BACKGROUND: Discrimination between epithelial ovarian cancer (EOC) and benign ovarian tumor (BOT) has always been difficult in clinical practice. We investigated the application of metabolomics in distinguishing EOC and BOT and tried to discover valuable biomarkers. METHODS: Plasma metabolomic profiling was performed using ultra-performance liquid chromatography mass spectrometry (UPLC/MS). Partial least-squares discriminant analysis was employed to classify EOC and BOT, and reveal their metabolic differences. The area under the receiver-operating characteristic curve (AUC) was utilized to evaluate the predictive performance of the metabolic profiles for external validation set. RESULTS: The metabolomic profiles consisting of 535 metabolites revealed a clear separation between EOC and BOT, with AUC of 0.86 for the external validation set. 6 metabolic biomarkers were identified, and the plasma concentrations of the 4 ascertained biomarkers (L-tryptophan, LysoPC(18:3), LysoPC(14:0), and 2-Piperidinone) were lower in EOC patients than those in BOT patients. Among them, tryptophan and LysoPC have been suspected to participate in cancer progression, and 2-Piperidinone might be a novel biomarker for EOC. CONCLUSIONS: Metabolomics could be used to discriminate EOC from BOT in clinical practice, and the identified metabolic biomarkers might be important on investigating the biological mechanisms of EOC.
BACKGROUND: Discrimination between epithelial ovarian cancer (EOC) and benign ovarian tumor (BOT) has always been difficult in clinical practice. We investigated the application of metabolomics in distinguishing EOC and BOT and tried to discover valuable biomarkers. METHODS: Plasma metabolomic profiling was performed using ultra-performance liquid chromatography mass spectrometry (UPLC/MS). Partial least-squares discriminant analysis was employed to classify EOC and BOT, and reveal their metabolic differences. The area under the receiver-operating characteristic curve (AUC) was utilized to evaluate the predictive performance of the metabolic profiles for external validation set. RESULTS: The metabolomic profiles consisting of 535 metabolites revealed a clear separation between EOC and BOT, with AUC of 0.86 for the external validation set. 6 metabolic biomarkers were identified, and the plasma concentrations of the 4 ascertained biomarkers (L-tryptophan, LysoPC(18:3), LysoPC(14:0), and 2-Piperidinone) were lower in EOC patients than those in BOT patients. Among them, tryptophan and LysoPC have been suspected to participate in cancer progression, and 2-Piperidinone might be a novel biomarker for EOC. CONCLUSIONS: Metabolomics could be used to discriminate EOC from BOT in clinical practice, and the identified metabolic biomarkers might be important on investigating the biological mechanisms of EOC.
Authors: David B Liesenfeld; Nina Habermann; Robert W Owen; Augustin Scalbert; Cornelia M Ulrich Journal: Cancer Epidemiol Biomarkers Prev Date: 2013-10-04 Impact factor: 4.254
Authors: Matthew F Buas; Haiwei Gu; Danijel Djukovic; Jiangjiang Zhu; Charles W Drescher; Nicole Urban; Daniel Raftery; Christopher I Li Journal: Gynecol Oncol Date: 2015-10-30 Impact factor: 5.482
Authors: Silvia Darb-Esfahani; Elena Ioana Braicu; Mika Hilvo; Ines de Santiago; Peddinti Gopalacharyulu; Wolfgang D Schmitt; Jan Budczies; Marc Kuhberg; Manfred Dietel; Tero Aittokallio; Florian Markowetz; Carsten Denkert; Jalid Sehouli; Christian Frezza Journal: Cancer Res Date: 2015-12-18 Impact factor: 12.701