MRI Analysis of Dynamic Changes in Posterior Compartment Structures in Young, Nulliparous Women

INTRODUCTION: Prior studies show differences in posterior compartment structures on magnetic resonance imaging (MRI) between resting and straining in patients with pelvic organ prolapse when compared to age- and parity-matched controls (Lewicky-Gaupp 2010). Further, nulliparous older women (>70) have larger resting levator bowl volume than younger women (<40) (Swenson 2020). OBJECTIVE: In this MRI measurement technique development pilot study, we perform a detailed analysis of the posterior compartment structures at rest and with straining in young, nulliparous subjects without prolapse symptoms and compare these MRI measurements to their POP-Q measurements. METHODS: Six OBGYN physicians from differing institutions and career stages met weekly to learn pelvic floor MRI measurement techniques. Using an MRI biorepository, mid-sagittal images were analyzed at rest and with straining in nulliparous subjects aged <40 years without prolapse symptoms. We utilized 3D Slicer software (slicer.org) to measure posterior vaginal wall length (PVW), urogenital hiatus (UGH), levator hiatus (LH), levator plate length (LP), and levator bowl volume (LBV). Six independent evaluators, five of which were novices, performed measurements that were averaged, and interrater reliability was assessed using the intraclass correlation coefficient (ICC). Statistical analysis was performed using Python; independent t-test and Pearson correlation were utilized. RESULTS: Eleven subjects were available for analysis. Table 1 shows demographic characteristics. Compared with resting measurements, UGH (p<.001), LH (p<.001), and LBV (p<.001) increased with straining without significant change in LP or PVW length. There was a strong positive correlation between MRI posterior fornix location and PVW length (r=.77, p<.001). There was a strong negative correlation between MRI cervix location and POP-Q C (r= −0.70, p=.01) and MRI posterior fornix and POP-Q D (r=−0.697, p=.02). ICC was .95 indicating excellent interrater reliability (Figure 1). CONCLUSIONS: MRI allows visualization of real-time changes within the same subject and provides insight into the biomechanics of the pelvic floor beyond what can be learned from textbooks and cadaveric dissections. In young, nulliparous women without prolapse symptoms, MRI reveals dynamic changes in posterior compartment structures between rest and strain. During straining, there is relaxation of the pelvic floor shown by increased UGH, LH, and LBV. Although straining did not change the length of LP or PWV, it likely changed the orientation of the levator plate, as evidenced by the increase in LBV. The strong correlations between MRI measurement and POP-Q measures reflect that these measures describe the same phenomenon. The ICC for this study showed excellent interrater reliability, demonstrating that targeted education and training in pelvic MRI analysis produces precise and accurate measurement results among novice learners (Table 2).