Abstract S4-06: PIK3CA mutation predicts resistance to anti-HER2/chemotherapy in primary HER2-positive/hormone-receptor-positive breast cancer – Prospective analysis of 737 participants of the GeparSixto and GeparQuinto studies

Abstract

Abstract Background: Phosphatidylinositol 3-kinase (PI3K)/AKT pathway aberrations are common in breast cancer (BC), PIK3CA mutations being the most common. Mutations are frequently found in hot-spots located in the helical and kinase domains (exons 9 and 20). Reported data is discrepant with regard to prognostic or predictive value of PIK3CA mutations especially in HER2+ve BC. We therefore investigated the frequency and prognostic associations of PIK3CA mutations in HER2+ve and triple negative (TN) primary BC by treated with neoadjuvant therapy. Methods: We prospectively evaluated PIK3CA mutations in the 512 participants of the neoadjuvant Geparsixto (G6) study (von Minckwitz et al. ASCO 2013) and validated in 225 participants of the GeparQuinto (G5) study (Untch et al. 2012). The G6 study investigates the effect of adding carboplatin to a non-pegylated liposomal doxorubicin/taxane combination for the treatment of patients with HER2+ve and TN primary BC. All HER2+ve patients received trastuzumab and lapatinib, the TN patients received bevacizumab. The G5 study showed that trastuzumab added to EC-Doc results in a significantly higher pCR rate than lapatinib. HER2, hormone receptors (HR), and Ki67 were centrally assessed in both studies. PIK3CA was genotyped in tumor material from formalin-fixed, paraffin embedded core biopsies taken before therapy with a tumor cell content of ≥20% using classical Sanger sequencing of exon 9 and 20. Results: In the G6 study, 595 patients with HER2+ve or TN primary BC have been randomized from 09/2011 to 11/2012. Median age was 47 years (range 21-78); most tumors were cT2 (65%); cN0 (57%); ductal invasive (93%), grade 3 (65%); within the HER2+ve group 62% were HR-positive. Currently, PIK3CA genotype is available from 512 randomized patients - 240 with HER2+ve and 272 with TN disease. Overall, 13.1% were found to have at least one mutation, in HER2+ve: 19.2% and TNBC: 7.7%. PIK3CA mutations were numerically more frequent in the HER2+ve/HR+ve compared to the HER2+ve/HR-ve group: 21.5% vs 15.4% respectively (p = 0.245. Overall, pCR rate was significantly lower in the PIK3CA mutant compared to wt group (22.7% vs. 43.6%; p = 0.001).This effect was only significant within the HER2+ve group (17.8% vs. 36.8%; p = 0.015) compared to TNBC (33.3% vs. 49%; p = 0.168). Within the HER2+ve/HR+ subgroup the PIK3CA mutant pts had a pCR rate of only 6.5% compared to 30.8% in the wt group (p = 0.005). In contrast there was no difference in pCR (42.9% vs. 46.1%) according to PIK3CA mutation status in the HER2+ve/HR-ve (p = 0.825) group. In the G5 study, 225 of 620 HER2+ve pts have biomaterial available for PIK3CA genotyping and central confirmation of HER2 and hormone-receptor status. The analyses are ongoing and the results for the trastuzumab and lapatinib treated cohorts will be presented at the meeting. Conclusion: Pts with PIK3CA mutant HER2+ve/HR+ve breast cancer are resistant to chemotherapy and dual anti-HER2 treatment. Other treatment options are needed to be tested in this group. The project has been funded within the EU-FP7 project RESPONSIFY No 278659. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr S4-06.