Objective: The purpose of this study was to compare the accuracy of fast spin-echo MR imaging that uses a transrectal surface coil with the accuracy of fast spin-echo MR imaging that uses a pelvic phased-array coil for preoperative local staging of uterine cervical carcinoma.
Materials and methods: Sixty-two patients with pathologically confirmed invasive cervical carcinoma who underwent MR imaging using both a pelvic phased-array coil and a transrectal surface coil were retrospectively evaluated. Both sets of images were reviewed in random order by three observers who then reached a consensus opinion at a second session. Among the 28 patients who underwent radical hysterectomy, findings from MR imaging and pathologic staging were compared. In the remaining nonsurgical cases, findings from MR imaging using a transrectal surface coil and MR imaging using a pelvic phased-array coil were compared with findings from pelvic examinations under general anesthesia.
Results: MR imaging using a transrectal surface coil revealed 50 (88%) of the 57 tumors found at surgery, and MR imaging using a pelvic phased-array coil revealed 44 (77%) of the 57 tumors, resulting in a statistically significant (p < .05) difference. In patients who underwent surgery (n = 28), the accuracy of the overall staging and the parametrial evaluation were not significantly different for MR imaging using a transrectal surface coil and MR imaging using a pelvic phased-array coil (89% and 96%, respectively, on MR imaging using a transrectal surface coil; 89% and 93%, respectively, on MR imaging using a pelvic phased-array coil). For the patients who did not undergo surgery (n = 34), MR imaging using a transrectal surface coil and MR imaging using a pelvic phased-array coil corresponded in 82% for overall staging and in 92% for parametrial invasion.
Conclusion: The use of a transrectal surface coil improved detection of small tumors of the uterine cervix but did not significantly improve the accuracy of overall staging or the accuracy of assessment of parametrial invasion.