C Coussoou1, V Laigle-Quérat2, D Loussouarn3, E Vaucel4, E Frampas2. 1. Service de radiologie et imagerie médicale, Hôtel-Dieu, CHU de Nantes, 1, place Alexis-Ricordeau, 44093 Nantes, France. Electronic address: coussoou.cecile@gmail.com. 2. Service de radiologie et imagerie médicale, Hôtel-Dieu, CHU de Nantes, 1, place Alexis-Ricordeau, 44093 Nantes, France. 3. Service d'anatomie cytologie pathologique, Hôtel-Dieu, CHU de Nantes, 1, place Alexis-Ricordeau, 44093 Nantes, France. 4. Service de gynécologie-obstétrique, hôpital Femme-Enfant-Adolescent, CHU de Nantes, 38, boulevard Jean-Monnet, 44093 Nantes, France.
Abstract
OBJECTIVES: To determine the diagnostic accuracy of magnetic resonance imaging (MRI) for local preoperative staging in endometrial cancer in our center (Centre Hospitalier Universitaire de Nantes: CHU), since the French National Cancer Institute's surgery recommendations publication in 2010, especially for the prediction of myometrial and cervical stromal invasion. METHODS: Retrospective monocentric study of consecutive women operated of endometrial cancer in gynecology department of CHU de Nantes, who underwent preoperative pelvic MRI in our Radiology department from November 2010 to November 2016. MRI data collected from initial report and compared to surgical histological findings as gold standard. RESULTS: Sixty-four patients were included. Deep myometrial invasion was present in 35 patients in MRI versus 34 patients on postoperative histology (5 false positives, 4 false negatives). Cervical stromal invasion was present in 9 patients in MRI versus 19 patients on postoperative histology (2 false positives, 12 false negatives). The sensitivity and the specificity were respectively 88.23% (95% confidence intervals (CI) [0.71-0.96]) and 83.33% (CI [0.64-0.93]) for the deep myometrial invasion; 36.84% (CI [0.17-0.61]) and 95.55% (CI [0.83-0.99]) for the cervical stromal invasion. CONCLUSION: Our results were comparable to the literature data, with a low sensitivity for the cervical stromal invasion detection, driving us to change our MRI protocol with optional high-resolution T2 sequences perpendicular to the cervical canal if necessary.
OBJECTIVES: To determine the diagnostic accuracy of magnetic resonance imaging (MRI) for local preoperative staging in endometrial cancer in our center (Centre Hospitalier Universitaire de Nantes: CHU), since the French National Cancer Institute's surgery recommendations publication in 2010, especially for the prediction of myometrial and cervical stromal invasion. METHODS: Retrospective monocentric study of consecutive women operated of endometrial cancer in gynecology department of CHU de Nantes, who underwent preoperative pelvic MRI in our Radiology department from November 2010 to November 2016. MRI data collected from initial report and compared to surgical histological findings as gold standard. RESULTS: Sixty-four patients were included. Deep myometrial invasion was present in 35 patients in MRI versus 34 patients on postoperative histology (5 false positives, 4 false negatives). Cervical stromal invasion was present in 9 patients in MRI versus 19 patients on postoperative histology (2 false positives, 12 false negatives). The sensitivity and the specificity were respectively 88.23% (95% confidence intervals (CI) [0.71-0.96]) and 83.33% (CI [0.64-0.93]) for the deep myometrial invasion; 36.84% (CI [0.17-0.61]) and 95.55% (CI [0.83-0.99]) for the cervical stromal invasion. CONCLUSION: Our results were comparable to the literature data, with a low sensitivity for the cervical stromal invasion detection, driving us to change our MRI protocol with optional high-resolution T2 sequences perpendicular to the cervical canal if necessary.