Premium
Ovarian Cancer Detection Using Targeted Plasma Metabolic Profiling
Author(s) -
Turner Cassidy,
Jasbi Paniz,
Shi Xiaojian,
Atlas Sarah,
Wei Yiping,
Gu Haiwei
Publication year - 2020
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.2020.34.s1.05499
Subject(s) - receiver operating characteristic , metabolomics , ovarian cancer , medicine , logistic regression , area under the curve , oncology , cancer , bioinformatics , biology
Ovarian cancer (OC) is the second most common form of gynecological cancer and is the most fatal among all forms of gynecological malignancies. Despite the important role of metabolic processes in the molecular pathogenesis of OC, robust metabolic markers to enable effective screening, rapid diagnosis, accurate surveillance, and therapeutic monitoring of OC are still lacking. In this study, we present a targeted liquid chromatography‐tandem mass spectrometry (LC‐MS/MS)‐based metabolic profiling approach for the identification of metabolic biomarker candidates that could enable expedited, highly sensitive and specific OC detection. Using this targeted approach, 90 plasma metabolites from many metabolic pathways of potential biological significance were reliably detected and monitored in 218 samples taken from the three groups of subjects (78 OC patients, 50 benign samples, and 90 healthy controls). Multivariate significance testing and receiver operating characteristic (ROC) analysis revealed 8 significant metabolites (p<0.05) with high magnitudes of fold change (FC>2) and high predictive accuracy [area under the curve (AUC)>0.90] for distinguishing healthy controls from OC patients. For discrimination of OC patients from benign samples, the results of our model refinement pipeline informed the construction of a logistic regression classifier defined by a 5‐metbaolite panel of potential plasma biomarkers, exhibiting roughly 75% predictive accuracy using a 50% random‐split training and testing paradigm for external validation. Multivariate ROC analysis showed enhanced classification performance for identification of OC patients from benign samples, relative to individual metabolites. Pathway analysis revealed significant disturbances in glycine, serine, and threonine metabolism; glyoxylate and dioxylate metabolism; the pentose phosphate pathway; and histidine metabolism. The results expand basic knowledge of the metabolome related to OC pathogenesis relative to both healthy and benign samples, revealing potential pathways or markers that can be targeted therapeutically. This study also provides a promising basis for the development of larger multi‐site projects to validate our findings across population groups and further advances the development of improved clinical care for OC patients.