Association between adjuvant posterior repair and success of native tissue apical suspension

Abstract

Background: Posterior repairs and perineorrhaphies are often performed in prolapse surgery to reduce the size of the genital hiatus. The benefit of an adjuvant posterior repair at the time of sacrospinous ligament fixation or uterosacral ligament suspension is unknown.

Objective: We aimed to determine whether an adjuvant posterior repair at transvaginal apical suspension is associated with improved surgical success.

Materials and methods: This secondary analysis of Operations and Pelvic Muscle Training in the Management of Apical Support Loss (OPTIMAL) trial compared 24-month outcomes in 190 participants who had a posterior repair (posterior repair group) and 184 who did not (no posterior repair group) at the time of sacrospinous ligament fixation or uterosacral ligament suspension. Concomitant posterior repair was performed at the surgeon's discretion. Primary composite outcome of "surgical success" was defined as no prolapse beyond the hymen, point C ≤ -2/3 total vaginal length, no bothersome bulge symptoms, and no retreatment at 24 months. The individual components were secondary outcomes. Propensity score methods were used to build models that balanced posterior repair group and the no posterior repair group for ethnographic factors and preoperative Pelvic Organ Prolapse Quantification values. Adjusted odds ratios were calculated to predict surgical success based on the performance of a posterior repair. Groups were also compared with unadjusted χ² analyses. An unadjusted probability curve was created for surgical success as predicted by preoperative genital hiatus.

Results: Women in the posterior repair group were less likely to be Hispanic or Latina, and were more likely to have had a prior hysterectomy and to be on estrogen therapy. The groups did not differ with respect to preoperative Pelvic Organ Prolapse Quantification stage; however, subjects in the posterior repair group had significantly greater preoperative posterior wall prolapse. There were no group differences in surgical success using propensity score methods (66.7% posterior repair vs 62.0% no posterior repair; adjusted odds ratio, 1.07; 95% confidence interval, 0.56-2.07; P = 0.83) or unadjusted test (66.2% posterior repair vs 61.7% no posterior repair; P = 0.47). Individual outcome measures of prolapse recurrence (bothersome bulge symptoms, prolapse beyond the hymen, or retreatment for prolapse) also did not differ by group. Similarly, there were no differences between groups in anatomic outcomes of any individual compartment (anterior, apical, or posterior) at 24 months. There was high variation in performance of posterior repair by surgeon (interquartile range, 15-79%). The unadjusted probability of overall success at 24 months, regardless of posterior repair, decreased with increasing genital hiatus, such that a genital hiatus of 4.5 cm was associated with 65.8% success (95% confidence interval, 60.1-71.1%).

Conclusion: Concomitant posterior repair at sacrospinous ligament fixation or uterosacral ligament suspension was not associated with surgical success after adjusting for baseline covariates using propensity scores or unadjusted comparison. Posterior repair may not compensate for the pathophysiology that leads to enlarged preoperative genital hiatus, which remains prognostic of prolapse recurrence.

Trial registration: ClinicalTrials.gov NCT00597935.

Keywords: apical suspension; colpopexy; perineorrhaphy; posterior colporrhaphy; posterior repair; prolapse; propensity score analysis; sacrospinous ligament fixation; transvaginal surgery; uterosacral ligament suspension; variation in practice; vault prolapse.

Figure 1.. Posterior Repair Performed by Surgeon

Each circle represents an individual surgeon. The y-axis shows the number of surgeries performed by that surgeon and the x-axis is the percent of those surgeries that included a Posterior Repair.

Figure 2. Predicted Probability of Surgical Success at 24 Months Plotted Against Baseline POPQ Strain GH

The predicted probabilities and the Wald 95% confidence limits (shaded area) are obtained from a logistic regression model predicting surgical success at 24 monthsbased on preoperative POPQ strain GH.

Figure 3. Predicted Probability of Surgical Success at 24 Months Plotted Against Baseline POPQ Point Bp

The predicted probabilities and the Wald 95% confidence limits (shaded area) are obtained from a logistic regression model predicting surgical success at 24 months based on preoperative POPQ point Bp.