Pelvic organ prolapse surgery and overactive bladder symptoms—a population-based cohort (FINPOP)

Introduction and hypothesis It is unclear how compartment of pelvic organ prolapse (POP) impacts overactive bladder (OAB) symptom severity or improvement after POP surgery. We hypothesized that anterior and apical prolapse are more strongly associated with OAB symptoms than posterior compartment prolapse. Methods A total of 2933 POP surgeries from a prospective population-based cohort were divided into two groups: (1) anterior and/or apical compartment surgery (± posterior repair), N = 2091; (2) posterior repair only, N = 478. Urinary frequency and urgency urinary incontinence (UUI) were evaluated using PFDI-20 (bothersome symptom: score 3–4) at baseline, 6, and 24 months. Association between degree of POP in specific compartments and symptoms at baseline was estimated with generalized linear models and between compartment of surgery and symptom improvement with generalized estimating equations. Results At least one bothersome symptom was reported by 40% at baseline, 14% at 6, and 19% at 24 months. At baseline, urinary frequency was associated with degree of anterior and apical and UUI with anterior compartment prolapse. Women undergoing surgery for anterior/apical compartment started with worse symptoms and experienced greater improvement than women undergoing posterior compartment surgery. Bothersome frequency resolved in 82% after anterior/apical and in 63% after posterior compartment surgery. Bothersome UUI resolved in 75% after anterior/apical and in 61% after posterior compartment surgery. After surgery, symptom severity was comparable between groups. Bothersome de novo symptoms occurred in 1–3%. Conclusions OAB symptoms are more strongly related to anterior and apical than to posterior compartment prolapse, but improvement is seen after surgery for any vaginal compartment.


Introduction
Overactive bladder (OAB) symptoms are common, affecting around 13% of women of all ages. The prevalence of these symptoms increases with age, and they can have a detrimental impact on the quality of life [1]. Community-based studies show that OAB symptoms are up to six times more frequent among women with pelvic organ prolapse (POP) compared with ageadjusted women without POP [2]. OAB symptoms also improve after POP surgery, implying a connection between the two [2]. Nevertheless, the role of POP as an explanatory pathology behind OAB remains unclear, and current guidelines do not list POP in the diagnostic algorithms for OAB [3,4].
Proposed mechanisms to explain the co-occurrence of OAB symptoms and POP include detrusor overactivity due to (1) bladder outflow obstruction, (2) bladder wall distension and stimulation of stretch receptors, and (3) traction and opening of the urethra triggering the emptying reflex [2]. Based on these theories, it is plausible that anterior compartment prolapse (i.e., bladder involvement) has a greater impact on the OAB symptoms than posterior compartment prolapse. However, the majority of studies have not found any correlation among the degree of anterior, apical, or posterior compartment prolapse and severity of OAB symptoms [2]. Furthermore, studies comparing symptom improvement after prolapse surgery for different vaginal compartments conclude comparable improvement regardless of the repaired compartment [5][6][7][8]. This lack of association, together with imperfect symptom resolution, challenges the rationale to perform POP surgery to address OAB symptoms [9,10].
To understand the relationship between OAB symptoms and POP, we [1] quantified the association between the degree of individual POP compartments and OAB symptoms before surgery and [2] examined whether symptom improvement after surgery is dependent on the repaired compartment. We hypothesized that OAB symptoms are more strongly related to the anterior and apical than to the posterior compartment prolapse.

Setting and participants
We used data from the national, prospective Finnish Pelvic Organ Prolapse Surgery Survey (FINPOP). The study setting, population, and methods of surgery have been reported in more detail previously [11]. All Finnish hospitals performing POP surgery were invited to participate and to recruit all patients scheduled to undergo prolapse surgery during 2015. Women unable to communicate in Finnish or Swedish were excluded. A total of 41 of 45 hospitals (91%) performing POP surgery participated. The FINPOP cohort includes 3535 POP operations representing 83% of POP operations performed nationwide during 2015 (National database: Care Register for Health Care).
The population of this study includes 2933 operations with preoperative clinical examination and symptom questionnaires available. The patient flow, exclusion criteria, and data availability are shown in Fig. 1. We excluded women receiving a procedure for stress urinary incontinence concomitantly (N = 25) or during the follow-up (N = 84) from the analyses regarding symptom improvement. Of six women receiving intradetrusor injections of botulinum toxin A between 6 and 24 months' follow-up, two were excluded from 24 months' analyses because they reported improvement in the OAB symptom scores. Since usage of OAB medication at baseline did not associate with fewer OAB symptoms (rather the opposite), we did not exclude these women from the analyses.

Data collection
All data in this study were collected prospectively, and information was not retrieved from hospital charts.
The preoperative degree of POP was assessed by the surgeons using the simplified Pelvic Organ Prolapse Quantification (POP-Q) system as a single most distal point (in centimeters from the hymen) for each vaginal compartment (Ba for anterior; Bp for posterior, and C for apical compartment). The stage of POP was determined according to the POP-Q system [12]. Vaginal length was not recorded, and therefore stages 3-4 for all compartments, as well as stages 0-1 for apical prolapse, were combined in the analyses. The surgeons also recorded participants' surgical history and details on the operation performed. The surgeons entered the data in the electronic study registry in a standardized form.
The participants completed standardized, self-administered questionnaires at baseline and at 6 and 24 months after the surgery. This included information on their medical, surgical, and obstetric history. Pelvic floor dysfunction was assessed with a validated, condition-specific quality-of-life instrument, Pelvic Floor Distress Inventory −20 (PFDI-20) [13,14]. The follow-up questionnaires were mailed to the participants, collected on paper or electronic forms based on participant's preference, and entered in the electronic study registry.
Information on the anti-incontinence procedures during the follow-up were retrieved from the Care Register for Health Care (coverage > 95%) [15].

Outcome measures
OAB symptoms were evaluated using two items in the PFDI-20: Item 15, 'Do you usually experience frequent urination?', assessed urinary frequency; 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?', assessed urgency urinary incontinence [14]. The scale for each symptom is as follows: 0: symptom not present; 1: symptom present but not at all bothersome; 2: symptom somewhat bothersome; 3: symptom moderately bothersome; 4: symptom quite a bit bothersome. We defined answers 3 and 4 as bothersome symptoms. Bothersome symptom was defined as resolved when bother score at follow-up was < 3.

Statistical analyses
We categorized the population into two groups based on the repaired compartment: (1) women who had surgery for anterior and/or apical compartment (± posterior compartment), i.e., anterior/apical group, and (2) women who had surgery for posterior compartment only, i.e., posterior group.
To further explore differences between anterior and apical repairs, we performed a secondary analysis dividing the anterior/apical group into women with (1) a vaginal procedure for anterior wall but not any type of apical procedure (anterior group); (2) any type of apical procedure but no vaginal procedure for anterior vaginal wall (apical group); (3) a vaginal procedure for anterior wall and any type of apical procedure (anterior and apical group).
We used a generalized linear model (ordinal logistic) to estimate the association between the baseline anatomy (Ba, Bp, C in centimeters) and symptom bother (ordinal scale 0 to 4). Multivariable models were fitted to control for prolapse in other compartments (Ba, Bp, C) and to adjust for potential confounders. The confounders (age, BMI, parity, smoking, previous POP surgery, previous antiincontinence surgery) were selected based on the knowledge from previous literature and from clinical experience using directed acyclic graphs [16]. Spearman's correlation coefficients did not indicate strong collinearity between the variables (all < 0.4). The ordinal logistic model yields odds ratios (OR) with 95% confidence intervals (CI) for a higher bother score with a centimeter increase in the extent of prolapse.
To estimate the association between the site/compartment of surgery and improvement after surgery, we used ordinal generalized estimation equations adjusting for confounders. To assess whether anterior/apical repair improved symptoms more compared with posterior repair, time * repair group interaction was included in the model. We also performed a sensitivity analysis adjusting for concomitant posterior repair.
Estimated marginal means from separate models with continuous dependent variables (instead of ordinal) were used to plot graphs to illustrate the results.

Ethical aspects
The study followed the ethical standards for human experimentation established by the Declaration of Helsinki of 1964, revised in 2013. The Research Ethics Committee of the Northern Savo Hospital District approved the study on May 20, 2014 (reference number 5//2014), and each participating hospital granted an approval for conducting the study. All participants gave written consent.

Characteristics of the study population
The characteristics of the study population (N = 2933) are presented in Table 1. Of women with follow-up data (N = 2602), 2091 (81%) underwent surgery for the anterior and/or apical (± posterior) compartment and 478 (19%) for the posterior compartment only (Fig. 1). One hundred fifty-seven (6%)

Prevalence of OAB symptoms at baseline
At baseline, 2346 women (82%) reported at least one OAB symptom of any degree and 1135 (40%) at least one bothersome (bother score 3 or 4) OAB symptom ( Table 2). Altogether 1303 (46%) women presented with both urinary frequency and urgency urinary incontinence of any degree, and 484 (17%) presented with bothersome urinary frequency and urgency urinary incontinence.

Symptom relief after surgery
The severity of urinary frequency and urgency urinary incontinence decreased after surgery for all compartments (anterior, apical, anterior and apical, posterior) during the 6-month follow-up (p < 0.008 for all). At 24 months, symptom severity remained better compared with the baseline except for urgency urinary incontinence among women undergoing posterior repair only (p = 0.186 for posterior group and < 0.001 for other groups) (Fig. 3). Women undergoing anterior/apical compartment surgery had worse symptom severity at baseline (< 0.001) and experienced greater symptom improvement after surgery than women undergoing surgery for posterior compartment only (time*group interaction < 0.001). There was no difference in the symptom severity after the surgery between the two groups (p > 0.05 for all) (Fig.  3). A sensitivity analysis adjusting for concomitant posterior repair yielded similar results (data not shown).
The prevalence of bothersome urinary frequency and urgency urinary incontinence in the total population decreased from 0.31 to 0.08 [relative risk (RR) 0.27] and 0.26 to 0.10 (RR 0.40) during 6 months' follow-up, respectively ( Table 2).

Discussion
Principal findings OAB symptoms among women undergoing POP surgery are common (urinary frequency or urgency urinary incontinence of any and bothersome degree observed in 82% and 40% of women) and depend on the compartment and severity of prolapse. The symptoms had stronger associations with the anterior and apical compartment than posterior compartment  The degree of prolapse explained only a small proportion of variation in the symptom severity at baseline (e.g., 11% absolute increase in the prevalence of bothersome urinary frequency from no anterior wall prolapse to stage months' symptom severity on the right. The scale of symptom severity: 4: symptom quite a bit bothersome; 3: symptom moderately bothersome; 2: symptom somewhat bothersome; 1: symptom present but not at all bothersome; 0: symptom not present. The thickness of the arrow is proportional to percentage [3][4]. Nevertheless, the correction of anatomy relieved a significant proportion of OAB symptoms. Six months after anterior/apical compartment surgery, urinary frequency bother score had decreased in 77% and urgency urinary incontinence in 62% of women. Although the symptoms did not correlate with the degree of posterior wall prolapse at baseline, OAB symptoms' bother score still decreased after posterior repair in 50-57% of women. Bothersome de novo symptoms were uncommon (1-3% 6 months postoperatively).
A systematic review concludes that OAB symptoms improve after POP surgery [2]. However, it has remained unclear how the improvement relates to surgery for specific vaginal compartments. We identified only four previous studies comparing anterior to posterior involvement, and these studies could not consistently demonstrate a difference between anterior and posterior compartment surgery [5][6][7][8].
Weak association between the degree of prolapse and OAB symptoms as well as incomplete symptom improvement after surgery imply that other factors explain a large part of the variation in OAB symptoms among the POP population. OAB is a nonspecific, complex, and multifactorial symptom syndrome frequent in the general population, and among men as well [1,2]. There appear to be several distinct subtypes of OAB with multiple different mechanisms. The underlying factors can overlap and have convoluted interactions. Several potential pathophysiological factors, including metabolic syndrome, affective disorders, gastrointestinal functional disorders, sex hormone deficiency, urinary microbiota, and subclinical autonomic nervous system dysfunctions, have been suggested [30]. Since surgery does not address these potential underlying causes, POP surgery will not lead to resolution of OAB symptoms in all cases.

Clinical implications
Patients, as well as clinicians, often assume that POP causes their OAB symptoms. Our data show that while the symptoms are not explained solely by the distorted anatomy, women nevertheless have high probability of symptom improvement without any further intervention.
Based on our results, POP can be considered a contributing factor to OAB symptoms, and it should be evaluated for in the diagnostic workup of women with these symptoms.

Strengths and limitations
To the best of our knowledge, this is the largest study on the topic. A national cohort including all levels of hospital care increases the generalizability of the results. Further strengths include prospective data collection and use of a validated questionnaire. Unlike previous studies [5][6][7], we excluded women with anti-incontinence procedures. This is important since anti-incontinence surgery may independently affect OAB symptoms [2]. We also analyzed the outcomes on the actual measurement scale instead of dichotomization.
Our study has limitations. The population is comprised of women scheduled for POP surgery. This may lead to referral bias overestimating the effect. Second, we did not use a specific scale for OAB or collect data on urinary urgency and nocturia, two additional symptoms of OAB. Third, we did not obtain frequency volume charts or urodynamic studies and lack objective measures of OAB. Fourth, we did not collect data on OAB medication at follow-up. Finally, the observational nature of the study precludes drawing a definite causal relationship between POP and OAB.

Conclusion
Urinary frequency and urgency urinary incontinence are common among women with POP and more strongly related to anterior and apical than to posterior compartment prolapse. Substantial symptom improvement is seen after surgery for any vaginal compartment, and bothersome de novo symptoms are rare. Residual postoperative symptoms are likely explained by the multifactorial nature of OAB symptoms.