Bais, Carlos

Abstract Extensive biological and clinical evidence supports the notion many CRC tumors are addicted to EGFR/KRAS/MAPK signaling. For example, mutations in KRAS, BRAF and other MAPK pathway components confer intrinsic or acquired resistance to anti-EGFR therapy. However, a significant proportion of acquired resistance cases have remained unexplained. ARID1A mutations are found in approximately 11% of CRC tumors, but a connection between ARID1A mutations and EGFR/MAPK signaling activation in CRC has not been previously suspected. By leveraging the genomic biomarker data collected from 333 patients at baseline and at progression who participated in the CALGB/SWOG 80405 trial, we found that mutations in ARID1A, a key component of SWI/SNF complex, were enriched in the cell free DNA of 6/16 of the patients whose tumors had become resistant to anti-EGFR therapy with ARID1A mutations detected at either timepoint (adj. p = 0.03, OR = 0.09, two-tailed fisher's exact test). In contrast, there was no evidence of enriched ARID1A mutations in patients treated with bevacizumab. To investigate the potential role of a broader ARID1A deficiency phenotype in resistance to cetuximab therapy, we developed an ARID1A mutant-like signature from TCGA that captures the transcriptional profile characteristic of ARID1A mutants. Using this signature to stratify CALGB patients based on gene expression data collected on tissue at diagnosis, we also found patients with an ARID1A mutant-like gene expression profile had worst outcome in patients treated with cetuximab than bevacizumab with adjustment for established clinical variables for both OS [p = 0.0002, hazard ratio (HR) 6.2, 95% confidence interval (CI) 2.4-16], and PFS [p = 0.0008, HR 4.7 (1.9 -12)], suggesting that ARID1A mutations may be also a mechanism of intrinsic resistance to anti-EGFR therapies. Gene expression analysis demonstrated that ARID1A deficiency leads to re-activation of an EGFR-like signature, suggesting reactivation of this pathway as a possible mechanism of resistance. Furthermore, CRISPR knockout of ARID1A in the cetuximab-sensitive CRC cell line NCI-H508 conferred elevated MAPK signaling relative to parental line and resistance to cetuximab in culture. Consistent with these findings, we also observed strong mutual exclusivity between ARID1A mutations and mutations in the EGFR/MAPK signaling pathway in more than 40,000 lung and colorectal cancer patients profiled in the FoundationCore® database. Strikingly, in lung cancers where EGFR mutations structural variants (SV) are more prevalent, EGFR SV was the top mutually exclusive alteration with ARID1A mutations in lung cancers. Of 5980 lung cancer patients with ARID1A and/or EGFR SVs, only 100 patients (1.7%) had both SVs. Taken together, our data suggest that ARID1A loss-of-function mutations may promote resistance to cetuximab by driving an EGFR-like transcription program in the absence of ligand-dependent activation of EGFR. Our results suggest that ARID1A defects could be potentially used as an exclusion biomarker for cetuximab treatment decisions, and they provide a rationale for exploring therapeutic MAPK inhibition in ARID1A mutant CRC patients. Citation Format: Radia Marie Johnson, Xueping Qu, Joshua Thomas, Yvonne Kschonsak, Ling-Yuh Huw, Fang-Shu Ou, Ethan Sokol, Nnamdi Ihuegbu, Oliver Zill, Omar Kabbarah, Anneleen Daemen, Richard Bourgon, Alan Venook, Federico Innocenti, Heinz-Josef Lenz, Felipe de Sousa e Melo, Carlos Bais. ARID1A mutations induce an EGFR-like gene expression signature and confer intrinsic and acquired resistance to cetuximab treatment in first line metastatic CRC [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr LB-063.

Abstract Background We aimed to develop a gene expression–based prognostic signature for isocitrate dehydrogenase (IDH) wild-type glioblastoma using clinical trial datasets representative of glioblastoma clinical trial populations. Methods Samples were collected from newly diagnosed patients with IDH wild-type glioblastoma in the ARTE, TAMIGA, EORTC 26101 (referred to as “ATE”), AVAglio, and GLARIUS trials, or treated at UCLA. Transcriptional profiling was achieved with the NanoString gene expression platform. To identify genes prognostic for overall survival (OS), we built an elastic net penalized Cox proportional hazards regression model using the discovery ATE dataset. For validation in independent datasets (AVAglio, GLARIUS, UCLA), we combined elastic net–selected genes into a robust z-score signature (ATE score) to overcome gene expression platform differences between discovery and validation cohorts. Results NanoString data were available from 512 patients in the ATE dataset. Elastic net identified a prognostic signature of 9 genes (CHEK1, GPR17, IGF2BP3, MGMT, MTHFD1L, PTRH2, SOX11, S100A9, and TFRC). Translating weighted elastic net scores to the ATE score conserved the prognostic value of the genes. The ATE score was prognostic for OS in the ATE dataset (P < 0.0001), as expected, and in the validation cohorts (AVAglio, P < 0.0001; GLARIUS, P = 0.02; UCLA, P = 0.004). The ATE score remained prognostic following adjustment for O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation status and corticosteroid use at baseline. A positive correlation between ATE score and proneural/proliferative subtypes was observed in patients with MGMT non-methylated promoter status. Conclusions The ATE score showed prognostic value and may enable clinical trial stratification for IDH wild-type glioblastoma.