelocytic leukemia, where they suppress the tumor suppressor RASSF1A and block granulocytic differentiation; treatment with all-trans-retinoic acid lowers these microRNAs, restores RASSF1A, and drives leukemic cells into arrest and apoptosis. In adults, a study of 307 AML patients genotyped for 42 variants across 15 DNA repair genes found that the ATM 4138C>T variant (rs3092856) was associated with striking chemotherapy refractoriness, seven of eight carriers resisted treatment, and inferior overall survival, a median of 7 versus 11 months, making it the largest AML cohort examining ATM polymorphisms to date.

Evidence from chronic myeloid leukemia and other myeloproliferative neoplasms remains thin but provocative. In a study of 476 CML patients in India, the ATM polymorphisms −5144A>T (rs228589) and C4138T (rs3092856) were associated with higher CML risk and worse scores on the EUTOS prognostic scale. A separate whole-genome sequencing study of 64 individuals from families with inherited myeloproliferative neoplasms, including essential thrombocythemia, polycythemia vera, and primary myelofibrosis, identified the germline ATM variant L2307F in some cases, although genomic instability was not evident across disease subtypes. Additional variants reported in MDS cohorts, including c.5753G>C, c.346A>G, and c.4060C>A, round out a growing but still fragmentary catalog of ATM alterations across the myeloid spectrum.

The review’s authors are candid about the limits of the evidence. The ten eligible studies varied widely in design, sample size ranging from 31 to 232 patients, laboratory methods, and measured outcomes, which precluded quantitative meta-analysis and left the overall evidence at low-to-moderate quality, best described as hypothesis-generating. Publication bias, retrospective designs, and population variability further constrain the inferences. Yet the authors argue the implications are clear enough to demand action: ATM alterations, spanning methylation, expression loss, somatic mutations, and inherited polymorphisms, behave as clinically relevant biomarkers across MDS, AML, and MPN, and the rs3092856 variant in particular emerges as a recurring thread linking chemoresistance and poor survival. Rather than treating ATM merely as a prognostic marker, the researchers call for multicenter studies with standardized methods, functional characterization of the mutations, integration of ATM status into treatment-oriented molecular panels, and exploration of therapeutic vulnerabilities, especially DNA repair inhibitors and synthetic lethality approaches that could finally convert this guardian gene’s failures into therapeutic opportunities.

Subject of Research: The prognostic and biological impact of ATM gene alterations in myeloid malignancies including MDS, AML, and myeloproliferative neoplasms

Article Title: ATM mutations in myelodysplastic neoplasms, acute myeloid leukemia, and myeloproliferative neoplasms: a systematic review

Article References: Laurindo, L. O., Parente, M. L. B., Sasahara, G. L., Goes, J. V. C., Sampaio, L. R., Pinheiro, Y. V., & Pinheiro, R. F. (2026). ATM mutations in myelodysplastic neoplasms, acute myeloid leukemia, and myeloproliferative neoplasms: a systematic review. Medical Oncology, 43(11), Article 300. https://doi.org/10.1007/s12032-026-03378-4

Image Credits: AI Generated

DOI: 10.1007/s12032-026-03378-4

Keywords: ATM gene, myelodysplastic neoplasms, acute myeloid leukemia, myeloproliferative neoplasms, DNA damage response, DNA methylation, genomic instability, rs3092856, miR-181a, chemoresistance, systematic review, precision medicine