Gang He1, Christina A Holcroft2, Marie-Claude Beauchamp3, Amber Yasmeen3, Alex Ferenczy4, Jennifer Kendall-Dupont5, Anne-Marie Mes-Masson6, Diane Provencher7, Walter H Gotlieb8. 1. Department of Diagnostic Medicine, Jewish General Hospital, McGill University, Montreal QC. 2. Centre for Clinical Epidemiology, Jewish General Hospital, McGill University, Montreal QC. 3. Lady Davis Institute for Medical Research, Montreal QC. 4. Department of Pathology, Jewish General Hospital, McGill University, Montreal QC. 5. Centre de recherche du Centre hospitalier de l'Université de Montréal/Institut du Cancer de Montréal, Université de Montréal, Montreal QC. 6. Centre de recherche du Centre hospitalier de l'Université de Montréal/Institut du Cancer de Montréal, Université de Montréal, Montreal QC; Département de médicine, Université de Montréal, Montreal QC. 7. Centre de recherche du Centre hospitalier de l'Université de Montréal/Institut du Cancer de Montréal, Université de Montréal, Montreal QC; Division de gynécologie oncologique, Université de Montréal, Montreal QC. 8. Lady Davis Institute for Medical Research, Montreal QC; Division of Gynecologic Oncology, Jewish General Hospital, McGill University, Montreal QC.
Abstract
OBJECTIVE: To investigate biomarkers and clinical parameters to distinguish ovarian cancers from benign ovarian tumours. METHODS: Serum biomarkers (CA 125, human epididymis protein 4 [HE 4], interleukin-18 [IL-18], leptin, macrophage migration inhibitory factor [MIF], fibroblast growth factor 2 [FGF-2], insulin-like growth factor, osteopontin, prolactin) and the risk of malignancy indexes I and II (RMI-I and RMI-II) scores were obtained prior to surgery in 52 patients with ovarian tumours (37 malignant and 15 benign). ROC curves were built for each individual marker, for logistic regression models using all markers, and for models combining both biomarkers and RMI scores. RESULTS: The model with nine biomarkers performed well (specificity 93%, sensitivity 84%) and was more reliable than the RMI-I or RMI-II alone. A regression model combining RMI-II and six of the biomarkers (CA 125, HE 4, IL-18, leptin, MIF, and FGF-2) allowed differentiation between the cancer and non-cancer cases in this pilot study. CONCLUSION: The regression models using biomarkers combined with clinical scoring systems warrant further investigation to improve triage of patients with ovarian tumours to enhance utilization of resources and optimize patient care.
OBJECTIVE: To investigate biomarkers and clinical parameters to distinguish ovarian cancers from benign ovarian tumours. METHODS: Serum biomarkers (CA 125, human epididymis protein 4 [HE 4], interleukin-18 [IL-18], leptin, macrophage migration inhibitory factor [MIF], fibroblast growth factor 2 [FGF-2], insulin-like growth factor, osteopontin, prolactin) and the risk of malignancy indexes I and II (RMI-I and RMI-II) scores were obtained prior to surgery in 52 patients with ovarian tumours (37 malignant and 15 benign). ROC curves were built for each individual marker, for logistic regression models using all markers, and for models combining both biomarkers and RMI scores. RESULTS: The model with nine biomarkers performed well (specificity 93%, sensitivity 84%) and was more reliable than the RMI-I or RMI-II alone. A regression model combining RMI-II and six of the biomarkers (CA 125, HE 4, IL-18, leptin, MIF, and FGF-2) allowed differentiation between the cancer and non-cancer cases in this pilot study. CONCLUSION: The regression models using biomarkers combined with clinical scoring systems warrant further investigation to improve triage of patients with ovarian tumours to enhance utilization of resources and optimize patient care.
Authors: Daphne Gschwantler-Kaulich; Sigrid Weingartshofer; Christine Rappaport-Fürhauser; Robert Zeillinger; Dietmar Pils; Daniela Muhr; Elena I Braicu; Marie-Therese Kastner; Yen Y Tan; Lorenz Semmler; Jalid Sehouli; Christian F Singer Journal: PLoS One Date: 2017-12-15 Impact factor: 3.240