Multiomic profiling of ER-positive HER2-negative breast cancer reveals markers associated with metastatic spread

Metastatic disease is the main cause of breast cancer (BC)-related deaths, but prediction of metastases remains challenging especially in the large and diverse group with estrogen receptor (ER)-positive, human epidermal growth factor receptor 2 (HER2)-negative tumors. Molecular tumor features beyond currently used markers could provide important information for stratifying metastatic risk. To allow for the discovery of new subtypes and molecular tumor features associated with metastatic spread, i.e., both lymph node and distant metastases, we here leverage advances in proteomic profiling of tumors.

We developed a protocol for proteome and phosphoproteome analysis using label-free data independent acquisition (DIA) liquid chromatography tandem mass spectrometry (LC-MS/MS) and integrated the generated data with parallel transcriptome data for the profiling of 182 ER-positive, HER2-negative primary BC tumors from the SCAN-B cohort.

A total of 13,571 protein groups, 7107 phosphopeptides and 13,085 expressed genes were quantified in at least 70% of the samples. The data showed clear differences between invasive lobular carcinoma and no special type cancers, including the hallmark loss of E-cadherin expression and differences in catenin levels. We identified potential new subtypes with differential immune infiltration patterns and survival through unsupervised consensus clustering. Additionally, by adopting an integrative, multiomic data analysis workflow, we identified several potential protein markers of both lymph node and distant metastases. For lymph node metastasis, the level of phosphorylated ES8L2 serine at position 570 (multivariable p value = 0.05, HR = 0.61, 95% CI 0.38-0.99) was associated with improved recurrence-free survival, and showed decreased abundance in lymph node positive cases. For distant metastases, on the other hand, proteins belonging to the heat shock protein 90 family were associated with worse distant recurrence-free survival (multivariable p value = 0.0058, HR = 2.10, 95% CI 1.24-3.55), with significantly higher abundance levels in patients with a distant recurrence event. These correlations with survival could also be validated in multiple external cohorts.

In summary, we present the most comprehensive matched multiomic dataset from ER-positive/HER2-negative BC tumors, not only serving as an invaluable resource for further advancing precision medicine but also allowing the discovery of potential biomarkers and providing unique insights into metastatic processes.

Sweden Cancerome Analysis Network-Breast: Genomic Profiling of Breast Cancer (SCAN-B), beginning 2010-08, NCT02306096.