2018 ASCO Annual Meeting!
Session: Gastrointestinal (Colorectal) Cancer
Type: Poster Session
Time: Sunday June 3, 8:00 AM to 11:30 AM
Location: Hall A
The characteristics of ARID1A mutations in colorectal cancer.
Gastrointestinal (Colorectal) Cancer
2018 ASCO Annual Meeting
Poster Board Number:
Poster Session (Board #88)
J Clin Oncol 36, 2018 (suppl; abstr 3595)
Author(s): Amir Mehrvarz Sarshekeh, Jonathan M. Loree, Ganiraju C. Manyam, Allan Andresson Lima Pereira, Kanwal Pratap Singh Raghav, Michael Lam, Jennifer S Davis, A. Dasari, Van Karlyle Morris, David Menter, Cathy Eng, Russell Broaddus, Mark Routbort, Rajyalakshmi Luthra, Dipen M. Maru, Michael J. Overman, Funda Meric-Bernstam, Scott Kopetz; University of Texas MD Anderson Cancer Center, Houston, TX; BC Cancer, Vancouver, BC, Canada; The University of Texas MD Anderson Cancer Center, Houston, TX; Instituto do Câncer do Estado de São Paulo, São Paulo, Brazil; Department of GI Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX
Background: AT-rich interactive domain 1A (ARID1A) is a component of the SWI/SNF chromatin remodeling complex that regulates gene expression. Inactivating mutations of ARID1A have been reported in a variety of cancers but data on characteristics and associated clinicopathologic features in colorectal cancer (CRC) are limited. Methods: Data for patients (pts) with CRC whose tumors underwent comprehensive genomic profiling were reviewed using the Cancer Genome Atlas (TCGA), Nurses’ Health Study and Health Professionals’ Follow-up Study (NHS/HPFS), AACR Project GENIE and MD Anderson Cancer Center databases. Results: Among 3127 pts, 196 (6.2%) had a mutation in ARID1A. Across the datasets, 249 mutations in ARID1A were identified. Mutations were more likely to be frameshift or nonsense as compared with mutations in other genes (64.0% vs. 9.1%, OR = 7.0, 95% CI 5.6-8.7; p< .001) and the majority were considered clonal by allele frequency (defined as > 25%). The mutation locations were broadly distributed, although 10 recurrent (hot-spot) regions were identified. ARID1A mutations were associated with MSI-H status (OR = 8.1, 95% CI 4.4-14.8; p< .001), with PIK3CA mutations (OR = 2.8, 95% CI 2.1-3.9; p< .001), and BRAF mutations (OR = 3.1, 95% CI 2.2-4.4; p< .001) but had inverse correlation with TP53 mutations (OR = 0.5, 95% CI 0.4-0.7; p< .001). Of note, 18/23 (74%) of tumors with ARID1A mutations were classified as consensus molecular subtype-1 (CMS-1) with OR of 17 (95% CI 4.8-63.7; p< .001). The mutations were associated with PPAR and HNF4 transcription factor activity. ARID1A mutations were more common in early stages (OR = 1.83, 95% CI 1.09-3.07; p= 0.019) and right-sided tumors (OR = 1.66, 95% CI 1.01-2.71; p= 0.034). There was no association between ARID1A mutation and race, gender, age at the time of diagnosis, grade, or presence of distant metastases. Conclusions: This is the largest study evaluating ARID1A mutations in CRC. The majority of mutations appear to be truncating and clonal, suggesting that they have functional significance. ARID1A-mutated tumors demonstrate enrichment of wild-type TP53 but they are more likely to have MSI-H, PIK3CA and BRAF mutations. The transcriptional signature may indicate future therapeutic strategies for this subgroup.