Abstract 874: Autoantibody biomarker discovery in basal-like breast cancer using nucleic acid programmable protein array

Abstract

Abstract Background: Basal-like breast cancer (BLBC) is a rare aggressive subtype of breast cancer (∼10-20% of all BC) with high recurrence rates and poor prognosis. BLBC disproportionately affects African American and younger women, has a high growth rate and occurs at young ages. These factors render mammographic screening challenging, necessitating the development of new methods for BLBC's early detection. Proteins released from small early tumors may only be present intermittently and get diluted to tiny concentrations in the blood, making them difficult to use as biomarkers. However, they can induce an autoantibody (AAb) response, which can amplify the signal and persist in the blood even if the antigen is gone. Accordingly, we assessed whether circulating AAbs could act as early detection biomarkers of BLBC. We have developed a sero-proteomic technology, called Nucleic Acid Programmable Protein Arrays (NAPPA), which tests the immune responses to thousands of proteins simultaneously. Results: Plasma samples were from the Polish Breast Cancer Study. BLBC patients were classified by immunohistochemical staining of ER-, PR-, HER2-, and either EGFR+ or CK5/6+. Case and control plasma were probed against ∼10,000 full-length human proteins on NAPPA to discover informative antigens, which were selected based on the following statistical metrics: 1) sensitivity at 95% specificity; 2) area under receiver operating characteristic curve (AUC); 3) partial AUC above 95% specificity; 4) Welch's t test. We then identified candidate AAbs in a training study combining the 800 top antigens on a single array. These arrays were screened with an independent training set of plasma from 50 each of BLBC cases and healthy controls, and 30 other BC cases. Preliminary analyses suggest 5 candidate biomarkers with sensitivities ranging from 26-32% at 94% specificity. The antigen list will be refined using AUC, and the sensitivity and partial AUC at 95%, 90% and 80% specificity levels. Permutation-based false discovery rates will be used to assess statistical significance, and the antigens will be combined into a biomarker panel using a random forests algorithm. Finally, we will test the individual markers and the antigen panel in a blinded independent validation set of 50 cases, 50 controls and 30 other BC cases. These tests will use cutoffs and algorithm determined entirely from the training data. We are assessing whether the candidate biomarkers are specific for cancer vs. controls and BLBC vs. other molecular subtypes of breast cancer. Summary: This describes the first immunoproteomic screen for AAb biomarkers of BLBC. Using protein microarrays displaying more than 10,000 folded unique human proteins, controls, BLBC and non-BLBC cases were tested in independent discovery and training stages identifying AAbs with potential diagnostic value. Blinded validation studies are in progress. Citation Format: Jie Wang, Jonine D. Figueroa, Garrick Wallstrom, Joshua Sampson, Eliseo Mendoza Garcia, Jason Steel, Jin Park, Karen S. Anderson, Louise Brinton, Montserrat Garcia-Closas, Jolanta Lissowska, Mark E. Sherman, Ji Qiu, Joshua LaBaer. Autoantibody biomarker discovery in basal-like breast cancer using nucleic acid programmable protein array. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 874. doi:10.1158/1538-7445.AM2014-874