Hiatus and pelvic floor failure patterns in pelvic organ prolapse: A 3D MRI study of structural interactions using a Level III conceptual model

Background: A large urogenital hiatus in Level III results in a higher risk of developing pelvic organ prolapse after birth and failure after prolapse surgery. Deepening of the pelvic floor and downward rotation of the levator plate have also been linked to prolapse. Currently we lack data that evaluates how these measures relate to one another and to prolapse occurrence and size.

Objective: This study uses measurements from a published conceptual model to compare women with and without prolapse to determine the magnitude of difference between cases and controls and to quantify the interrelationships among different aspects of pelvic floor shape and structure.

Study design: Ninety-one women with anterior predominant prolapse and uterus in situ who had 3D MRI and 30 similar women with normal support were studied. Resting scans were used to avoid the influence of the prolapse dilating the hiatus. Measurements assessed three domains: hiatus size (urogenital and levator hiatus); length of the surrounding pelvic floor muscles (pubovisceral, puborectal, iliococcygeal muscles); the shelf-like posterior pelvic floor (levator plate shape, levator bowl volume), and bony pelvic dimensions. Effect sizes were calculated and principal component shape analysis performed to evaluate levator plate shape. A-scores were calculated and a value greater than 1.68 (95^th percentile) was considered the "failure" criterion. Frequency and severity of structural support site failure were analyzed by prolapse size.

Results: Resting urogenital and levator hiatal areas were 68% and 59% larger in the prolapse group compared to controls. These area enlargements were 2-4 times larger than the anterior-posterior dimension enlargements (urogenital hiatus 36%; levator hiatus 13%). The greatest muscle length differences between groups occurred in the pubovisceral (34%) and puborectal (25%) muscles compared to the iliococcygeal muscle (8%)-roughly half the area differences. Levator bowl volume was 63% deeper with prolapse. Urogenital hiatus and levator hiatus areas were strongly correlated with pubovisceral and puborectal muscle length (.7 to .8), while iliococcygeal muscle length had lower correlations (.4 to .5). Levator bowl volume correlated strongly with hiatal enlargement (.7 to .8) and muscle length (pubovisceral and puborectal muscles), moderately so with levator plate and iliococcygeal muscle, and weakly with bony dimension. Failure frequency increased with prolapse size for urogenital hiatus anterior-posterior (p=.001) and area (p=.019). By contrast, levator hiatus area was similar for all prolapse sizes (p=.288), while levator hiatus anterior-posterior failure was more common in larger prolapses (p=.018) but with smaller percentages of failure than levator hiatus area (p<.01). Both levator bowl volume (p=.015) and levator plate (p=.045) trended toward increasing failure with larger prolapse sizes. Among women with enlarged urogenital hiatus at straining, 43% and 28% had normal urogenital hiatus anterior-posterior and area at rest, respectively.

Conclusion: Changes in the shape and dimensions of the pelvic floor are complex and are not captured by a single measure (such as the urogenital hiatus anterior-posterior dimension, which does not capture its lateral expansion). The failure patterns were different between small and large prolapses. Understanding why could lead to improved prevention and treatments for Level III failures.