Abstract 2331: Metabolic profiling of pre and post treatment ovarian cancer tissues using in situ mass spectrometry imaging

Abstract Chemotherapy resistance represents a major problem in ovarian cancer prognosis and reliable indicators of chemotherapy sensitivity remain an unmet need. Here, we examined metabolic profiles of pre- and post-neoadjuvant chemotherapy (NACT) tumor samples from patients with high-grade serous ovarian cancer (HGSC). We performed spatially resolved, desorption electrospray ionization-mass spectrometry (DESI-MS) on sections from clinically annotated HGSC samples collected according to a laparoscopic triage algorithm from patients pre- and post-NACT. Patients were considered to be either excellent (ER) if they presented complete response or only microscopic disease left at interval surgery, or poor responders (PR) if they presented stable or progressive disease after NACT. Tumor tissues from 50 patients (29 ER and 21 PR, 13 ER matched pre and post NACT, 13 PR matched pre and post-NACT) were eligible for the analyses Data relative to the detected metabolites, were collected from epithelium and stroma separately. We first identified differentially upregulated metabolites in ER and PR groups and performed pathway analysis. The over-represented pathways in the chemo-naïve PR tissues included the nucleotide metabolism pathway in epithelial areas, and degradation of cysteine and homocysteine, metabolism of amino acids and derivatives, as well as sulfur amino acids metabolism in the stromal areas (p values FDR adjusted 0.045). In general, both the epithelium and stroma of PR tissues of chemo-naïve tumors, contained a higher number and a higher variety of lipid species compared with ER tissues. In post-NACT tumor epithelial areas, there was global downregulation of metabolites across all patient samples, compared to the pre-chemo specimens; this was more evident in ER patients. In post-NACT tumors from PR patients, glutathione synthesis and recycling pathway, and sulfite oxidation to sulfate pathway and many glycerolipids species were upregulated in the epithelium, while phosphatidic acids were upregulated in the stroma. Lastly, we evaluated the predictive power of the metabolic data extracted by the epithelial areas of chemo-naïve tissues, in classifying patients as ER or PR. We used 10-fold cross-validation to create a ridge regression model for the classification of ER and PR samples from pre-NACT tumor tissues. This model, based on 78 differentially expressed metabolites (small molecules and lipids), correctly classified patients as ER or PR with an accuracy of 76.2% (sensitivity 84.6% and specificity 62.5%). Our findings indicate altered metabolic pathways in ovarian cancer tissues based on response to NACT. The identification of differentially activated pathways, such as sulfate metabolism and glutathione metabolism in tumors with poor response, suggest novel actionable targets to overcome chemotherapy resistance in patients with HGSC. Citation Format: Sara Corvigno, Meredith Spraldin, Michael Keating, Sunil Badal, Igor Pereira, Elaine Stur, Nicholas Bateman, Waleed Barakat, Thomas Conrads, Sanghoon Lee, George Maxwell, Susan K. Lutgendorf, Premal Thaker, Jinsong Liu, Nicole Fleming, Katheleen Darcy, Livia S. Eberlin, Anil K. Sood. Metabolic profiling of pre and post treatment ovarian cancer tissues using in situ mass spectrometry imaging [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2331.