A comparison between dynamic pelvic magnetic resonance imaging and videoproctography in patients with constipation

Abstract

This study attempts to compare the diagnostic efficacy of dynamic pelvic magnetic resonance imaging with that of videoproctography for the presence of rectocele, sigmoidocele, and intussusception as well as the measurement of anorectal angle and perineal descent in constipated patients. Patients volunteering for the study and fulfilling the criteria for videoproctography to evaluate constipation were also scheduled for dynamic pelvic magnetic resonance imaging. Patients undergoing videoproctography were placed in the left lateral decubitus position, after which 50 ml of liquid barium paste was introduced into the rectum. After this, approximately 100 ml of thick barium paste similar to stool in consistency was injected into the rectum, and the patient was instructed to defecate while video images were taken. For dynamic pelvic magnetic resonance imaging, air, to be used as contrast, was allowed to accumulate in the rectum via examination with the patient in the prone position. A capsule was taped to the perineal skin immediately posterior to the anal orifice for marking. Sagittal and axial T1 images were obtained through the pelvis at 8-mm intervals with dynamic breathhold sagittal images of the anorectal region obtained at rest and during strain and squeeze maneuvers. Total acquisition time per maneuver was approximately 19 seconds. The tests were performed by different examiners blinded to the result of the other evaluation. The investigations were independently interpreted, findings compared, and patients questioned regarding their impression of dynamic pelvic magnetic resonance imaging and videoproctography. From June 1996 to April 1997, 22 patients (15 females) with a mean age of 68 (range, 21-85) years underwent both videoproctography and dynamic pelvic magnetic resonance imaging. Dynamic pelvic magnetic resonance imaging was only able to detect 1 of 12 (8.3 percent) anterior rectoceles and one of two (50 percent) posterior rectoceles identified by videoproctography. It failed to recognize any of the rectoanal intussusception (zero of four) but did show 9 of 12 (75 percent) sigmoidoceles. Significant discrepancy of measurement of the anorectal angle and perineal descent exists between the two studies, and dynamic pelvic magnetic resonance imaging was not able to detect any (0 of 11) of the patients with increased fixed perineal descent and only half (one of two) of the patients with increased dynamic perineal descent noted on videoproctography. All 22 patients preferred dynamic pelvic magnetic resonance imaging over videoproctography because of greater comfort. Occasionally, the increased cost of new technology can be justified by the enhanced diagnostic yield. The ability to avoid unnecessary surgery or, conversely, to continue to search for otherwise occult pathology that can be surgically corrected justifies routine application of these new tools. However, this study has shown that, despite a cost of approximately ten times more for dynamic pelvic magnetic resonance imaging than for videoproctography, no clinical changes were made. Thus, on the basis of this study, we cannot endorse the routine application of dynamic pelvic magnetic resonance imaging for the evaluation of constipated patients. In certain selected individuals, it may play a role, but further study is necessary to clarify its exact role.