Ovarian cancer (OC) is the 5th leading cause of cancer-related deaths for U.S. women and the deadliest gynecological disease. Lack of symptoms in addition to the deficiency of highly specific biomarkers for detection typically result in only 25% of OC cases being diagnosed at FIGO stage I. High-grade serous carcinoma (HGSC) is the most prevalent form of OC, but three rarer histological subtypes also exist? endometrioid, clear cell, and mucinous. An effective screening strategy for early diagnosis would be particularly advantageous since 5-year OC survival rates can be as high as 90%. Unfortunately, protein biomarkers such as CA-125 do not have sufficient positive predictive value to be useful from a clinical perspective. We hypothesize that useful information regarding early stage HGSC and other ovarian cancers can be found in the serum metabolome. Our pilot studies in both humans and OC models, such as the double-knockout Dicer-Pten mouse recently developed by our team members, show great promise in this regard? average sensitivity and specificity for early detection have reached 97.8% and 99.0% in banked human serum samples, and up to100% in mice. These results have prompted us to perform a much deeper investigation of metabolome alterations associated with early stage ovarian cancers in larger serum sample sets, and over time. We will perform metabolomics experiments in mice and banked de-identified human serum samples with much higher coverage than before by ?data fusing? various modes of ultraperformance liquid chromatography-mass spectrometry (UPLC-MS) and nuclear magnetic resonance (NMR), coupled with pathway-centric data analysis. We also propose supplementing serum-level metabolomics experiments with deep-coverage tissue mass spectrometry imaging (MSI) in both 2-D and 3-D, using a combination of matrix-assisted laser desorption/ionization (MALDI) and desorption electrospray ionization (DESI), which have complementary ionization mechanisms. Furthermore, we propose to depart from the commonly used approach of tentatively identifying spectral features by only using accurate masses, and implement a ?deep metabolite annotation? approach that uses both ?fused? high-resolution techniques (high field Orbitrap MS, MS/MS, 2-D NMR) and a new technology based on collisional cross section predictions for both travelling wave and drift tube ion mobility-MS.

Public Health Relevance

High-grade serous ovarian cancer is the most common and deadliest type of ovarian cancer; it is diagnosed mostly at an advanced stage at which the cancer has already spread beyond the ovary or the fallopian tube, to the abdominal cavity. This advanced-stage diagnosis inevitably causes high mortality. In this project, we propose to study molecular changes associated with early disease both in mouse models that closely mimic human disease and in banked de-identified human serum samples, with the aim of ultimately designing a robust diagnostic panel.

National Institute of Health (NIH)
National Cancer Institute (NCI)
Research Project (R01)
Project #
Application #
Study Section
Enabling Bioanalytical and Imaging Technologies Study Section (EBIT)
Program Officer
Knowlton, John R
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Georgia Institute of Technology
Schools of Arts and Sciences
United States
Zip Code