Synthetic lethality of ERBB2 and CCND1 in breast cancer at scale
Authors: Rishi Nair, Evan P White, Roy Khalife, Anthony M Magliocco
Journal: Molecular Cancer Therapeutics | Volume 23, Issue 6_Supplement
Publisher: The American Association for Cancer Research
Originally presented at AACR
Published June 10 2024
Abstract
Introduction: Synthetically lethal mutations offer unique molecular targets for oncologic therapy. At scale, synthetic lethality (SL) is observed when alterations to one gene alone do not correspond with worse overall survival (OS), but simultaneous expression with another gene does correspond with worse OS. Mutually exclusive gene expression is not a requirement for SL. Existing breast cancer (BC) literature suggests synthetic lethality between CKS1B and PLK1, as well as BRCA1/BRCA2 and PARP1. Unaltered RB1 and altered CCND1 are also known to exhibit SL in the HER2-deficient environment of triple-negative BC. CCND1 and ERBB2 are typically amplified in invasive ductal carcinoma, with ERBB2 amplifications correlating to larger tumor size. This relationship is not observed in invasive lobular carcinoma, although expression of E2F1, a downstream transcription factor of CCND1, is inversely correlated with tumor grade. Methods: Our in silico approach investigated SL between ERBB2 and CCND1 in breast invasive carcinoma without distinction to lobular or ductal origin (TCGA, PanCancer Atlas 2018). We identified 17 genes of interest based on molecular alterations present in >10% of the sample (n=1084) and common mentions in BC literature. SL was determined if alterations to one gene alone did not significantly worsen OS, while alterations to two genes corresponded with worse OS. Results: Twenty-three SL interactions were observed. CCND1 and ERBB2 were selected because neither gene significantly worsened OS alone (p=0.915 & p=0.0951 respectively). However, patients with alterations to both genes had significantly worse OS compared to those with alterations to one gene alone. CCND1 and ERBB2 alterations were present in 14.9% and 13.7% of the sample, respectively. HR=2.568 (p=0.0138) for SL patients compared to those with CCND1 alterations alone; HR=2.244 (p=0.0497) for SL patients compared to those with ERBB2 alterations alone. Mutual exclusivity was not significantly observed (p=0.084). Multivariate analysis was performed using Cox regression models; neither age (older or younger than 55 years) nor race (White, Black, Asian) confounded our results due to p>0.05. Conclusion: Our findings support potential therapeutic use of CCND1 targets in HER2+ BC, specifically for patients with SL between CCND1 and ERBB2. These include PLK1 inhibitors, commonly used in estrogen receptor-positive patients with CDK4/6 inhibitor resistance, along with HER2 inhibitors. A combination of MAP2K and PIK3CA inhibitors should also be investigated in HER2+ BC patients with CCND1 amplification, since HER2+ colorectal cancer patients have been demonstrated to fare better with this treatment.