UTILITY OF PREOPERATIVE DYNAMIC PELVIC FLOOR MAGNETIC RESONANCE IMAGING (MRI) FOR PELVIC ORGAN PROLAPSE (POP)

M. Canter, MD, J. Verosko, MD, A. Mark, MD and C.B. Iglesia, MD

   Objective: We sought to determine whether preoperative dynamic MRI of the pelvic floor agrees with physical exam and surgical findings in patients with POP, and whether this diagnostic modality would ultimately change the proposed surgical approach.

   Materials and Methods: The study consisted of a case series of 11 patients underwent a full urogynecologic history/questionnaire and physical examination including pelvic organ prolapse quantification (POP-Q), barrier multichannel urodynamic evaluation, dynamic MRI of the pelvis, and surgical evaluation. Statistical analyses included Fisher's exact test and the McNemar test. Inclusion of 11 patients insured 80% power for detecting a difference between paired proportions of 0.6 when the proportion of discordant pairs was expected to be 0.65, based on a two-sided McNemar's test with a 0.05
significance level (nQuery Advisor, Version 2.0).

   Results: Mean age was 56.9 years (range 39-69) and mean parity 2.5 (range 1-7). Seven (63.6%) were Caucasian and 4 (36.3%) were black. Ten patients (91%) had prior hysterectomy; 5 (45.5%) had prior surgery for prolapse; and 4 (36.4%) had prior incontinence surgery. Symptoms included urge incontinence in 63.6% and stress incontinence in 45.5%. Nine patients (81.8%) had stage III prolapse, one (9%) had Stage IIA and 1 had Stage IV prolapse. There was perfect agreement between MRI and operative findings with respect to cystoceles (p = 0.006) and perineal descent (p = 0.018). MRI missed 7 of the 9 enteroceles found at surgery (enterocele 18.2% by MRI versus 81.8% by surgery, p = 0.016), but did not incorrectly identify any enterocele where none were found at surgery. Five of the enteroceles missed by MRI were suspected during physical exam. Only peritoneoceles were shown on MRI in 5 of the 7 missed enteroceles. MRI also missed 4 of the 7 rectoceles found at surgery. In 2/11 (18.2%) patients, MRI findings of a peritoneocele unsuspected on physical exam changed the proposed surgical approach from a completely transvaginal to a laparoscopic repair. One of these two patients was thought to have a sole cystocele but was found to have an anterior enterocele and vault prolapse on laparoscopy. The other patient was presumed to have a large rectocele and was found to have a sliding omentocele beneath the rectovaginal septum, in addition to the rectocele, on laparoscopy.

   Conclusions: Substantial discrepancies were found between MRI and operative prolapse findings in patients with advanced pelvic organ prolapse. Reasons for MRI underestimation compared to surgical findings include limitations related to supine straining as opposed to upright straining position, and the presence of significant small bowel adhesions. However, when MRI indicates an unexpected peritoneocele, the laparoscopic approach can be helpful.

   Key words: pelvic organ prolapse, pelvic floor imaging, dynamic MRI