Vaginal examination has limitations in assessing the degree and extent of recurrent prolapses. Magnetic resonance defecography (MR def) can fully investigate whether an anterior compartment prolapse is isolated or combined with apical and/or posterior compartment defects, as well as detect defecatory dysfunction or rectal intussusception which may require specific interventions.[1] We have based our decision for surgical correction of recurrent anterior compartment prolapse (RACP) on MR def findings[2] and evaluate in this report the impact of these imaging findings on our pelvic organ prolapse (POP) repair outcomes.

Following IRB approval, a retrospective review of MR def findings for anterior compartment prolapse (ACP) and patient outcome after POP repair with a minimum 6 months follow-up was performed. Patients were only included if they had a prior hysterectomy to have more uniform population. Physical exam was based on the POP -Q system that utilizes six points, Aa and Ba for anterior vagina, Ap and Bp for posterior vagina, point C and D for cervix/vault with point D not utilized in women with prior hysterectomy. All physical exams were performed by an FPMRS trained attending or fellow. VCUG was performed according to a published protocol in the lateral standing position.[3] MR def were performed in the supine position according to a previously published protocol.[2] Example MR def images are shown in Figure 1. Based on MR findings, isolated ACP was corrected vaginally with native tissue repair, while dual or triple compartment prolapses were corrected via open or robotic approaches with mesh interposition. The primary outcome was recurrence of cystocele stage 2 or greater on examination and/or need for POP reoperation.

From 2015-2020, a total of 53/90 women with MR def prior to POP repair for ACP had a previous hysterectomy and were included in the study. The other 37 did not have prior hysterectomy and were not included in the results. Patients were divided into open/robotic approach (n=29) and vaginal approach (n=24). Demographics were similar between the two groups. The median age was 67 (63-73) years and median follow-up was 1.5 (0.8-2.3) years. 
VCUG cystocele grade results were correlated with MR def cystocele grade results (p=0.0093) but 9 (19%) patients with grade 1 cystocele on VCUG were noted to have grade 2 or 3 cystocele on MR def. Anterior compartment prolapse stage based on the Ba point from physical exam was also correlated with MR def anterior compartment prolapse (p=0.0026) with only 7 (14%) being upgraded from stage 0 or 1 on physical exam to grade 2 or 3 on MR def (Table 3). Apical compartment prolapse based on the C point from physical exam was not correlated with MR def results (p=0.20). Fifteen (39%) with stage 0 or 1 apical compartment prolapse on physical exam had grade 2 or 3 based on MR def. 5 (14%) with stage 2 or 3 on physical exam were noted to be grade 0 or 1 on MR def.
All patients had stage 2 or greater ACP. Eighteen (34%) had minor involvement of other compartments and 16 (30%) had involvement of all three compartments. 13 (72%) of those with minor apical and posterior compartment involvement had a vaginal surgical approach and 13 (81%) of those with triple-compartment involvement had an open/robotic surgical approach (p=0.0088) (Table 1).
Cystocele recurrence on physical exam occurred in 8 (15%) of patients with a slightly higher rate in those with a vaginal surgical approach (25% vs 7%, p=0.12), although not statistically significant. Only 1 (2%) patient needed a reoperation during the follow-up period and her initial repair was performed vaginally.

Women were selected for the study only if they had prior hysterectomy due to the increased incidence and complicated nature of POP in this patient subset. They were split into groups based on open/robotic vs vaginal approach with no differences noted in patient demographics between the two groups. Although not statistically significant, a larger number of patients in the open/robotic group had prior BNS surgery, which was expected given the increased complexity of compartmental anatomy in patients with prior surgery. The median follow-up time for patients was relatively short due to the recent operations in many of the patients, but the authors have extensive experience in prolapse repair and longer term follow-ups for the utilized operations.
Clinical exam and VCUG ACP results were correlated with MR def findings, but a substantial percentage (14% and 19%, respectively) did show increased ACP on MR def compared to exam and VCUG. Apical compartment prolapse assessed on physical exam was not correlated with MR def findings and 39% of patients had greater prolapse on MR def. Posterior compartment prolapse was difficult to assess on most patients clinically and was not able to be compared to MR def results. The disagreement between physical exam, VCUG and MR def support the utility of MR def in assessing POP, particularly in the apical and posterior compartments.
One of the weaknesses of the study was that all patients did not receive the expected surgical approach based on MR def results. Despite this, most patients with minor apical and posterior compartment involvement had a vaginal surgical approach and most of those with triple-compartment involvement had an open/robotic surgical approach. In the post-operative follow-up period, there was a low cystocele recurrence rate on physical exam and an even lower percentage of patients needed reoperation.

The similar recurrence rates with each surgical approach, low overall reoperation rate and disagreement between physical exam, VCUG and MR def results support the use of MR def as an imaging modality to help guide surgical decision for ACP.

Khatri G, et al. Abdom Radiol. 2019
Khatri G, et al. Magn Reson Imaging Clin N Am. 2017
Arbique G, et al. Urology. 2006