Abstract 1486: Fatty acid and phospholipid profiling of liver tumor tissue: correlation with in vivo molecular PET imaging of phosphocholine synthesis

Abstract

Abstract Alterations in phospholipid and fatty acid metabolism may play key roles in hepatocarcinogenesis. Phosphocholine synthesis can be imaged and quantified in-vivo by positron emission tomography (PET) using radiopharmaceuticals bearing the choline moiety. Results from this PET assessment of tissue phospholipid metabolism were compared to phospholipid and fatty acid profiles of both tumor and liver parenchymal tissues in patients with liver cancer undergoing hepatectomy. Fluorine-18 fluorocholine PET/CT of the liver was completed <2 weeks prior to surgery as part of a clinicaltrials.gov-registered (NCT01395030) IRB-approved clinical study with written informed consent. Paired tumor and parenchymal tissue samples were collected in liquid nitrogen (20 hepatocellular carcinoma, 3 with intrahepatic cholangiocarcinoma), freeze-dried, homogenized and extracted with methanol/CHCl3/water containing BHT as preservative. The organic phase was analyzed by orbitrap LCMS after electrospray ionization in positive mode to quantify fatty acids (after derivatization) and phospholipid metabolites. Tumor and liver standardized uptake value (SUV) measurements were obtained from static PET images. Tumors exhibited significantly lower levels of docosahexaenoic acid, docasopentanoic acid, alpha-linolenic acid and significantly higher levels of palmitoleic acid, eicosadienoic acid, arachadonic acid, and linoleic acid than peritumoral liver tissue. Tumors also exhibited significantly higher SUVs than surrounding liver. Highest tumor SUV correlated significantly with tumor levels of palmitoleic acid, gamma linolenic acid, CDP choline, phosphatidylcholine, phosphatidylethanolamine and phosphatidylinsolitol. In contrast, no significant correlations were noted between tumor SUV and peri-tumoral levels of these compounds. Peritumoral liver SUV was significantly correlated with levels of myristic acid, eicosaptaenoic acid, gamma-linolenic acid, palmitoleic acid, docosapentaenoic acid, linoleic acid, oleic acid, and palmitic acid in peritumoral liver tissue. Furthermore, parenchymal liver SUV was inversely correlated with the Knodell histologic activity index (HAI), as well as histologic features of portal inflammation, piecemeal necrosis, and fibrosis. These histology features were also inversely correlated with liver tissue concentrations of docosahexaenoic acid, CDP-choline, total phosphaticylcholine, and total phosphatidylinositol. No significant correlations were found between tumor fatty acid/phospholipid levels and Edmonson-Steiner tumor grade. These preliminary results may indicate multiple inter-related metabolic alterations involving phospholipid and fatty acid metabolism in primary liver cancer. Chronic liver disease appears to be associated with diminishing hepatic phospholipid reserve. Note: This abstract was not presented at the meeting. Citation Format: Sandi Alexander Kwee, Adrian A. Franke, Laurie J. Custer, Xingnan Li, Linda L. Wong. Fatty acid and phospholipid profiling of liver tumor tissue: correlation with in vivo molecular PET imaging of phosphocholine synthesis. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1486. doi:10.1158/1538-7445.AM2015-1486