Discipline: Nanoscience
Subcategory: Biomedical Engineering
Session: 1
Paula Angarita Rivera - Indiana University-Purdue University of Indianapolis and Marian University
Co-Author(s): Mark Woollam, Indiana University-Purdue University Indianapolis, Indianapolis, IN; Meghana Teli, Indiana University-Purdue University Indianapolis, Indianapolis, IN ; Ali Daneshkhah, Indiana University-Purdue University Indianapolis, Indianapolis, IN ; Amanda Siegel Mangilal Agarwal, Indiana University-Purdue University Indianapolis, Indianapolis, IN
One in eight women will likely test positive for breast cancer in their lifetime. For this reason, it is important to diagnose potential patients before cancer metastasizes. Screening test are the primary method used by physicians to detect breast cancer in patients before a biopsy is performed. However, from the biopsy only 20% of those patients have breast cancer. The current screening method is not very efficient and there is no other method that accurately determines if a patient has breast cancer or not. Volatile organic compounds (VOCs) are metabolites that can serve as biomarkers for different types of diseases. An effective method to analyze these VOCs is using gas chromatography-mass spectrometry. There are two common types of mass analyzers which differentiate ions by their mass to charge ratio (m/z): single-quadrupole (GC-MS) and quadrupole time-of-flight (GC-QTOF). The goal is to demonstrate that these mass analyzers can identify the same statistically significant biomarkers that can distinguish between localized and delocalized breast cancer. A total of 18 mouse urine samples were analyzed by using both instruments. eight urine samples were collected from mice with cancer in the mammary pad and ten from mice with cancer metastasized to the bone. Multivariate analysis such as Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA) and Leave One Out Cross Validation (LOOCV) were used to identify the statistically significant biomarkers present in breast cancer. The results demonstrate that a total of 6 biomarkers are present in both data sets (95% accuracy). Furthermore, these biomarkers, if confirmed in human studies, could be used to design a resistive based sensor to detected breast cancer non-invasively in human urine.
Funder Acknowledgement(s): Thank you to my mentors and INDI for this opportunity. I would like to thank National Science Foundation for their support through the Research Experience for Undergraduate Students (REU) program (Award # 1659688).
Faculty Advisor: Dr. Mangilal Agarwal, agarwal@iupui.edu
Role: I ran the samples in the GC-QTOF for this study as well I did the statistical analysis for both sets of data compare in this research (GCMS and GC-QTOF). I was responsible for all the results of the project.