Abstract 5300: Loss of function of two closely linked tumor suppressors contributes to biologic aggressiveness in endometrial cancers: Co-mutation of CTCF and ZFHX3

Therapies for recurrent or advanced endometrioid endometrial cancer (EEC) are ineffective, and no targeted therapies, or genetic markers to identify high-risk patients exist. There is an urgent need to identify factors contributing to aggressiveness. We performed whole exome sequencing of 8 early stage endometrial cancers that subsequently recurred. Two chromosome 16q tumor suppressor genes, CTCF and ZFHX3, were identified as candidates associated with recurrence. The endometrial Cancer Genome Atlas proved that both of these genes are frequently mutated in EEC. To further explore the relationship between CTCF and ZFHX3 mutations and ECC tumorigenesis/aggressiveness, targeted deep sequencing of both genes was performed on 541 EEC tumors. We relied on microsatellite typing and a novel Bayesian method to assess copy number alterations (CNAs). The Bayesian method's use of deep sequencing data to predict CNAs showed 97% concordance with microsatellite typing. The overall mutation rate (SNVs and small indels) was 24% and 18% for CTCF and ZFHX3, respectively. The majority of mutations were loss of function (stop, frameshift or splice site), with CTCF at 89% and ZFHX3 at 59% loss of function defects. There was a significant excess of mutational co-occurrence in both genes (P = 0.003). Deletion/CNA was also common: 79 of 497 informative tumors had CNAs involving both genes (16%). Taken together, the high rate of co-mutation and co-deletion suggests selective advantage for inactivation of both tumor suppressors. Patients with tumors having mutation or CNA of both CTCF and ZFHX3 had significantly shorter recurrence-free survival (RFS) compared to wt (P = 0.007, HR = 2.4, 95%CI = 1.3-4.4). Mutation/loss of both genes was significantly associated with features portending aggressive disease: high grade (P60 years, P = 0.02). There was no association with stage. Although our ability to determine the contributions of CTCF and ZFHX3 individually was limited by sample size and their frequent co-occurrence, we assessed them separately as independent prognostic markers. It is noteworthy that mutation/CNA of ZFHX3 significantly correlated with shorter RFS (P = 0.003, HR = 2.4, 95%CI = 1.4-4.0), whereas no such association was seen with CTCF. In a multivariable analysis including grade, stage, LVSI and therapy, ZFHX3 mutation/CNA remained an independent prognostic marker (P = 0.04). We are currently validating these results in a larger independent cohort. Here we describe emerging markers that could be useful for identifying high-risk patients. Our data point to what may be a complex interplay between the ZFHX3 and CTCF tumor suppressors in EEC tumorigenesis. Citation Format: Christopher J. Walker, Mario A. Miranda, Matthew O'Hern, Kevin Coombes, Ralf Bundschuh, David G. Mutch, Paul J. Goodfellow. Loss of function of two closely linked tumor suppressors contributes to biologic aggressiveness in endometrial cancers: Co-mutation of CTCF and ZFHX3. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 5300. doi:10.1158/1538-7445.AM2015-5300