Abstract
We aimed to evaluate the accuracy of ultrasonography with gynecologic examination performed by a gynecological oncologist and magnetic resonance imaging (MRI) interpreted by a radiologist for the local and regional staging of patients with early-stage cervical cancer. The study was a single-site sub-analysis of the multi-institutional prospective, observational Total Mesometrial Resection (TMMR) Register Study, which included all consecutive study patients from Gdynia Oncology Center. Imaging results were compared with pathology findings. A total of 58 consecutive patients were enrolled, and 50 underwent both ultrasonography and MRI. The accuracy of tumor detection and measurement errors was comparable across ultrasonography and MRI. There were no significant differences between ultrasonography and MRI in the accuracy of detecting parametrial involvement (92%, confidence interval (CI) 84-100% vs. 76%, CI 64-88%, p = 0.3), uterine corpus infiltration (94%, CI 87-100% vs. 86%, CI 76-96%, p = 0.3), and vaginal fornix involvement (96%, CI 91-100% vs. 76%, CI 64-88%, p = 0.3). The importance of uterine corpus involvement for the first-line lymph node metastases was presented in few cases. The accuracy of ultrasonography was higher than MRI for correctly predicting tumor stage: International Federation of Gynecology and Obstetrics (FIGO)-2018: 69%, CI 57-81% vs. 42%, CI 28-56%, p = 0.002, T (from TNM system): 79%, CI 69-90% vs. 52%, CI 38-66%, p = 0.0005, and ontogenetic tumor staging: 88%, CI 80-96% vs. 70%, CI 57-83%, p = 0.005. For patients with cervical cancer who are eligible for TMMR and therapeutic lymphadenectomy, the accuracy of ultrasonography performed by gynecological oncologists is not inferior to that of MRI interpreted by a radiologist for assessing specific local parameters, and is more accurate for local staging of the disease and is thus more clinically useful for planning adequate surgical treatment.