Background: Even after successful treatment of primary breast tumors, there is a continued risk of recurrence. The risk varies between subtypes and there are ongoing efforts that aim to improve prediction of such risks for individual patients. Detection of subclinical metastases might be achieved by biomarkers in blood. In this study, we profiled protein expression in blood plasma from patients with known clinical outcome (recurrence vs no recurrence) to identify prognostic markers of breast cancer recurrence.

Methods: The subjects and specimens were made available through the Clinical Breast Care Project using IRB-approved protocols. We analyzed blood plasma samples taken at the time of diagnosis from consented patients who subsequently relapsed (33 cases) as well as those with no disease recurrence (31 controls). Based on hormone receptor and lymph node status the samples were grouped as: ER-/HER2- (17 cases/15 controls), ER+/LN+ (10/10) and ER+/LN- (6/6). We used aptamer-based SOMAscan assay platform to study the expression of 1252 proteins. We analyzed the protein expression data by using their coding genes in order to apply the Gene Set Enrichment Analysis method (GSEA v.2, Broad Institute). Pathway databases of KEGG, REACTOME, BIOCARTA and C4 collection were used. Significant gene sets were called at 5% FDR, and overlaps and low coverage gene sets (Tags <70%) were removed. Statistical analysis and clustering were done using R.

Results: Unsupervised clustering showed some difference in signal in the ER+/LN- group. Even though there was a lack of significantly differentiated proteins between the cases and controls of this group, many significant gene sets were identified. After applying the cutoff filters and removing the overlaps, there were 5 gene sets enriched with the pathway collection, involved in B-cell receptor signaling, mRNA metabolism, tight junction and SCF-KIT signaling. Similarly, 9 gene sets from the MORF compendium were differentially expressed with the C4 collection and included neighborhood genes of NME2, ACTG1, EIF3S2, AP2M1, DAP3, UBE2I, NPM1, AATF and NPM1. In contrast, neither differentially expressed proteins nor gene sets were identified from the ER+/LN+ and ER-/HER2- groups. Since the sample size of the ER+/LN- group was small, we conducted a similar analysis by randomly choosing 6 case and control samples in the other two groups respectively. There were still no differentially expressed proteins or gene sets identified above the specified cutoff parameters.

Conclusion: Using plasma protein expression data we identified underlying gene sets differentially expressed between ER+/LN- patients who had cancer recurrence and no recurrence. Many genes in these sets were already known biomarkers (e.g. PTEN, AKT1, STAT3, SET etc.). These results can be used for understanding patterns of recurrence in different cancer subtypes. Further research is needed to estimate the clinical significance of these gene products.

The views expressed in this article are those of the author and do not reflect the official policy of the Department of Army/Navy/Air Force, the Department of Defense, or U.S. Government.

Citation Format: Praveen Kumar A, Kovatich AJ, Biancotto A, Cheung F, Davidson-Moncada JK, Kvecher L, Liu J, Ru Y, Kovatich AW, Deyarmin B, Fantacone-Campbell JL, Hooke JA, Raj Kumar PK, Rui H, Hu H, Shriver CD. Analysis of breast cancer recurrence using gene set enrichment analysis [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P4-09-14.