Correlation of receptor activator of nuclear factor kappa b (RANK) expression in breast cancer (BC) at the time of diagnosis with recurrence-free survival (RFS) and risk of bone-dominant metastases (BDM) in the I-SPY 1 trial (CALGB 150007/150012; ACRIN 6657).

Abstract

2 Background: RANK is a tumor necrosis factor receptor family protein that is critical to bone turnover. The ligand to RANK (RANKL) can induce metastases from RANK-expressing tumor cells, and treatment with a RANKL inhibitor can reduce BC metastases (mets) in pre-clinical models and there is an FDA approved drug that targets RANKL. Osteoprotegerin (OPG) is a RANKL decoy receptor that blocks the interaction of RANK/RANKL. I-SPY 1 was a multicenter neoadjuvant trial with well annotated gene expression data and detailed clinical outcomes; we studied the impact of RANK expression on both RFS and BDM. Methods: OPG/RANK/RANKL pathway expression in core biopsies was evaluated at diagnosis in patients (pts) on I-SPY 1, then correlated with ER, intrinsic subtype, stage, response to chemotherapy, site of mets and RFS using the student t-test. Results: Gene arrays were performed on pretreatment tumor from 221 pts; 149 pts have available data with n being ER+. At a median follow-up of 3.5 yrs, 36 pts developed recurrence; 13 pts had BDM, 15 had non-bone mets (NBDM), and 8 pts were not evaluable. RANK, but not OPG or RANKL, was more highly expressed in ER- vs ER+ tumors (p=0.04) and in basal vs luminal A/B subtypes (p= 0.04), but no difference in expression based on tumor grade, node status or response to chemotherapy. Higher (>50%) RANK and lower (<50%) OPG expression correlated with worse RFS (p=0.03). Pts who developed BDM had higher expression of RANK, but not RANKL or OPG compared those with NBDM (p=0.05). After adjusting for ER and intrinsic subtypes, RANK expression was even more significantly correlated with risk for BDM (p=0.004). Conclusions: Higher RANK and lower OPG expression in primary tumor tissue correlates with poor RFS, and high RANK expression is associated with increased risk for BDM. These data suggests that targeting the OPG/RANK/RANKL pathway could be a promising strategy in preventing BDM in pts with high risk BC.