Comparing the efficacy of imaging techniques in detecting myometrial invasion, cervical involvement and pelvic lymph-nodal metastasis in endometrial cancer

Abstract

Aim: To compare sensitivity, specificity, positive predictive and negative predictive value of preoperative imaging techniques for detecting myometrial invasion, cervix involvement, and alsopelvic lymph nodal metastasis in endometrial cancer Materials and Methods: The medical records of patients who underwent an operation for endometrial cancer in the years between 2005 and 2017 were collected from the database at our institution. Preoperative imaging reports of 252 ultrasonography (USG), 89 computerized tomographies (CT),147 magnetic resonance imaging (MRI) of patients with endometrial cancer, and postoperative pathologic reports were collected and compared. Results: In our study 252 ultrasonography (USG), 89 computerized tomography (CT), 147 magnetic resonance imaging (MRI) examinations were evaluated. Among deep myometrial invasion; all imaging modalities have low specificities (respectively 26.7%, 37.9%, and 32.4%) but higher sensitivities (respectively 68.5%, 79.1%, and 89.4%). To rule out cervical invasion all modalities have high and comparable sensitivities (respectively 98.3%, 95.0%, and 87.0%). On the other hand, USG has superiority to detect cervical invasion over CT and MRI (respectively 71.4%, 15.4%, and 22.7%). CT has much highersensitivity than MRI for detection of pelvic lymph node metastasis (87.5% vs 53.1%). Conclusion: Preoperative imaging modalities have high sensitivities for deep myometrial invasion, but low detection rates for cervical involvement and pelvic lymph node metastasis. MRI should be the preferred modality for myometrial invasion, on the other hand, The USG is much better to detect cervical involvement. CT has superiority on other imaging modalities among lymph node metastasis.