Abstract A005: MGMT as a pan-cancer biomarker in temozolomide preclinical studies using models from the NCI PDMR

Abstract Introduction: Deficiency in MGMT expression, through low gene/protein expression and/or silencing of the MGMT promoter along with a functional DNA mismatch repair pathway, significantly increases a cancer cell’s sensitivity to temozolomide (TMZ) treatment, particularly in glioblastoma. To understand whether MGMT deficiency affects response to TMZ in other cancer types, preclinical drug studies were performed in patient-derived xenograft (PDX) models selected from the NCI Patient-Derived Models Repository (PDMR: https://pdmr.cancer.gov). MGMT deficiency was measured using a CLIA-validated promoter methylation assay (PMR assay), protein expression by immunohistochemistry (IHC), and gene expression by RNAseq. Here we evaluated the concordance of MGMT status across different assays and their correlations with response to TMZ. Methods: Eighty-eight pan-cancer PDX models, including solid tumors and sarcomas, were used in TMZ preclinical studies. MGMT deficiency was determined using the following criteria: promoter methylation ratio ≥ 2 by PMR assay, low protein expression (tumor nuclear staining < 30%) by IHC, or low gene expression (z-score of normalized count < -1.65) by RNAseq. For each model, PDX samples from 2 passages were evaluated to assess consistency in MGMT promoter methylation and protein expression during passaging. The survival metric of relative median delay in tumor quadrupling for event-free survival (EFSx4) was used to quantify tumor growth delay of the models treated with TMZ. Results: Out of 88 models, 7 models were identified as MGMT deficient and 71 models as MGMT proficient across all three assays, resulting in an assay concordance rate of 88.64% (78/88). Among the 10 discordant models, 3 were MGMT promoter unmethylated, but showed low MGMT protein and gene expression. Conversely, 5 models were MGMT promoter methylated, but had high MGMT gene and protein expression. Possible reasons for these discrepancies include differences in sensitivity of the assays and post-transcriptional/translational regulation of MGMT gene/protein expression. One urothelial/bladder cancer model showed inconsistent IHC results and one Hurthle cell neoplasm showed inconsistent PMR results between 2 passages suggesting tumor heterogeneity within these models. Comparing the EFSx4 response metric and MGMT status, MGMT-deficient models treated with TMZ had a mean EFSx4 of 1.9, 1.8, and 2.1 when detected by PMR assay (n=13), IHC (n=10), and RNAseq (n=11), respectively. These durations were statistically significant (p<0.0001, Wilcoxon rank sum test) compared with MGMT proficient models treated with TMZ (mean EFSx4 = 1.1). Conclusion: Three MGMT assays showed good concordance in TMZ preclinical studies using PDMR models. Substantial responses to TMZ suggest that this drug represents an effective treatment for patients with MGMT deficiency, regardless of cancer type. Citation Format: Li Chen, Shahanawaz Jiwani, Lindsay Dutko, Vishnu Rahul Kannan, Maria Saeed, Alyssa Chapman, Lisa Brown, Nikitha Nair, Corinne Camalier, Biswajit Das, Ting-Chia Chang, Chelsea McGlynn, Howard Stotler, Shannon Uzelac, Justine Mills, Tiffanie Chase-Miner, Cindy Timme, Yvonne Evrard, Laura Kuhlmann, Dianne Newton, Melinda Hollingshead, Chris Karlovich, Mickey Williams, James Doroshow. MGMT as a pan-cancer biomarker in temozolomide preclinical studies using models from the NCI PDMR [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2025 Oct 22-26; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2025;24(10 Suppl):Abstract nr A005.