Abstract 1611: 4-protein signature predicts outcome to tamoxifen treatment in estrogen receptor positive recurrent breast cancer

Abstract

Listen

Translate article

Abstract Tamoxifen is one of the most frequently used drugs in the treatment of estrogen receptor positive breast cancer. Although tamoxifen significantly increases patient survival in the adjuvant setting, a major factor that decreases its efficacy in the recurrent setting is drug resistance. The aim of our study is to identify a protein signature predicting outcome of tamoxifen treatment using high resolution mass spectrometry. A total of 112 snap frozen breast tumor tissues were gathered from patients that had received tamoxifen as first line treatment upon recurrence. Samples were divided into a discovery (n= 56; Erasmus Medical Center) and validation sets (n=56; multi-center cohort). Based on progression of disease from start of tamoxifen therapy, patients were categorized having either Good (time-to-progression; ttp>6 months) or Poor (ttp<6 months) outcome. Approximately 4,000 tumor cells (∼200ng protein) were microdissected from each tissue, which were lysed and subsequently trypsin-digested. MS analyses were performed on a nano-liquid chromatography LTQ-Orbitrap mass spectrometer. Protein identification and quantitation were performed through MaxQuant. Significantly differentially abundant proteins (p<0.05) were filtered through multivariate analysis and used to build a predictor for tamoxifen outcome. A total of 2741 proteins were identified in both datasets. Comparative proteome analysis showed differential expression of 99 proteins in the discovery set. A subset of 4 proteins (PDCD4, OCIAD1, G3BP2, and CGN) derived from the multivariate model was selected to develop a predictor for tamoxifen outcome. The 4 protein abundances in the discovery samples were used to build a receiver operating characteristic curve, in which Youden J was selected as grouping cutoff. Group prediction was tested in the validation set through log rank test, which resulted in a significant ttp difference (p=0.004; HR=2.33, 95% CI [1.30-4.16]) between Good and Poor outcome. Our signature significantly predicts Poor outcome patients with 86.7% sensitivity, 41.46% specificity, 35.1% PPV, and 89.5% NPV in the multi-center cohort. OCIAD1, PDCD4, and CGN were significantly enriched in patients with Good outcome, while G3BP2 was more abundant in patients with Poor outcome. We are currently validating these results by immunohistochemistry in a large independent cohort. In addition, we are developing a targeted, multiplexed, quantitative MS-based assay for larger scale clinical validation. Our 4-protein signature significantly predicts outcome to tamoxifen treatment for recurrent breast cancer. Large scale clinical validation will enable more tailored treatment of breast cancer patients with advanced ER+ disease. In addition, our findings may help elucidate the mechanisms underlying endocrine therapy resistance in recurrent breast cancer. Citation Format: Tommaso De Marchi, NingQing Liu, Mila Tjoa, Christoph Stingl, Marcel Smid, Maxime P. Look, Mieke A. Timmermans, Rene BH Braakman, Mark Opdam, Sabine Linn, Paul Span, Fred CGJ Sweep, John WM Martens, Theo M. Luider, John A. Foekens, Arzu Umar. 4-protein signature predicts outcome to tamoxifen treatment in estrogen receptor positive recurrent breast cancer. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 1611. doi:10.1158/1538-7445.AM2014-1611