Are we underestimating urinary retention following intravesical onabotulinumtoxinA in women with overactive bladder and a history of prior pelvic surgery?

1
INTRODUCTION
The international continence society's (ICS) definition of overactive bladder (OAB) is a disease process culminating in a complex of symptoms including urinary urgency associated with nocturia or urinary frequency plus or minus urge incontinence and in the absence of infective or anatomical pathology.
1
The negative impacts of OAB on health‐related quality of life as well as a significant economic burden are well documented.
2
,
3
First‐line OAB treatment consists of lifestyle modifications which if unsuccessful is followed by pharmaceutical anticholinergic agents. Third‐line treatment is intravesical OnabotulinumtoxinA (BTN/A) which is becoming a well‐used tool for management of overactive detrusor refractory to medical management, with its efficacy and safety profile well documented in the literature.
4
,
5
,
6
A commonly reported side effect following this procedure is urinary retention requiring intermittent catheterization (IC). Allergan®, the primary botox preparation used for this procedure, conducted a clinical trial that reports urinary retention requiring IC at approximately 6% for patients posttreatment of idiopathic detrusor overactivity.
4
Despite this, internationally, multiple studies report rates of IC notably higher, up to 43% in some populations.
6
,
7
,
8
This study aims to review the experience of a specialist urological surgeon to assess if current quoted rates of urinary retention are in line with prior experiences; as well as to identify preoperative predictors for retention to help guide future informed consent and patient management.

2
MATERIALS AND METHODS
2.1
Data collection
Barwon Health ethics committee ethics approval: Ref: QA/74001/VICBH‐2021‐253767(v1). Electronic medical records were collected for all female patients who underwent intravesical BTN/A for management of idiopathic detrusor overactivity. Patient data was retrospectively collected from February 2016 to March 2021. Exclusion criteria included: neurogenic bladder dysfunction; incomplete preoperative urodynamic assessment; previous history of IC; prior urinary retention; or lost to follow up. A secure deidentified database was generated to store patient information.
The following patient demographic information was collected: patient age; significant past medical history including neurological conditions, diabetes, prior pelvic radiotherapy, preoperative urinary retention, prior urethral sling, prior vaginal prolapse surgery, prior sacral nerve stimulator (SNS), urethral stenosis, and evidence of positive preoperative urinary tract infection (UTI) and any evidence of treatment. Urodynamic assessment results were collected including preoperative postvoid residual (PVR), maximum detrusor pressure at peak flow during urodynamics (pDet@Qmax), evidence of obstruction on urodynamics, bladder capacity, and urodynamic proven detrusor overactivity (DO).
Outcome measures included urinary retention defined as any patient requiring initiation of IC postoperatively, and postoperative PVR.

2.2
Surgical method
All patients had a preoperative urine microculture completed 1 week preoperatively. If a positive result was identified, they received a minimum of 3 days of appropriate sensitive antibiotics before the procedure. Each patient had 100 units of BTN/A diluted in 30 ml of normal saline, delivered to the bladder in 1 ml aliquots sparing the trigone. Patients were seen 2–3 weeks postoperatively and a postvoid bladder scan was completed. Decision to initiate IC was made on a case‐by‐case basis by the urologist, who has a special interest in functional urology, dependent on postoperative voiding volumes and PVR, and clinical concern for poor effective bladder emptying.

2.3
Statistical analysis
Patient demographic and clinical characteristics are presented. Continuous variables are presented as means and standard deviations for normally distributed data, or medians and interquartile ranges if skewed. Respectively, Student t tests or Mann–Whitney–Wilcoxon tests are used to compare characteristics between patients who required IC and those who did not. Categorical variables are presented as numbers and percentages, using χ
2 tests of comparison. A p < 0.05 is considered significant.
The associations between requirement of postoperative intermittent catheterization and age, preoperative PVR, prior sling, prior vaginal prolapse surgery, preoperative UTI, maximum detrusor pressure at peak flow during urodynamics (pDet@Qmax), detrusor overactivity, and bladder capacity (10 ml increments) were assessed with univariate, and multiple logistic regression models.
Proportions of patients requiring IC are presented for each clinically standard category of postoperative PVR volume.
A subsequent comparison of urodynamics assessment for patients presenting with prior sling is presented, using χ
2
, Student t tests or Mann–Whitney U tests as appropriate.
All calculations were performed in Stata version 17.0.
9

3
RESULTS
Data were collected for 106 patients. Of these, 12 were excluded from analyses; two patients were found to be neurogenic, five had missing or incomplete preoperative video urodynamics assessment, one had IC before Botox, one had an episode of preoperative urinary retention and three were missing outcome data. Thus, 94 patients form the analysis data set.
Demographics and preoperative characteristics of patients (total, and grouped by required intermittent catheterization) are presented in Table 1.
In total, 34 (36%) patients in the sample required IC to treat urinary retention, 2–3 weeks postoperatively. The average age was 69.7 years (SD 17.2). There were higher percentages of patients with preoperative UTI, prior sling, and prior vaginal prolapse surgery amongst patients requiring IC, compared to patients who did not require IC. Five patients had both prior sling and prior vaginal prolapse surgery, three of whom showed postoperative urinary retention. Twenty‐nine percent (n = 10) of all patients requiring IC had a preoperative UTI. Thirty‐two patients (94%) requiring IC had a preop PVR < 101 ml (median 0 ml [interquartile range [IQR]] 0–20 ml) with no evidence of a difference in preoperative PVR between patients requiring IC and those who did not. We note that only four (4%) patients showed elevated preoperative PVR (>100 ml) in our sample (range 120–242 ml); two of whom required postoperative IC (preoperative PVR 130 and 242 ml).
Table 2 shows the results of univariate and multivariable logistic regression analyses assessing the relationship between clinical factors thought to predict urinary retention and the need for IC.
At the univariate level, there was strong evidence of an association between requirement of IC and preoperative UTI, prior sling, and prior vaginal prolapse surgery. Patients with preoperative UTI had over four times the odds of requiring IC than patients with no UTI (OR: 4.25, 95% CI: 1.31–13.08, p = 0.016). It is worth noting that all patients were treated with appropriate antibiotic therapy before their procedure.
In a multivariable model assessing all risk factors for urinary retention, only prior sling remained a significant predictor of IC (OR: 7.35, 95% CI: 1.59–34.11, p = 0.011), however, numbers in our final model were reduced due to data loss (n = 76 with complete data available) so associations would have been difficult to detect. Odds of requiring IC remained close to three times that for prior vaginal prolapse surgery and preoperative UTI compared to patients without these conditions, however, did not meet the level of statistical significance.
In a subsequent subset analysis of prior sling patients, we found no statistical evidence of differences in preoperative urodynamics assessments or preoperative PVR volumes between the 11 patients who required IC, and the 5 patients that did not (Table 3). All four of the prior sling patients presenting with a preoperative UTI required IC.
Greater percentages of patients were initiated on IC as postoperative PVR volume increased (Table 4). The median postoperative PVR of patients initiated on IC was 387 ml (IQR: 306–537) compared to 54 ml (19–149) in patients not requiring IC. Of the 34 patients requiring IC, 26 (76%) had a postoperative PVR greater than 300 ml (range: 306–1139 ml). One patient who had a relatively low postoperative PVR of 147 ml was commenced on IC as had poor effective voids, only passing 20–50 ml at a time. One patient with a postoperative PVR > 400 ml was encouraged to commence IC, however, due to patient reluctance did not commence and was placed on prophylactic antibiotics. They were followed up again 2 weeks later for clinical review with repeat assessment showing PVR < 50 ml.

4
DISCUSSION
The symptoms of bladder overactivity are impacting over 16% of our population,
10
however when pharmacotherapy options fail, intravesical BTN/A has been shown to be a safe and effective treatment in multiple clinical trials.
4
,
5
However these trials have not included key patient populations commonly seen in consulting rooms including prior bladder or pelvic surgeries. Accurate assessment and recognition of patients most at risk for urinary retention and need for intermittent catheterization are still to be identified. Clinical trials have previously reported IC rates of 6%,
4
however we report a larger rate of IC initiation of 36%. A key difference in our study is the inclusion of patients which were representative of what we believe to be a “typical” population for an Australian urologist, including typical management challenges of preoperative workup, such as identification of asymptomatic UTI needing treatment. The odds of requiring IC was seven times larger for patients with prior mid‐urethral sling surgeries, three times larger for patients found to have prior vaginal prolapse surgery, and three times larger for patients found to have a positive preoperative UTI despite appropriate antimicrobial therapy, a common management challenge for the urological surgeon. Perhaps the asymptomatic bacteriuria may be indicative of mild voiding dysfunction.
Mixed stress and urge incontinence are common with one in three patients with urinary incontinence believed to have mixed aetiology,
11
as such we need to consider inclusion of patients with prior continence surgery when looking to assess rates of adverse outcomes post intravesical BTN/A. Although identifying that these patients were at an increased risk of IC in our study, we could not conclude that there was any evidence of a difference in any of the urodynamics assessment factors (including exclusion of any obvious urethral obstruction) in the patients with prior mid‐urethral sling operations, although numbers are small so we acknowledge differences will be hard to detect.
Although prior studies by Osborn et al.
7
have illustrated that increased preoperative PVR was a predictor of postoperative urinary retention, we did not find and cannot conclude similar findings from our sample. We postulate this to be indicative of a small number of patients with preoperative PVR > 50 ml (n = 10) within our sample rather than a true effect. This is demonstrated in Table 1. We suggest that an explanation for this is that this surgeon regularly uses SNS devices and that perhaps patients who were found to have higher preoperative PVR would have been offered SNS as a surgical alternative, or perhaps these patients had already been commenced on IC and did not meet inclusion criteria.
Despite being limited by retrospective methodology, these results are in line with internationally recognized studies with similar rates of intermittent catheterization post 100 units BTN/A,
7
which are up to six times those reported from Chappel et al.
4
It is difficult to determine if Chappel et al. trial excluded patients with positive preoperative urine microculture results, but they did exclude key patients who had prior sling, prior vaginal prolapse surgery. If all patients with these clinical characteristics were excluded from our data set, the overall rate of IC would be 22% (n = 13). We note that in our setting, the decision to initiate IC is based on a case‐by‐case patient assessment, looking at clinical need and shared decision making processes which includes but is not defined by limits of postoperative PVR as in the previously mentioned clinical trial by Chappel et al.
Patients who undergo intravesical botox injections for idiopathic overactive bladder (iOAB) would benefit from a greater understanding of the risk of urinary retention and IC when consenting to intravesical BNT/A for iOAB. Further studies should look to larger cohorts, preferably with multisite participation and analysis to evaluate if this trend is more prevalent within urological practice, with the inclusion of patients representative of complex urological history.
We acknowledge that the retrospective nature of our data collection is a limitation and that our review was limited by a single surgeon, with a follow‐up time that was shorter than comparable studies.
7
Small study numbers may have impacted the strength of evidence in our findings.

5
CONCLUSION
Prior urethral sling surgery was found to be associated with a higher rate of initiation of intermittent self‐catheterization with weaker evidence of association between IC and preoperative UTI, and prior vaginal prolapse surgery. The rate of IC initiation of 36% was higher than reported in prior clinical trials. The inclusion of patients with elevated preoperative PVR, and preoperative asymptomatic urinary tract infection may have explained differences in rates of IC initiation but mores closely reflects a typical population seen in the clinical setting.

AUTHOR CONTRIBUTIONS

Angela Holmes: Data curation, Formal analysis, Investigation, Project administration, Visualisation, Writing – original draft, Writing – review and editing. Ellen Kelsey: Data curation, Formal Analysis, Writing – original draft. Amy Wombwell: Data curation. Charlotte Molesworth: Formal Analysis, Data curation. Kathryn McLeod: Conceptualization, Project administration, Supervision, Writing – review and editing.

CONFLICT OF INTEREST
The authors declare no conflict of interest.

ETHICS STATEMENT
Approved by local ethics governance; Research Ethics, Governance and Integrity Unit (REGI) Barwon Health, Geelong, Vic, Australia, Ref: QA/74001/VICBH‐2021‐253767(v1). Not required as per our ethics department as low impact research, with only retrospective file review and nil impact on patient planned care.