Prospective analysis of circulating metabolites and endometrial cancer risk
Laure Dossus, Eirini Kouloura, Carine Biessy, Vivian Viallon, Alexandros P. Siskos, Niki Dimou, Sabina Rinaldi, Melissa A. Merritt, Naomi E. Allen, Renée T. Fortner, Rudolf Kaaks, Elisabete Weiderpass, Inger T. Gram, Joseph A. Rothwell, Lucie Lécuyer, Gianluca Severi, Matthias B. Schulze, Therese Haugdahl Nøst, Marta Crous‐Bou, Maria-Jose Sánchez, Pilar Amiano, Sandra M. Colorado‐Yohar, Aurelio Barricarte Gurrea, Julie A. Schmidt, Domenico Palli, Claudia Agnoli, Rosario Tumino, Carlotta Sacerdote, Amalia Mattiello, Roel Vermeulen, Alicia K. Heath, Sofia Christakoudi, Konstantinos K. Tsilidis, Ruth C. Travis, Marc J. Gunter, Hector C. Keun
Abstract
BackgroundEndometrial cancer is strongly associated with obesity and dysregulation of metabolic factors such as estrogen and insulin signaling are causal risk factors for this malignancy. To identify additional novel metabolic pathways associated with endometrial cancer we performed metabolomic analyses on pre-diagnostic plasma samples from 853 case-control pairs from the European Prospective Investigation into Cancer and Nutrition (EPIC).MethodsA total of 129 metabolites (acylcarnitines, amino acids, biogenic amines, glycerophospholipids, hexoses, and sphingolipids) were measured by liquid chromatography-mass spectrometry. Conditional logistic regression estimated the associations of metabolites with endometrial cancer risk. An analysis focusing on clusters of metabolites using the bootstrap lasso method was also employed.ResultsAfter adjustment for body mass index, sphingomyelin [SM] C18:0 was positively (OR1SD: 1.18, 95% CI: 1.05–1.33), and glycine, serine, and free carnitine (C0) were inversely (OR1SD: 0.89, 95% CI: 0.80–0.99; OR1SD: 0.89, 95% CI: 0.79–1.00 and OR1SD: 0.91, 95% CI: 0.81–1.00, respectively) associated with endometrial cancer risk. Serine, C0 and two sphingomyelins were selected by the lasso method in >90% of the bootstrap samples. The ratio of esterified to free carnitine (OR1SD: 1.14, 95% CI: 1.02–1.28) and that of short chain to free acylcarnitines (OR1SD: 1.12, 95% CI: 1.00–1.25) were positively associated with endometrial cancer risk. Further adjustment for C-peptide or other endometrial cancer risk factors only minimally altered the results.ConclusionThese findings suggest that variation in levels of glycine, serine, SM C18:0 and free carnitine may represent specific pathways linked to endometrial cancer development. If causal, these pathways may offer novel targets for endometrial cancer prevention.