Abstract P5-12-05: Phenotypic discordance between primary and metastatic breast cancer (MBC) in a large scale real-life multicentre French cohort

Background: Therapeutic options at diagnosis for metastatic breast cancers (MBC) differ largely according to cancer phenotype (namely, hormone receptor (HR) and HER2 status). Reported discordance rates between primary tumor and metastasis vary widely in literature, with a median of 18% for estrogen receptor, 31% for progesterone receptor and 10% for HER2. The present study aimed to compare phenotypic profiles between primary and MBC in the real-life setting. Patients (pts) and methods: Epidemio-Strategy and Medical Economics (ESME)MBC data platform (NCT03275311) is a French national, multicenter, observational cohort using clinical trials' methodology for data capture, monitoring and quality controls. At the time of analysis, it comprised data of 16703 consecutive newly diagnosed MBC pts (1/01/08-31/12/14) treated in 18 French comprehensive cancer centres. The primary endpoint was the discordance rate of HR and HER2 status between primary tumor and MBC (biopsy of metastatic site done within 6 months of MBC diagnosis). Only patients with both histological reports available were considered. Potential factors associated with phenotype discordance were assessed in a multivariate logistic regression. Results: 2933 out of 16703 (17.6%) had a biopsy in the first 6 months of metastatic disease. HR and/or HER2 status was available in 1677 pts. The discordance rate between primary and matched MBC was 14.2% (222/1566) for HR: loss of expression in 72.5%, gain in 27.5%. For HER2, the discordance rate was 7.8% (84/1076): 45.2% of losses and 54.8% of gains of expression. The primary HR+/HER2+ subgroup had the highest rate of changes: 53% (49/92) with either a loss of HR (43%), loss of HER2 (43%) or a loss of both (14%). 18% (33/181) of primary triple-negative breast cancer (TNBC) had a phenotypic change with a majority of HR gain (79%). In multivariate analysis, administration of adjuvant chemotherapy +/- targeted therapy was the sole independent predictor of HR status modification (OR: 1.73, 95%CI 1.27-2.36, p=0.001). The presence of a mixed histology was the only predictor of HER2 discordance (OR =2.57, 95%CI 1.19-5.55, p=0.016). Patient characteristics Total population (n=16703)Pts with primary and MBC phenotype available (n=1677)Age at metastatic diagnosis Median (range)61 (19-99)60 (24-93)De novo MBC4507 (27.1%)221 (13.2%)Number of metastatic sites Median (range)1 (1-9)2 (1-7)MBC sites Brain1200 (7.2%)138 (8.2%)Visceral7755 (46.4%)928 (55.3%)Non-visceral7748 (46.4%)611 (36.4%)Phenotypic profileN = 2933N=1677TNBC356 (18.5%)272 (19.3%)HR+/HER2-1251 (65.0%)917 (65.2%)HR-/HER2+150 (7.8%)105 (7.5%)HR+/HER2+168 (8.7%)112 (8%)Missing1008271Metastatic site samplingN=2933N=1677Bone692 (24.2%)419 (25.5%)Liver514 (18.0%)355 (21.6%)Skin379 (13.3%)203 (12.4%)Node306 (10.7%)169 (10.3%)Lung258 (9.0%)168 (10.2%)Pleura283 (9.9%)121 (7.4%)CNS/CSF*132 (4.6%)42 (2.6%)Other or multiple296 (10.3%)165 (10.0%)Missing7335* CNS= central nervous system, CSF=cerebro-spinal fluid Conclusion: Biopsy and phenotype re-evaluation of MBC early in the disease course has a confirmed potential significant therapeutic impact in this large scale real life setting and should be proposed as often as possible. Citation Format: Lefevre S, Lusque A, Joyon N, Arnould L, Penault-Llorca F, MacGrogan G, Treilleux I, Vincent-Salomon A, Haudebourg J, Maran-Gonzalez A, Charafe-Jauffret E, Courtinard C, Franchet C, Verriele V, Brain E, Tas P, Delaloge S, Filleron T, LaCroix-Triki M. Phenotypic discordance between primary and metastatic breast cancer (MBC) in a large scale real-life multicentre French cohort [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P5-12-05.