Abstract PS4-05-05: EGFR amplification and PI3K pathway mutations identify a subset of breast cancers that synergistically respond to EGFR and PI3K inhibition

Abstract Background: Epidermal Growth Factor Receptor (EGFR) family signaling is commonly dysregulated in cancer by amplification or activating mutations. Although studies have investigated dual EGFR/PI3K inhibition in breast cancer, they have not determined biomarkers that predict success. Here, we present a subset of patients with EGFR amplification and PI3Kinase pathway mutations in breast cancer that can be synergistically targeted by dual EGFR/PI3K inhibition. Methods: Frequency of EGFR amplification and PI3K pathway mutations (PIK3CA, PIK3R1/2, PTEN, AKT1/2/3, TSC1/2/3) were studied through datasets from cBioPortal and Caris Life Sciences. Triple negative breast cancer (TNBC) cell lines used include BT20 (amplified EGFR, PIK3CA activating mutation), MDA-MB-468 (amplified EGFR, PTEN deletion), and MDA-MB-231 (no EGFR or PI3K alterations). EGFR amplification was determined by immunoblot and fluorescent in situ hybridization, and PI3K mutations by sequencing. Signaling was determined by immunoblot, drug synergy by cell viability, cell death by propidium iodide staining, cell cycle analysis by flow cytometry, and xenograft animal study. Caris Life Science CODEai was used to evaluate real-world overall survival (OS) obtained from insurance claims and calculated from date of tumor biopsy to last contact using Kaplan-Meier estimates. Statistical significances were determined by chi-square and Mann-Whitney U test, student’s t-test with 2 tailed comparisons assuming equal variance, one-way or two-way ANOVA were performed as indicated. P-values of ≤ 0.05 were considered significant. Results: EGFR amplification occurs in approximately 1-2.5% of breast cancer patients, more frequently in TNBC (2.45-6.7%) and ER-/HER2+ (1.3-6.5%) breast cancers, and in the molecular basal (2.33-8.1%) and HER2 enriched (1.87-5.4%) subtypes. EGFR amplified patients in cBioPortal and in the Caris datasets had worse median OS compared to those with non-amplified EGFR (mOS: 111.1 m vs 164.6 m, p=0.0269, and 21.6 m vs 32.9 m, HR 0.69, 95% CI 0.58 - 0.82, p