Vaginal morphology and position associated with prolapse recurrence after vaginal surgery: A secondary analysis of the DEMAND study

Abstract

Objective: To identify vaginal morphology and position factors associated with prolapse recurrence following vaginal surgery.

Design: Secondary analysis of magnetic resonance images (MRI) of the Defining Mechanisms of Anterior Vaginal Wall Descent cross-sectional study.

Setting: Eight clinical sites in the US Pelvic Floor Disorders Network.

Population or sample: Women who underwent vaginal mesh hysteropexy (hysteropexy) with sacrospinous fixation or vaginal hysterectomy with uterosacral ligament suspension (hysterectomy) for uterovaginal prolapse between April 2013 and February 2015.

Methods: The MRI (rest, strain) obtained 30-42 months after surgery, or earlier for participants with recurrence who desired reoperation before 30 months, were analysed. MRI-based prolapse recurrence was defined as prolapse beyond the hymen at strain on MRI. Vaginal segmentations (at rest) were used to create three-dimensional models placed in a morphometry algorithm to quantify and compare vaginal morphology (angulation, dimensions) and position.

Main outcome measures: Vaginal angulation (upper, lower and upper-lower vaginal angles in the sagittal and coronal plane), dimensions (length, maximum transverse width, surface area, volume) and position (apex, mid-vagina) at rest.

Results: Of the 82 women analysed, 12/41 (29%) in the hysteropexy group and 22/41 (54%) in the hysterectomy group had prolapse recurrence. After hysteropexy, women with recurrence had a more laterally deviated upper vagina (p = 0.02) at rest than women with successful surgery. After hysterectomy, women with recurrence had a more inferiorly (lower) positioned vaginal apex (p = 0.01) and mid-vagina (p = 0.01) at rest than women with successful surgery.

Conclusions: Vaginal angulation and position were associated with prolapse recurrence and suggestive of vaginal support mechanisms related to surgical technique and potential unaddressed anatomical defects. Future prospective studies in women before and after prolapse surgery may distinguish these two factors.

Keywords: MRI; angulation; dimension; hysterectomy; hysteropexy; pelvic organ prolapse; position; prolapse recurrence; vagina; vaginal mesh.

Figure 1.

Illustration of the data processing steps of the vaginal analysis. (A) Axial magnetic resonance (MRI) scan of the participant at recovery (rest period following strain without prolapse reduced). The vagina was manually segmented with the lumen excluded. Vaginal segmentations across multiple MRI slices were stacked to reconstruct a 3D surface model of the vagina with zero thickness. To remove aliasing (sharp edges) from the 3D vaginal model, a non-biased smoothing algorithm was applied such that the model’s original shape and volume was preserved. (B) The 3D vaginal model was then placed in a morphometry algorithm to perform model-based vaginal measurements through anatomical landmark detection. First, the 3D model was iteratively sliced in the axial direction every 1.5 mm (half of the MRI slice thickness). For each 3D vaginal slice (represented as a thin ribbon), the right (red point) and left (blue point) lateral edges were extracted and the centroid (black point) of the vaginal slice was calculated. The straight-line distance between the lateral edges defined the transverse width of the vaginal slice. This iterative process was used to find the maximum transverse width and establish the right (red points) and left (blue points) lateral margins and centerline (black points) of the vagina. The vaginal centerline points were split in half to represent the upper (green points) and lower (purple) vagina. A line of best fit was calculated for each set to define the upper (green line) and lower (purple) vaginal axes. The pelvic coordinate system (PCS, dashed arrows) was then applied to calculate physiological vaginal position (black points) and morphology (angulation, dimension) measures in 3D space.

Figure 2.

Visualization of the vaginal position and morphology (angulation, dimension) measures. (A) Vaginal position measures. Position (black points) of the vaginal apex and mid-vagina in the coronal and sagittal plane with respect to the 3D pelvic coordinate system (PCS). The X-, Y-, and Z-coordinates of each point correspond to the medial-lateral, anterior-posterior, and superior-inferior position, respectively. The orientation of the axes PCS indicates the positive direction. (B) Vaginal angle measures. Angles of the upper (green angle) and lower (purple angle) vagina in the coronal and sagittal plane. The coronal angles are with respect to the Z- (superior-inferior) axis and the sagittal angles are with regard to the Y- (anterior-posterior) axis. The upper-lower vaginal angles represent the angle formed at the transition point between the lower and upper vagina. In the coronal plane, it is calculated as the difference between the upper and lower vaginal coronal angles. In the sagittal plane, it is the sum of the upper and lower vaginal sagittal angles (gray angle). (C) Vaginal dimension measures. The vaginal length (black dotted double arrow) is given by the length of the vaginal centerline. The maximum transverse width (solid black double arrow) is given by the largest straight-line distance between the right (red point) and left (blue point) lateral margin of the vagina across all vaginal slices along the total vaginal length. The vaginal surface area and volume are given by the surface area and the amount of space enclosed by the surface of the 3D vaginal model, respectively.