PIK3CA and MAP3K1 alterations imply luminal A status and are associated with clinical benefit from pan-PI3K inhibitor buparlisib and letrozole in ER+ metastatic breast cancer

Mellissa J Nixon,David B Solit,Michael F Berger,Carlos L Arteaga,Luigi Formisano,Justin M Balko,M Valeria Estrada,Steven J Isakoff,Andrés Forero-Torres,Lewis C Cantley,Helen Won,Paula I González-Ericsson,Melinda E Sanders,Ingrid A Mayer,Eric P Winer

doi:10.1038/s41523-019-0126-6

Abstract

Clinical trials have demonstrated the efficacy of combining phosphoinositide 3-kinase (PI3K) inhibitors with endocrine therapies in hormone therapy-refractory breast cancer. However, biomarkers of PI3K pathway dependence in ER+ breast cancer have not been fully established. Hotspot mutations in the alpha isoform of PI3K (PIK3CA) are frequent in ER+ disease and may identify tumors that respond to PI3K inhibitors. It is unclear whether PIK3CA mutations are the only biomarker to suggest pathway dependence and response to therapy. We performed correlative molecular characterization of primary and metastatic tissue from patients enrolled in a phase Ib study combining buparlisib (NVP-BKM-120), a pan-PI3K inhibitor, with letrozole in ER+, human epidermal growth factor-2 (HER2)-negative, metastatic breast cancer. Activating mutations in PIK3CA and inactivating MAP3K1 mutations marked tumors from patients with clinical benefit (≥6 months of stable disease). Patients harboring mutations in both genes exhibited the greatest likelihood of clinical benefit. In ER+ breast cancer cell lines, siRNA-mediated knockdown of MAP3K1 did not affect the response to buparlisib. In a subset of patients treated with buparlisib or the PI3Kα inhibitor alpelisib each with letrozole where PAM50 analysis was performed, nearly all tumors from patients with clinical benefit had a luminal A subtype. Mutations in MAP3K1 in ER+ breast cancer may be associated with clinical benefit from combined inhibition of PI3K and ER, but we could not ascribe direct biological function therein, suggesting they may be a surrogate for luminal A status. We posit that luminal A tumors may be a target population for this therapeutic combination.

Highlights

Estrogen receptor-positive (ER+) breast cancer is the most common clinical subtype of breast cancer, and afflicts over 140,000 patients each year in the US.[1]
Using correlation of the expression of each tumor to the luminal A and luminal B PAM50 centroids, we found a strong correlation of tumors with both MAP3K1 and PIK3CA mutations with a luminal A-like expression pattern vs. a luminal B expression (p < 0.0001; Fig. 3a)
It is important to note that a similar association between MAP3K1 and PIK3CA alterations with outcome on alpelisib were independently identified in a preliminary report from ASCO 2018.26 there are clear caveats to modeling the molecular effects of lossof-function of MAP3K1 with PIK3CA mutations in breast cancer cell lines, we did not observe a direct effect of MAP3K1 loss on phosphoinositide 3-kinase (PI3K) inhibitor sensitivity in vitro

Summary

INTRODUCTION

Estrogen receptor-positive (ER+) breast cancer is the most common clinical subtype of breast cancer, and afflicts over 140,000 patients each year in the US.[1]. We analyzed generation sequencing data from a targeted panel of known recurrently mutated genes in cancer in a cohort of patients from two phase Ib trials of buparlisib and alpelisib, each in combination with the aromatase inhibitor letrozole The goal of this analysis was to determine possible biomarkers aside from PIK3CA mutations status that were associated with clinical benefit to the combination. The longest-responding patient, whose tumor harbored a MAP3K1 L380S mutation, was excluded in this analysis, despite the fact that this is a recurrent mutation in breast cancer (AACR GENIE data v4.1), suggesting it is a loss-of-function mutation Patients with both PIK3CA and MAP3K1 alterations appeared to achieve the greatest benefit from buparlisib and letrozole relative

RESULTS

DISCUSSION

Findings

METHODS