Abstract P1-07-15: Revisiting chromosome 17q copy number aberrations in early high-risk breast cancer

Background — Aim: HER2 and TOP2A gene status are of prognostic/predictive relevance and are broadly determined with fluorescent in situ hybridization (FISH). For this purpose, probes against (A) 17p11.1-q11.1 (>5Mb, CENtromere), (B) 17q12 (600Kb, including HER2 among other genes), and (C) 17q21-22 (500Kb, similarly including TOP2A) are used, and gene pathology is determined by B/A and C/A signal ratios according to consensus cut-offs. However, chromosome 17 (chr17) and 17q pathology is complex and may be misinterpreted by this approach, especially in cases with low copy gains. Herein, we profiled A, B and C signals in 1027 adjuvantly treated, early breast cancer patients. Methods: Maximal A, B, C copy numbers (raw FISH data) were submitted to hierarchical clustering as continuous variables. Clusters were compared with clinicopathological parameters, conventional FISH ratios, and were evaluated for their prognostic significance (overall survival, OS) with respect to tumor Ki67 status (13% cut-off), TopoIIa protein (5% cut-off) and ER/PgR positivity (1% cut-off). Results: Clustered values resulted into 832 (81%) putatively chr17 stable tumors (ABC low ) and 195 17q unstable tumors. Out of these, 43 (4%) had possible chr17 gain (ABC high ), 30 (3%) HER2/TOP2A gain (BC high ), and 122 (12%) HER2 gain (B high ). Unstable 17q was more frequent in Ki67 high, ER/PgR negative, high-grade tumors, and were almost exclusively present in luminal-HER2 and HER2-enriched tumors (all Pearson9s p9s high , BC high and B high with the respective classical CEN17, HER2/TOP2A and HER2 gains was overall high , BC high and B high copy profiles with Ki67 status (Wald9s p = 0.007) and TopoIIa immunopositivity (Wald9s p = 0.018) were observed. Among patients with low Ki67 tumors, 5-year survival rates were 95% B high , 90% ABC low , 88% ABC high and 63% BC high (BC high vs. ABC low , HR=3.65, 95%CI 1.32–10.09, p = 0.013), while among patients with high Ki67, no such difference was observed. Similarly, patients with TopoIIa negative - BC high tumors were associated with shorter OS (HR = 2.63, 95%CI 1.06–6.48, p = 0.036) compared to ABC low tumors. By contrast, the interaction observed between copy profiles and ER/PgR status (Wald9s p = 0.011) was in the opposite direction. Among patients with ER/PgR negative tumors, those with BC high and ABC high tumors survived longer than those with ABC low tumors (HR = 0.29, 95%CI 0.12–0.73, p = 0.008); no such difference was observed in ER/PgR positive tumors. Conclusions: Profiles of 17p11-q11, 17q12 and 17q21-22 copies according to broadly used FISH probes reveal distinct subgroups of breast cancer with possibly different versions of chromosomal instability that are only partially concordant with classical CEN17, HER2 and TOP2A gain status. The prognostic significance of these profiles seems to depend on tumor proliferation and hormonal status, as well as on the extent of the affected areas on 17q. Validation of these profiles for prospective application may aid in a more accurate assessment of early high-risk breast cancer patients. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P1-07-15.