Long-term Urinary outcomes after transvaginal uterovaginal prolapse repair with and without concomitant midurethral slings

Abstract

Importance:

Many providers place concomitant midurethral slings during pelvic organ prolapse repair, yet growing evidence supports staged midurethral sling placement.

Objectives:

To compare urinary function after transvaginal uterovaginal prolapse repair with and without midurethral sling.

Study Design:

Secondary analysis of the SUPeR trial (hysterectomy with uterosacral ligament suspension vs. mesh hysteropexy). Our primary outcome was Urinary Distress Inventory (UDI-6) through 5 years compared between women with and without a concomitant sling within prolapse repair arms. Sling effect was adjusted for select clinical variables and interaction terms (α=0.05).

Results:

The sling group included 90 women (43 hysteropexy, 47 hysterectomy) and the no sling group included 93 women (48 hysteropexy, 45 hysterectomy). At baseline, the sling group reported more bothersome stress (66% vs 36%, p<0.001) and urgency incontinence (69% vs 48%, p=0.007). For hysteropexy, there were no significant long-term differences in UDI-6 scores or bothersome urine leakage between sling groups. For hysterectomy, women with sling had better UDI-6 scores across timepoints [adjusted mean difference −5.1, 95% confidence interval (CI) −9.9 to −0.2]; bothersome stress and urgency leakage were less common in the sling group [stress adjusted odds ratio (AOR) 0.1 (95% CI 0.0 to 0.4); urge AOR 0.5 (95% CI 0.2 to 1.0)]. Treatment for stress incontinence over 5 years was similar in the sling (7.9%) vs no sling groups (7.6%).

Conclusions:

5-year urinary outcomes of concomitant midurethral sling may vary by type of transvaginal prolapse surgery, with possible benefit of midurethral sling at the time of vaginal hysterectomy with apical suspension but not after mesh hysteropexy.

INTRODUCTION

Urinary symptoms are extremely common in women with pelvic organ prolapse.¹ Although preexisting overactive bladder symptoms generally improve after prolapse surgery,² persistent or de novo urinary incontinence symptoms remain problematic.

Symptomatic and occult stress urinary incontinence (SUI) are often treated at prolapse surgery with a concomitant midurethral sling (MUS).³ Placement of a concomitant MUS at prolapse surgery significantly reduces post-operative de novo SUI.⁴ However, the number needed to treat with MUS to prevent one case of urinary incontinence at 12 months was 6.3.⁴ While many providers continue to place concomitant MUS for treatment and prevention of SUI, a growing body of evidence demonstrates spontaneous resolution of pre-operative SUI (symptomatic and occult) after prolapse repair and lends support to a staged approach for MUS. A staged MUS strategy is one in which the determination to treat SUI is delayed until after prolapse repair. Both prospective and retrospective studies have reported placement of proportionately fewer MUS with a staged approach; though follow-up was limited.^5–7

The Pelvic Floor Disorders Network Study of Uterine Prolapse Procedures Randomized Trial (SUPeR) provides an opportunity to assess the longer-term impact of a concomitant MUS on urinary function and quality of life in a prospective, well-characterized population of women undergoing transvaginal uterovaginal prolapse repairs. Our objective was to compare urinary outcomes across 5 years after transvaginal uterovaginal prolapse repairs with and without MUS. We hypothesized that after controlling for baseline SUI bother, postoperative urinary outcomes would not differ between women with and without MUS regardless of prolapse repair.

METHODS

This is a secondary analysis of the SUPeR trial, a randomized controlled trial which compared treatment success after transvaginal sacrospinous hysteropexy with graft (Uphold® LITE) (hysteropexy group) and vaginal hysterectomy using uterosacral ligament suspension (hysterectomy group), in women with symptomatic uterovaginal prolapse beyond the hymen. Full protocol details as well as 3- and 5-year outcomes have been published.^8–10 The protocol did not specify treatment for stress urinary incontinence leaving the decision to surgeon’s discretion. In the original study, participants provided written informed consent, and IRB approval was obtained at each site. The primary aim of this secondary analysis was to assess the impact of performing a concomitant MUS on urinary function and condition-specific quality of life over time. Our primary outcome was the short form Urinary Distress Inventory Score (UDI-6) subscale of the 20 question Pelvic Floor Distress Inventory (PFDI-20)¹¹ ascertained at 6 weeks, 6, 12, 18, 24, 36, 48 and 60 months after surgery, comparing results between women with and without concomitant MUS. The UDI-6 short-form of the 28 question UDI assesses urinary function (score range 0–100, with higher scores indicating worse symptoms).^11,12 To aid in clinical interpretation, we converted the minimal clinically important difference (MCID) published for the long form UDI to a MCID for the UDI-6. Using previously published conversion formulas, the MCID for the UDI-6 is 5.8.^12,13 Analyses were conducted on data from all randomized subjects.

Secondary outcomes included postoperative bothersome stress urine leakage [defined as an affirmative response to question 17 on the PFDI-20 ‘Do you usually experience urine leakage related to coughing, sneezing, or laughing?’ with bother qualified as ‘Somewhat’, ‘Moderately’, or ‘Quite a bit’], bothersome urgency urine leakage [defined as an affirmative response to question 16 on the PFDI-20 ‘Do you usually experience urine leakage associated with a feeling of urgency, that is, a strong sensation of needing to go to the bathroom?’ and bother of ‘Somewhat’ or greater], Urinary Impact Questionnaire (UIQ) score (subscale of the Pelvic Floor Impact Questionnaire¹¹), Incontinence Severity Index (ISI) score¹⁴, Patient Global Impression of Improvement (PGI-I) score (improvement defined as ‘much better’ or ‘very much better’)¹⁵, and Short Form Health Survey – 12 (SF-12)¹⁶ Physical and Mental component scores through 5 years. We also assessed aggregate secondary outcomes through 5 years including new or worsening SUI, urgency urinary incontinence (UUI), and voiding dysfunction; postoperative urinary incontinence treatment, and postoperative complications related to MUS placement (including mesh exposure, mesh erosion, or sling revision).

Longitudinal models for repeated outcomes collected across post-operative visits through 60 months were fit to assess the association with concomitant MUS. Because the primary aim of this study was to assess the impact of MUS on postoperative urinary outcomes, outcomes assessed after postoperative UI treatments were excluded from modeling, as they could reflect improvements due to any subsequent treatments and not the MUS. Models accounted for within-participant correlations across visits using an auto-regressive order 1 structure. Since the decision to perform a MUS was not randomized and the indication was prevention or treatment of SUI, models were adjusted for preoperative bothersome stress urine leakage as an a priori specified covariate. All models included the three-way interaction between concomitant MUS, apical pelvic organ prolapse repair, and visit as well as all the pairwise interaction terms. To account for the imbalance in potential risk factors when evaluating the effect of the non-randomized MUS on urinary function and quality of life outcomes, preoperative clinical and demographic characteristics were considered for inclusion as additional model covariates based on clinical importance or a statistical difference between MUS groups at a significance level of 0.20. Longitudinal models for SF-12 Physical and Mental component scores additionally adjusted for baseline scores which differed between women with and without MUS. Differences between sling groups were estimated within SUPeR apical repair groups if interaction terms were statistically significant. Otherwise, differences were averaged across the apical repair groups. If outcomes did not vary over time, then estimates were averaged across time. The word ‘average’ as utilized in the results refers to a mean estimate.

Due to the low frequency of aggregate outcomes, no models were fitted and descriptive statistics are shown by apical repair. All tests were conducted at a significance level of 0.05 and no adjustments were made for multiple comparisons. SAS v9.4 (Cary, NC) was used for statistical analysis.

RESULTS

There were 90 women in the MUS group (43 hysteropexy, 47 hysterectomy) and 93 women in the No MUS group (48 hysteropexy, 45 hysterectomy). The overall mean age was 66 ± 7.3 years. At baseline, the MUS group included fewer participants with Stage ≥3 pelvic organ prolapse, more participants with bothersome SUI and UUI, and had higher UDI-6 scores (Table 1). Preoperative bothersome stress urine leakage (a priori specified), preoperative advanced POP-Q stage (≥ 3) and postvoid residual (chosen based upon significance, Table 1) as well as age and BMI (chosen based upon clinical relevance) were selected as model covariates. Four women in each SUPeR surgical group reported previous SUI surgery, only one of whom was in the MUS group. Due to small numbers this variable was not adjusted for in analysis.

Table 1.

Potential Risk Factors by Concomitant Midurethral Sling (MUS)

Characteristic, n (%) or Median (P25, P75) unless otherwise specified	N	MUS N=90	No MUS N=93	Risk Difference/Location Shift (95% CI) a	P-value a
Patient Demographics
Age, years	183	65.6 (59.7, 72.1)	66.6 (61.6, 71.4)	−0.3 (−2.5 to 1.8)	0.79
Caucasian	183	78 (87)	80 (86)	1 (−10 to 11)	>0.99
Hispanic or Latina	178	9 (10)	7 (8)	3 (−6 to 12)	0.61
Married/living with partner	183	56 (62)	63 (68)	−6 (−19 to 8)	0.44
Higher education after high school	178	54 (62)	56 (62)	1 (−14 to 15)	>0.99
Medicaid/Medicare	183	51 (57)	53 (57)	0 (−15 to 14)	>0.99
Medical History
History of smoking	183	25 (28)	23 (25)	3 (−10 to 16)	0.74
Current smoker	183	0 (0)	4 (4)	−4 (−11 to 0)	0.12
Body mass index (kg/m2)	183	27.7 (24.4, 31.2)	28.3 (25.3, 31.1)	−0.4 (−1.5 to 1.0)	0.63
Gravidity	183	3.0 (2.0, 4.0)	3.0 (2.0, 4.0)	0.0 (0.0 to 0.0)	0.88
Vaginal parity	183	3.0 (2.0, 3.0)	2.0 (2.0, 3.0)	0.0 (0.0 to 0.0)	0.59
Postmenopausal	183	87 (97)	91 (98)	−1 (−8 to 5)	0.68
Prior stress urinary incontinence surgery	183	1 (1)	7 (8)	−6 (−14 to 0)	0.06
Prior pelvic organ prolapse surgery	183	2 (2)	7 (8)	−5 (−13 to 1)	0.17
Other prior pelvic surgery	183	48 (53)	42 (45)	8 (−7 to 23)	0.30
Baseline Pelvic Floor Measurements
POP-Q measurement, cm
Ba	183	3.0 (1.0, 4.0)	3.0 (2.0, 4.0)	0.0 (−1.0 to 0.0)	0.41
Bp	183	0.3 (−2.0, 2.0)	0.0 (−2.0, 3.0)	0.0 (−1.0 to 1.0)	0.81
C	183	0.0 (−2.0, 3.0)	0.0 (−3.0, 3.0)	0.0 (−1.0 to 1.0)	0.84
TVL	183	9.0 (9.0, 10.0)	9.0 (8.0, 10.0)	0.0 (0.0 to 1.0)	0.06
Advanced POP-Q stage b (≥ 3)	183	66 (73)	80 (86)	−13 (−25 to −1)	0.04
Postvoid residual, mL	182	30.0 (10.0, 55.0)	45.0 (10.0, 100.0)	−8.0 (−20.0 to 2.0)	0.19
Baseline Patient Reported Outcome Scores
Urinary Function
Bothersome stress urine leakage c	183	59 (66)	33 (36)	30 (15 to 44)	<0.001
Bothersome urge urine leakage d	183	62 (69)	45 (48)	21 (4 to 34)	0.007
Urogenital Distress Inventory (UDI-6) score	183	47.9 (25.0, 66.7)	33.3 (16.7, 50.0)	12.5 (4.2 to 20.8)	0.004
Urinary Impact Questionnaire (UIQ) score	183	21.4 (0.0, 42.9)	14.3 (0.0, 33.3)	0.0 (0.0 to 9.5)	0.24
Incontinence Severity Index (ISI) score	183	4.0 (2.0, 8.0)	2.0 (0.0, 4.0)	2.0 (1.0 to 4.0)	<0.001
General Quality of Life
Short Form Health Survey (SF-12)
Physical component score	181	49.9 (38.6, 55.3)	44.4 (33.2, 53.8)	3.1 (0.2 to 6.3)	0.04
Mental component score	181	53.5 (43.5, 57.5)	54.8 (47.1, 58.4)	−1.9 (−4.4 to 0.2)	0.09
Concomitant Pelvic Organ Prolapse (POP) Repairs
Hysteropexy apical POP repair (vs. hysterectomy)	183	43 (48)	48 (52)	−4 (−18 to 11)	0.66
Anterior POP repair	183	72 (80)	75 (81)	−1 (−13 to 11)	>0.99
Posterior POP repair/perineorrhaphy	183	52 (58)	50 (54)	4 (−11 to 19)	0.66

Data are Median (P25, P75) for continuous measures and n (%) for categorical measures unless otherwise specified. P25=25^th Percentile; P75=75^th Percentile; CI=Confidence Interval; POP-Q=Pelvic Organ Prolapse Quantification; TVL=total vaginal length; POP=Pelvic Organ Prolapse.

^aFor nominal categorical measures, the p-values were obtained from Fisher’s exact test and exact risk difference and 95% CI limits were obtained by exact methods based on the score statistic. For continuous measures, p-values were obtained using Wilcoxon Rank-Sum test and location shift and 95% confidence intervals were obtained with a Hodges-Lehmann estimation of location shift. All tests were conducted at a significance level of 0.20 for selection of risk factors to account for the imbalance in preoperative characteristics when evaluating the effect of the non-randomized MUS on urinary function. No adjustments for multiple comparisons were made.

^bPelvic Organ Prolapse Quantification (POP-Q) Stages: Stage 2-The vagina is prolapsed between 1 cm above the hymen and 1 cm below the hymen; Stage 3-The vagina is prolapsed more than 1 cm beyond the hymen but is not everted within 2 cm of its length; Stage 4-The vagina is everted to within 2 cm of its length.

^cBothersome stress urine leakage is defined as a positive response to PFDI-20 item 17 ‘Do you usually experience urine leakage related to coughing, sneezing, or laughing?’ and a degree of bother of ‘Somewhat’, ‘Moderately’, or ‘Quite a bit’ to the follow-up question ‘If yes, how much does it bother you?’.

^dBothersome urge urine leakage is defined as a positive response to PFDI-20 item 16 ‘Do you usually experience urine leakage associated with a feeling of urgency, that is, a strong sensation of needing to go to the bathroom?’ and a degree of bother of ‘Somewhat’, ‘Moderately’, or ‘Quite a bit’ to the follow-up question ‘If yes, how much does it bother you?’.

The effect of MUS on urinary outcomes varied by type of prolapse repair. For hysteropexy, there was a significant interaction between MUS and visit timeframe (p <0.001), where women with MUS reported transient lower (better) UDI-6 scores compared to women without MUS only at 6 weeks [adjusted mean difference (AMD) −9.0 (95% CI −16.0 to −1.9)] with no significant differences at other timepoints (Figure 1, Supplemental Table S1). Similarly, the effect of MUS on ISI score varied by visit timeframe (interaction p=0.02) with lower scores among MUS at earlier postoperative timepoints, although these individual differences at each timepoint did not reach significance (Supplemental Table S1). No significant interaction between MUS and time was found for the UIQ, and scores were not significantly different between MUS groups averaged across time [AMD −1.6 (−5.6 to 2.4), p=0.43] (Supplemental Table S1). Notably, we observed a significant MUS and visit interaction in outcomes collected at 6 weeks (UDI-6 and ISI) as opposed to those with initial collection at 6 months (UIQ). Bothersome stress leakage was more common in women without MUS at early timepoints, but this difference did not persist and was not significant after 12 months (Figure 2, Supplemental Table S3). Bothersome urgency leakage did not differ between MUS groups [adjusted odds ratio (AOR) 1.1 (0.6 to 1.9), p=0.83].

Figure 1.

Urogenital Distress Inventory (UDI-6) Score Adjusted Means and 95% Confidence Intervals by Apical Pelvic Organ Prolapse Repair and Concomitant Midurethral Sling (MUS)^a

^a Adjusted means, 95% confidence intervals, and p-values comparing the continuous outcome by concomitant midurethral sling (MUS) across post-operative visits through 60 months are obtained from general linear models controlling for within-participant correlations across visits with an auto-regressive order 1 structure and adjusted for MUS, visit, apical pelvic organ prolapse repair, three way interaction between MUS, apical pelvic organ prolapse repair, and visit as well as all the pairwise interaction terms, and the following pre-operative characteristics to account for the imbalance in these potential risk factors when evaluating the effect of the non-randomized MUS on the outcome: age, body mass index, advanced POP-Q stage (≥ 3) ^b, postvoid residual, and bothersome stress urine leakage ^c. The p-value corresponding to the three-way interaction is obtained from the type 3 partial sums of squares hypothesis test on the fitted model and the p-values for the MUS and visit interaction by apical pelvic organ prolapse repair are obtained by performing a partitioned analysis of the LS-means for the three-way interaction. The p-values corresponding to the MUS main effect by apical pelvic organ prolapse repair are obtained by performing a partitioned analysis of the LS-means for the interaction between MUS and apical pelvic organ prolapse repair.

^b Pelvic Organ Prolapse Quantification (POP-Q) Stages: Stage 2-The vagina is prolapsed between 1 cm above the hymen and 1 cm below the hymen; Stage 3-The vagina is prolapsed more than 1 cm beyond the hymen but is not everted within 2 cm of its length; Stage 4-The vagina is everted to within 2 cm of its length.

^c Bothersome stress urine leakage is defined as a positive response to PFDI-20 item 17 ‘Do you usually experience urine leakage related to coughing, sneezing, or laughing?’ and a degree of bother of ‘Somewhat’, ‘Moderately’, or ‘Quite a bit’ to the follow-up question ‘If yes, how much does it bother you?’.

Figure 2.

Percent with Bothersome Stress Urine Leakage ^a by Apical Pelvic Organ Prolapse Repair and Concomitant Midurethral Sling (MUS) ^b

^a Bothersome stress urine leakage is defined as a positive response to PFDI-20 item 17 ‘Do you usually experience urine leakage related to coughing, sneezing, or laughing?’ and a degree of bother of ‘Somewhat’, ‘Moderately’, or ‘Quite a bit’ to the follow-up question ‘If yes, how much does it bother you?’. All events reported at or after 36 months are aggregated to the 36-month visit.

^b P-values comparing the binary outcome by concomitant midurethral sling (MUS) across post-operative visits through 60 months are obtained from generalized linear models with logit link controlling for within-participant correlations across visits with an auto-regressive order 1 structure and adjusted for MUS, visit, apical pelvic organ prolapse repair, three way interaction between MUS, apical pelvic organ prolapse repair, and visit as well as all the pairwise interaction terms, and the following pre-operative characteristics to account for the imbalance in these potential risk factors when evaluating the effect of the non-randomized MUS on the outcome: age, body mass index, advanced POP-Q stage (≥ 3) ^c, postvoid residual, and bothersome stress urine leakage ^a. The p-value corresponding to the three-way interaction is obtained from the type 3 partial sums of squares hypothesis test on the fitted model and the p-values for the MUS and visit interaction by apical pelvic organ prolapse repair are obtained by performing a partitioned analysis of the LS-means for the three-way interaction. The p-values corresponding to the MUS main effect by apical pelvic organ prolapse repair are obtained by performing a partitioned analysis of the LS-means for the interaction between MUS and apical pelvic organ prolapse repair.

^c Pelvic Organ Prolapse Quantification (POP-Q) Stages: Stage 2-The vagina is prolapsed between 1 cm above the hymen and 1 cm below the hymen; Stage 3-The vagina is prolapsed more than 1 cm beyond the hymen but is not everted within 2 cm of its length; Stage 4-The vagina is everted to within 2 cm of its length.

For hysterectomy, no significant interactions were found between MUS and time, so MUS group estimates were averaged across timepoints. Women with MUS had lower (better) UDI-6 scores averaged across timepoints compared to women without MUS [AMD −5.1 (−9.9 to −0.2), p=0.04, Figure 1, Supplemental Table S2]. Similarly, women in the MUS group reported significantly lower UIQ scores (lower adverse impact) across time points [−4.7 (−8.6 to −0.8), p=0.02]. Both bothersome stress (Figure 2) and urgency leakage were less common in the MUS group averaged over time [stress AOR 0.1 (0.0 to 0.4), p<0.001; urge AOR 0.5 (0.2 to 1.0), p=0.04, Supplemental Table S4]. However, there was no significant group difference in ISI scores averaged across time [AMD −0.6 (−1.5 to 0.3), p=0.20, Supplemental Table S2].

No significant differences were found between MUS and No MUS averaged across apical repairs in SF-12 Physical [AMD 0.1 (−1.6 to 1.8), p=0.95] or Mental [AMD 0.9 (−0.7 to 2.4), p=0.23] scores after adjusting for preoperative scores which differed at baseline (data not shown). Over 85% of women with or without MUS reported that their condition was “much better” or “very much better” on the PGI-I (Supplemental Tables S3 and S4). Only descriptive statistics are reported for PGI-I as the longitudinal model adjusting for apical repair did not converge due to extremely high rates of improvement in women with and without MUS.

Overall treatment for SUI over 5 years was similar in the MUS groups overall [MUS 7/89 (7.9%) vs No MUS 7/92 (7.6%), Table 2] or by procedure type [mesh hysteropexy: MUS 5/43 (11.6%) vs No MUS 4/47 (8.5%); hysterectomy: MUS 2/46 (4.3%) vs No MUS 3/45 (6.7%), Table 2]. Very few women in either of the prolapse repair arms underwent subsequent surgical intervention for SUI 3/91 women (3.3%, 2 MUS and 1 bulking) in the hysteropexy arm and 3/92 (3.3%, 3 MUS) in the hysterectomy arm] (Table 2).

Table 2.

Aggregate Outcomes through 5 years by Apical Pelvic Organ Prolapse Repair and Concomitant Midurethral Sling (MUS)

	Hysteropexy		Hysterectomy
Aggregate Outcome, n/N (%) unless otherwise specified	MUS N=43	No MUS N=48	MUS N=47	No MUS N=45
Efficacy Outcomes
New or worsening stress urinary incontinence (SUI) a	8/43 (19)	13/48 (27)	3/47 (6)	11/45 (24)
New or worsening urge urinary incontinence (UUI) b	16/43 (37)	16/48 (33)	12/47 (26)	15/45 (33)
New or worsening voiding dysfunction c	4/43 (9)	2/48 (4)	6/47 (13)	2/45 (4)
Treatment for urinary incontinence d	10/43 (23)	8/47 (17)	6/46 (13)	10/45 (22)
SUI treatment e	5/10 (50)	4/8 (50)	2/6 (33)	3/10 (30)
UUI treatment f	8/10 (80)	6/8 (75)	5/6 (83)	7/10 (70)
Safety/Tolerability Outcomes among patients with MUS g
Midurethral sling mesh exposure	0/43 (0)		2/47 (4)
Midurethral sling mesh erosion	0/43 (0)		0/47 (0)
Midurethral sling revision	3/43 (7)		0/47 (0)

CI=Confidence Interval

^aNew or worsening stress urinary incontinence (SUI) are identified based on the collection of new or worsening SUI complication and SUI treatment ^d at follow-up visits 6 weeks through 5 years.

^bNew or worsening urge urinary incontinence (UUI) are identified based on the collection of new or worsening UUI complication and UUI treatment ^d at follow-up visits 6 weeks through 5 years.

^cNew or worsening voiding dysfunction are identified based on the collection of new or worsening difficulty emptying bladder complication at follow-up visits 6 weeks through 5 years.

^dTreatment for urinary incontinence (UI) are identified based on the collection of UI treatment at follow-up visits 6 weeks through 5 years. Further classification of treatment for SUI and/or UUI was based on manual review by clinical expert.

^eInitial treatment for stress urinary incontinence (SUI) reported includes the following among hysteropexy mesh: Bulking (n=1), MUS (n=2), Pelvic floor muscle therapy (PFMT) (n=6). Among hysterectomy, initial treatment for SUI reported includes the following: MUS (n=3), PFMT (n=2).

^fInitial treatment for urge urinary incontinence (UUI) reported includes the following among hysteropexy mesh: Overactive Bladder Medication (n=11), PFMT (n=2), PTNS (n=1). Among hysterectomy, initial treatment for UUI reported includes the following: Overactive Bladder Medication (n=10), PTNS (n=2).

^gAmong patients with MUS, safety/tolerability outcomes such as mesh exposure, mesh erosion, and sling revision are identified based on specific complications/adverse events collected at follow-up visits 6 weeks through 5 years and the systematic collection of open-ended adverse events from surgery through 5 years.

DISCUSSION

This analysis found that the impact of MUS on urinary symptoms reported on the UDI-6 over 5-years varied by type of concomitant transvaginal uterovaginal prolapse surgery. Benefits of MUS were noted when performed with vaginal hysterectomy and uterosacral ligament suspension but not with mesh hysteropexy except at 6-weeks postoperatively when the MUS group reported fewer symptoms.

Similarly, the hysteropexy MUS group did not differ in other urinary symptom outcomes over time, such as SUI bother or condition-specific quality of life. In contrast, for those undergoing hysterectomy, a concomitant MUS resulted in lower (better) urinary incontinence symptom severity and SUI bother as well as better condition-specific quality of life compared to women without MUS averaged across all time points.

The explanation for the differential impact of MUS on urinary symptom outcomes between the types of apical prolapse repairs is unclear. At 5 years, this study’s parent SUPeR trial found that the only anatomic POP-Q measure that differed between hysteropexy vs. hysterectomy with uterosacral ligament suspension was TVL with an average 0.9 cm longer length in the hysteropexy arm. Neither the anterior compartment nor apical POP-Q points differed between groups.⁸ It might be assumed that these latter POP-Q points, which are highly correlated with one another in women with severe prolapse,¹⁷ would have the greatest effect on the continence mechanisms, reflecting straightening of the anterior vagina and decreasing urethral resistance or kinking.¹⁸ Given that the anterior vaginal wall POP-Q measurements remained similar between groups at 5 years, this variable does not explain the differential urinary outcomes between prolapse repair groups demonstrated in the current study.

It appears that differences in apical repairs have differential effects on MUS and continence mechanisms. While speculative, it is possible that placement of mesh in the anterior vaginal compartment at the time of prolapse repair may compromise urinary outcomes after concomitant MUS by altering MUS positioning or tensioning intraoperatively or during the healing process. Further investigation should assess whether anterior vaginal mesh, placed either vaginally or laparoscopically, impacts MUS position and longer-term function.

Currently, no standard algorithm exists to recommend a concomitant MUS versus a staged procedure in women undergoing prolapse surgery. Evidence exists to support both strategies. Studies indicate that MUS performed concomitantly with vaginal prolapse repair in women with symptomatic preoperative SUI decreases the risk of having persistent SUI. A meta-analysis found that transvaginal prolapse repair with concomitant MUS in women with SUI improves the risk of subjective SUI by approximately one-third and decreases the need for additional incontinence surgery.¹⁹ The vast majority of women in these studies underwent native tissue repairs and approximately two-thirds underwent apical repairs.⁵ Conversely, a staged approach is considered a reasonable option, as some SUI symptoms resolve after prolapse repair alone without undergoing concomitant incontinence procedures. In a multicenter randomized trial comparing vaginal prolapse repair with and without MUS among women with preoperative SUI, 26% reported complete resolution of SUI without a MUS, whereas 56% of women with a concomitant MUS had persistent SUI at 1 year follow-up.⁵ Other non-randomized studies of women with prolapse and SUI symptoms undergoing prolapse surgery including other approaches such as sacrocolpopexy demonstrated similar findings; resolution of SUI was seen in 30–69% of those without a MUS.^6,20 Based on our findings and the above summary of the literature, it is plausible that efficacy of concomitant MUS differs in the setting of anterior vaginal hysteropexy mesh for apical suspension. This suggests that predictive models regarding postoperative SUI should account for the specific prolapse procedure being performed.^21,22

The lack of a published MCID for the UDI-6 makes it challenging to assess the clinical significance of our findings. To aid in the clinical interpretation, we performed conversions of the published MCID for the long form UDI.¹³ Prior studies have demonstrated that a MCID for the UDI is 11¹², which converts to a MCID of 5.8 on the UDI-6 scale. We found that among women undergoing hysterectomy, those with concomitant MUS had average UDI-6 scores that were 5.1 points better than those without MUS. Though this difference is smaller than the converted MCID, it should be noted that the true MCID for our study population may differ from those published based on populations with urinary incontinence, since not all of our participants had urinary incontinence at baseline. Yet, using the available conversion, we demonstrate potentially clinically meaningful differences between MUS groups.

It should be noted that very few women in either of the prolapse repair arms underwent subsequent SUI procedures. Retreatment for SUI over 5-years was infrequent and did not differ by MUS group. Limitations of prior retrospective research have cited that postoperative SUI treatment rates could be falsely low over the short-term if patients were unable to pursue another procedure within the same calendar year because of cost, insurance deductibles or time off work.^6,7 Our data demonstrated that SUI re-treatment remained low over 5 years even among those who do not undergo concomitant MUS. This emphasizes the importance of both long-term follow-up in surgical outcomes studies and further patient-centered work to evaluate reasons why patients do not seek treatment for persistent or de novo SUI after prolapse repair.

This study’s strengths include its population of a well-characterized cohort of women undergoing prolapse repair from 9 clinical sites across the US. The participants were randomized to two transvaginal surgical apical repairs and prospectively followed. This study utilized validated questionnaires to assess urinary symptoms in long-term follow-up, filling a gap in the current literature. Longitudinal analyses were performed accounting for intra- and inter-subject variation between groups. The major limitation of this study is that women were not randomized to MUS, introducing potential bias when comparing urinary symptoms between MUS groups. Though we a priori controlled for preoperative bothersome stress urine leakage in modeling, there may have been other unmeasured differences between women who did and did not undergo concomitant MUS. Given that this was an exploratory analysis and no adjustments were made for multiple comparisons, our results should be interpreted with caution. At the same time, we found consistent results across the UI outcomes, lending support to our findings. Additionally, incontinence was not objectively characterized. Lastly, our study findings may be of less clinical relevance following the FDA suspension of vaginal mesh for prolapse,²³ and do not necessarily reflect outcomes after mesh sacrocolpopexy.

In conclusion, the current study demonstrated that the efficacy and durability of a concomitant MUS and subsequent postoperative urinary symptoms may be differentially impacted by the type of uterovaginal prolapse repair. A concomitant MUS at the time of hysterectomy with uterosacral ligament suspension was associated with better urinary outcomes over time. In contrast, there was no sustained benefit of concomitant MUS at the time of mesh hysteropexy. Few women in either MUS group pursued subsequent surgery for SUI through 5 years, supporting a patient’s preference in decision making between concomitant or staged management of potential or symptomatic SUI. To guide surgeons and optimize patient outcomes, further research is warranted to evaluate the impact of other apical prolapse procedures such as sacrocolpopexy on the efficacy of concomitant MUS and postoperative urinary symptoms.