270P - Dissecting molecular heterogeneity of luminal breast cancers using an ion mobility DIA proteomic approach

Background

Estrogen positive/HER2 negative (ER+/HER2-) early (e) breast cancers (BC) is a heterogeneous entity. Parallel Accumulation Serial Fragmentation with a data-independent acquisition (DIA-PASEF) approach enables the accurate and reproducible label-free quantification of large proteomes. Combined with bioinformatics analysis, the aim of this study was to explore if this approach could identify subgroups of ER+/HER2- eBC of different prognosis, that has never been conducted, to our knowledge, in a large homogeneous cohort of ER+/HER2- eBC.

Methods

Frozen primary tumors were collected from ER+/HER2- eBC patients treated in the ICO cancer center between 2006 and 2009 for a first occurrence of unilateral invasive carcinoma of no special type, with a surgical management first. All patients received endocrine therapy and chemotherapy in the adjuvant setting. Each sample was analyzed with a DIA approach and calculated with Spectronaut in library-free mode. Only proteins identified with 4 peptides were retained.

Results

Proteomic data from 150 patients, were analyzed. Median follow-up was 51 years. Thirty patients presented with metastatic relapse (MR). To assign samples to specific molecular profiles, unsupervised analyses were performed using consensus clustering. Forty-three samples were distributed in cluster 1 (C1; 29%); 62 in cluster 2 (C2; 41%) and 45 in cluster 3 (C3; 30%). Ninety-six, 55 and 71 % of patients were classified by immunochemistry as luminal B in C3, C2 and C1, respectively. C2 shows the best distant metastasis-free survival and C3 was enriched in SBR high tumors. The biological processes underlying cluster C3 are related to proliferation, MYC and mTORC1 pathway activation and interferon response. C2 was more characterized by hormone metabolic process with higher expression of PIP while C1 presented mesenchymal features: EMT, coagulation, fatty acid metabolism, adipogenesis, hypoxia.

Conclusions

New high throughput global proteomic is able to discriminate between ER+/HER2- eBC into subgroups with different prognoses. C1 has biological characteristics (invasion, EMT) that are not covered by current systemic treatments and may open up new therapeutic opportunities.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

6 Oncoproteomic Unit and Prot’ICO facility, Institut de Cancérologie de l’Ouest Pays de la Loire, 15 rue Boquel, 49 055 Angers, France.

Funding

Fondation de l'avenir.

Disclosure

A. Patsouris: Financial Interests, Institutional, Advisory Board: Lilly, Gilead, AstraZeneca, Novartis, Pfizer, Daiichi; Financial Interests, Institutional, Other, teaching: Gilead; Financial Interests, Institutional, Invited Speaker: NOVARTIS. V. Verriele: Financial Interests, Institutional, Advisory Board: Daiichi Sankyo. M. Campone: Financial Interests, Institutional, Advisory Board: AstraZeneca, Novartis, Sanofi, Daiichi Sankyo, Lilly, Stemline, Gilead, Seagen; Financial Interests, Institutional, Invited Speaker: Novartis, Lilly. All other authors have declared no conflicts of interest.