Therapeutic potential of AURKA inhibition in ER+ inflammatory breast cancer

Background: ER+ inflammatory breast cancer (IBC) is under-appreciated and under-investigated as an aggressive subset of ER+ breast cancer, despite having rates of metastasis comparable to triple negative (TN) IBC. Recent genome-wide profiling of tumors from patients with metastatic breast cancer has shown that IBC samples have a higher frequency of AURKA amplification than non-IBC samples. Our pilot studies in IBC validate these findings and suggest AURKA is amplified in TNIBC. We hypothesize that Aurora Kinase A (AURKA) is a viable therapeutic target in metastatic IBC and in particular ER+ IBC and may exhibit synthetic lethality with ARID1A and SMARCA4 in IBC. Methods: We reviewed RNA and DNA data from clinically ordered BostonGene tumor assays for IBC versus non-IBC at MDACC. Using ER- IBC cell lines we examined the protein expression of AURKA and Alisertib efficacy was tested using WST assay in IBC cell lines. Senescence was examined using beta-gal staining and ELISA assay. In vivo efficacy was studied in a novel ER+ IBC patient derived xenograft (PDX) MDA-BCM-IBC-102. Results: We previously demonstrated using a large multi-institutional IBC data set including 137 IBC patient tissues analyzed using Affymetrix gene expression arrays, no clear clustering among ER response genes by ER+ subtypes and high AURKA expression in both ER+ and ER- cases. Comparing 23 ER+ IBC to 164 ER+ NON-IBC patient samples from Boston Gene, we find AURKA is amplified in 70% cases and a statistically significant inverse relationship between AURKA and SMARCA4 in IBC. Expression of AURKA by immunoblotting in IBC cell lines demonstrates AURKA expression in four IBC cell lines, including two HER2+ (IBC3 and KPL4) and two triple negative (A3250 and SUM149). Higher doses of Alisertib promote reversible senescence assayed by SA-betagal assay and high IL6 by ELISA validated senescence associated secretory phenotype. In collaboration with the PDX program at Baylor College of Medicine led by Dr. Lewis, our team has developed first novel ER+ PDX model from an ER+ IBC patient. In vivo study demonstrated combination treatment with Alisertib and fulvestrant significantly inhibited tumor growth in ER+ PDX models at 28 days compared to either monotherapy, indicating potential synergistic efficacy. Conclusions: AURKA is commonly amplified in IBC and often associated with gene alterations that may predict synthetic lethality to AURKA targeting. In ER+ IBC AURKA amplification is inversely correlated with putative synthetic lethal partner SMARCA4. We have characterized IBC models with high AURKA expression and demonstrated sensitivity to AURKA inhibition with fulvestrant. Senescence at higher doses of Alisertib in TNIBC models may provide a direction for targeting resistance or escape. Further studies in ER+ organoids are ongoing.