Introduction

Bladder outlet obstruction (BOO) is a known complication of any antiincontinence surgery, with estimates ranging between 2.3 and 22% [1–3]. A more accurate figure is probably reflected in the metaanalysis completed by the American Urological Association Stress Urinary Incontinence Clinical Guidelines Panel that reported the incidence of urinary retention for more than 4 weeks after sling surgery to be 8%, with the risk of permanent urinary retention generally not exceeding 5% [4].

Symptoms of BOO include obstructive voiding symptoms such as complete or partial urinary retention, hesitancy, positional voiding, double voiding, and/or irritative symptoms such as urgency, frequency, and urge incontinence (Table 1). Studies involving patients with BOO after antiincontinence surgery report that 60–84% of patients present with obstructive symptoms and 55–75% present with irritative symptoms [5–7]. BOO is probably underrecognized because irritative symptoms are often overlooked in the presence of normal emptying. Recurrent urinary tract infections may also be a presenting complaint.

Table 1 Symptoms of BOO
Full size table

Surgical treatment of BOO is generally via urethrolysis, with abdominal, vaginal, and combined approaches being utilized. The degree of urethral mobilization in these procedures can vary significantly along a spectrum that ranges from simple sling/suture incision with minimal urethral dissection to full circumferential mobilization of the urethra. The efficacy of urethrolysis procedures relieving the obstructive symptoms for this population of patients ranges between 50 and 100%, while irritative symptoms resolve in 2.3 to 100% of patients [2, 5–19]. Multiple authors have found resolution of urinary retention to be higher than complete resolution of irritative voiding symptoms [5, 16, 19]. Although the success of the procedure is generally good, treatment of patients who fail an initial procedure can be difficult, and there is scant literature support to guide the management of these patients. To our knowledge, there is only one published series with more than 20 patients, which examines the outcomes of repeat urethrolysis [16].

The purpose of our study was to determine our success in relieving irritative and/or obstructive voiding symptoms in our patients undergoing both primary and repeat transvaginal urethrolysis (TVU) procedures. Postoperative complications including de novo stress urinary incontinence (SUI) were also evaluated after these procedures.

Materials and methods

A retrospective review was conducted with all patients who underwent a TVU procedure for iatrogenic BOO by the senior author between 2001 and 2005 at a single institution. In all patients, complete history and physical examination and postvoid residual (PVR) determination were done. Multichannel videourodynamic studies (UDS) were routinely performed. An exception to this was made in several patients who were referred with primarily obstructive voiding symptoms that began immediately after an antiincontinence procedure, assuming they had a fixed urethra on examination, they had an elevated PVR, and they did not previously have a urethrolysis or sling incision/excision procedure performed. Cystoscopy was performed in situations where there was a high degree of suspicion for an intravesical or intraurethral foreign body. Preoperative symptoms of our patients included urinary urgency, frequency, urge incontinence, hesitancy, voiding difficulty, incomplete emptying, positional voiding, and urinary retention requiring intermittent self-catheterization. With the exception of one patient who had an acontractile detrusor, the urodynamic studies of our patients revealed relatively high-pressure, low-flow voiding phases, although strict urodynamic criteria were not solely used as the basis for recommending a urethrolysis procedure. A detrusor pressure of greater than 20 cm H2O at the point of maximal flow during the voiding phase, in association with a maximum flow of less than 12 ml/s, was the general parameter used to define the high-pressure, low-flow voiding. All patients were diagnosed with BOO based on a combination of the their symptoms, relationship to antiincontinence procedure and onset of symptoms, PVR, physical examination, and urodynamic studies. Patients with follow-up of less than 3 months and patients with an intravesical or urethral foreign body were excluded from this study.

Patients were categorized into three groups based upon the extent of the urethrolysis performed. All patients underwent a transvaginal procedure performed through a midline vaginal incision at the level of the bladder neck. Lateral dissection was then utilized to separate the vaginal epithelium from the paraurethral fascia to the level of the inferior pubic ramus [12]. Patients with this degree of dissection were designated as group 1. This situation was generally reserved for patients with BOO for less than 1 year after a sling procedure, and it also involved incision and/or excision of the suburethral portion of the sling. Group 2 consisted of extensive bilateral dissection into the retropubic space [7]. After the lateral dissection as described above, the endopelvic fascia was perforated with scissors to enter the retropubic space. Blunt dissection with the surgeon’s finger was used aggressively to free the urethra and bladder from all attachments to the under surface of the pubic bone. Any permanent sutures or sling material encountered was resected or sharply incised. Group 3, most often, but not exclusively, reserved for repeat procedures, involved adding a suprameatal incision to the procedure described for group 2, and then wrapping a Martius fat pad flap around the urethra [7]. In all patients, the extent of the dissection was continued until there was sufficient urethral mobility as demonstrated by posterior rotation of the urethra when gentle traction was applied to the catheter.

The preoperative and postoperative statuses of patients were compared for the following groups: (1) patients with preoperative obstructive voiding symptoms (cured/failed), (2) patients with preoperative catheter dependence (cured/failed), (3) patients with preoperative irritative symptoms (cured/partially cured/failed). New onset SUI was also noted. For patients with preoperative obstructive voiding symptoms, cure was defined as the resolution of these symptoms and normalization of PVR (< 100 ml). In the preoperative catheter dependence group, cure was defined as the lack of need for intermittent or indwelling catheter. For patients with preoperative irritative or OAB symptoms, cure was defined as complete resolution of irritative symptoms without the need for anticholinergic agents. Partial cure was defined as an improvement of irritative symptoms, and/or the use of anticholinergic agents to address irritative symptoms, while fail was defined as no change or worsening of OAB symptoms. Comparisons between the patients cured and failed (preoperative obstructive voiding symptoms and preoperative catheter dependence) or cured, partially cured, or failed (irritative symptoms) were made in terms of age, interval between antiincontinence procedure and urethrolysis, history of a prior urethrolysis, urodynamic parameters (detrusor pressure at maximum flow or maximum flow rate), preoperative PVR, length of follow-up, and type of urethrolysis procedure performed. Statistical analysis was performed using SAS System statistical software (SAS, Cary, NC, USA), using t, chi-square, and ANOVA tests where appropriate. The analysis was initially performed using data from all the patients in the study. A post hoc analysis was done using data from patients undergoing repeat procedures.

Results

A total of 55 women who underwent 61 TVU procedures between 2001 and 2005 were included in this analysis. Twenty-three of the patients had a prior urethrolysis procedure performed. One patient was excluded due to lack of follow-up. The mean age of the patients was 58 years. The mean time between the urethrolysis and the antiincontinence procedure was 33.9 months. Mean follow-up was 20.4 months (Table 2). The group distribution is as follows: groups 1–7, groups 2–31, and groups 3–23.

Table 2 List of patient variables for both all patients and those just having repeat procedures
Full size table

Obstructing antiincontinence procedures were identified as a bladder-neck sling in 21 patients, a retropubic urethropexy in 13, a synthetic midurethral sling in 14, needle-suspension procedures in 2, and unknown or other procedures in 5. Multichannel UDS were performed preoperatively in 49 of the 55 patients (all patients who underwent two procedures by the senior authors each had two UDS done). The average maximum flow was 7.6 ml/s (seven French catheter in place) with an average P det of 54 cm H2O at this flow. Detrusor overactivity was noted on the UDS of 28 patients.

The average PVR for our patients was 165 ml preoperatively vs 57 ml postoperatively, which was statistically different (p < 0.05). Preoperative presentation was both obstructive (either symptoms of PVR > 100 ml) and irritative in 32 patients, irritative only in 15 patients, and obstructive only in 14 patients. Determination of success for the procedure was evaluated separately for obstructive and irritative problems. Of the 46 patients with obstructive voiding preoperatively, 38 (83%) patients were considered cured based upon the resolution of obstructive voiding symptoms and normalization of PVR (< 100 ml). None of these patients required intermittent catheterization postoperatively (two of the patients who were not cured each had two procedures performed by the senior author; these patients are counted twice). Both the preoperative and postoperative PVRs were lower in the cured group, as well as the mean number of prior urethrolysis procedures performed in these patients (p < 0.05). No statistical differences were found between the groups in terms of age, interval between antiincontinence procedure and urethrolysis, urodynamic parameters (detrusor pressure at maximum flow or maximum flow rate), type of urethrolysis performed, type of prior antiincontinence procedure, or length of follow-up. Forty-two of these patients required intermittent self-catheterization preoperatively, and eight (19%) required this postoperatively (two patients who were catheter dependent both pre- and postoperatively underwent two procedures by the senior author; these patients are counted twice). Of the 47 patients with irritative symptoms preoperatively, 21 (45%) patients were completely cured of their symptoms, 16 (34%) were partially cured, and 10 (21%) considered their symptoms unchanged (failed). In comparing the patients with irritative symptoms by outcome groups, there were no differences in age, interval between antiincontinence procedure and urethrolysis, history of a prior urethrolysis, preoperative PVR, postoperative PVR, urodynamic parameters (detrusor pressure at maximum flow or maximum flow rate), or length of follow-up. Using ANOVA, the distribution of outcomes for resolution of irritative complaints differed based upon the type of urethrolysis performed (p < 0.05) (Table 3).

Table 3 Outcome groups compared in terms of preoperative obstructive symptoms (and/or including elevated PVR), preoperative catheter dependence, and preoperative irritative symptoms
Full size table

Of the 23 repeat procedures performed (21 patients), 5 patients had TVU procedures with extensive retropubic dissection (group 2), and 18 patients had the added suprameatal incision and Martius flap (group 3). These patients had undergone an average of 1.72 prior urethrolysis procedures. A post hoc analysis was done involving the data only from patients having a repeat procedure. Of the 18 patients with obstructive symptoms or clinical findings, 13 (72%) were cured and 5 (28%) failed. Both the preoperative and postoperative PVRs were less among those patients cured compared to those who had failed procedures (preoperative PVR, cured vs failed, 172 vs 355 ml, p < 0.05), (postoperative PVR, cured vs failed, 22 vs 79 ml, p < 0.05, respectively). None of the other parameters mentioned above was different between the groups. Nine (64%) of the 14 patients who were catheter dependent preoperatively were considered cured. Seventeen patients had irritative complaints before the repeat procedure. Ten (59%) were cured, six (35%) were partially cured, and one (7%) was left unchanged (failed). No differences were noted between the groups in terms of the analyzed variables (Table 3). The numbers were too low to appropriately compare differences in outcomes based upon the type of procedure performed.

Complications after this procedure were relatively low. Postoperatively, 16% (9/61) of the patients had SUI. Of these nine patients, three had mild SUI symptoms before the procedure and felt that the symptoms were unchanged postprocedure (one of these patients had two urethrolysis procedures and is counted twice), and one had SUI symptoms before the procedure but had worse symptoms postoperatively. Only three patients sought treatment for their SUI, with two being successfully treated with injectable agents and the third awaiting placement of a bladder-neck sling. The other patients felt that their symptoms were mild enough that they did not want additional treatment. The “degree” of urethral dissection did not seem to be associated with postoperative SUI, as two patients were from group 1 (2/7–29%), five patients from group 2 (5/31–16%), and two patients from group 3 (2/23–9%). Also of note, six of the patients with SUI underwent primary procedures, and three patients with SUI had a repeat procedure performed. No patients reported new onset OAB complaints. A small cystotomy occurred in two patients while performing the urethrolysis. In both patients, this was recognized and repaired, and it did not prevent the urethrolysis from being completed. Another patient developed a wound infection at the incision site on the labia from the Martius flap. This was managed with packing, and the incision closed by secondary intention.

Discussions

BOO is an important complication that should be recognizable by all surgeons who perform antiincontinence procedures. Our average interval between antiincontinence procedure and urethrolysis was 34 months, which exemplifies the fact that physicians often overlook this clinical condition. This is especially true among patients with primarily irritative symptoms, who present with a normal PVR, most probably because these patients get labeled as having de novo overactive bladder. One should have a high degree of suspicion for BOO in patients with de novo or significantly worse irritative voiding symptoms that persist beyond 3 months postoperatively, especially when the symptoms are not controlled with standard doses of anticholinergic medications. A urodynamic study is very helpful in evaluating patients that fall into this category.

The diagnosis of BOO should involve a compilation of the findings from the history, physical examination and PVR, as well as other studies that include multichannel videourodynamics and cystoscopy. Cystoscopy is essential in evaluating patients for whom one has a high degree of suspicion for a foreign body in the bladder or urethra. Patients may present with either obstructive or irritative symptoms in these situations. Although we generally use a voiding detrusor pressure at maximum flow above 20 cm H2O and a maximum flow rate below 12 ml/s as cutoffs for defining BOO in our practice, we recognize that there is no consensus as to what numbers best define BOO in the female patient [20, 21]. These urodynamic parameters must be placed in the context of the other findings. Furthermore, patients with an apparent acontractile or hypocontractile bladder, but with other signs of BOO, should not be excluded from being candidates for a urethrolysis procedure on the account of these numbers as we have been successful in relieving voiding dysfunction symptoms in some of these patients. A temporal relationship between the onset of symptoms and the antiincontinence procedure, along with a history of normal voiding before the procedure, should guide one’s decision to perform a urethrolysis.

Our finding of increased success in the resolution of obstructive complaints compared to irritative complaints is similar to that reported in other series [5, 16, 19]. One possibility is that some of the patients who were not cured of their irritative symptoms represent a group of patients with underlying OAB symptoms due to another etiology that preceded the antiincontinence procedure. It is reasonable to assume that relieving the obstruction in these patients would only improve, but not cure, these patients. Another possibility is that some of these patients truly had de novo urge incontinence, irrespective of any obstruction, after the surgery. Despite everything, even in patients undergoing repeat procedures, a sizable number of patients experienced an improvement in their irritative symptoms. Therefore, even if a patient has failed a prior urethrolysis procedure, it is very reasonable to recommend urethrolysis to patients with persistent BOO, who present with mostly irritative complaints. It is important to convey to these patients, however, that complete resolution of these symptoms may not feasible, although they are generally significantly improved. Before performing a repeat procedure, one should reevaluate the patient, taking care to rule out other etiologies for these symptoms, especially neurogenic processes. There should continue to be physical and/or urodynamic evidence of persistent obstruction as well. Also, one should take into account the type or previous procedures performed and the extent of prior dissection.

In examining outcomes in relation to preoperative parameters among our patients, we noted that the preoperative PVR was lower in patients cured of these symptoms compared to those considered failed. This difference was not noted when comparing irritative outcomes. This is consistent with the fact that many patients with BOO will present with only irritative symptoms, having normal PVRs and no obstructive voiding complaints. When comparing the outcomes of all patients with preoperative obstructive symptoms, the patients who were cured had undergone fewer prior urethrolysis procedures. This parameter was not statistically different in any of the other comparison groups, including those that involved only patients undergoing repeat procedures. Other authors have found a delay to urethrolysis, a higher PVR, or a prior failed procedure to be associated with less successful outcomes of the procedure [5, 13, 22]. Importantly, no finding appears to be consistently predictive of a lack of success and should not preclude surgical management.

It is interesting to note that when examining outcomes relative to the type of procedure performed, the distribution in the irritative symptom outcomes was statistically different. Although our numbers are relatively small, failure appeared to be most associated with group 2. We speculate that some of these patients redeveloped retropubic scarring resulting in persistent irritative symptoms. Additional studies with larger numbers of patients should be done to further evaluate this finding. There were no differences in outcomes between the types of urethrolysis procedures with regard to treatment of preoperative BOO symptoms/elevated PVR or preoperative catheter dependence.

Management of patients with a prior failed TVU can be challenging. Some authors recommend a retropubic approach in this scenario [8, 23]; however, we feel that a TVU allows for optimal access to the retropubic attachments in the region of the urethrovesical junction, allowing for circumferential mobility of the entire urethra if desired. In a series of 24 patients undergoing a repeat urethrolysis, the surgical approach, retropubic vs transvaginal, did not affect success of the procedure [16]. In contrast, a cohort study of 16 patients undergoing a primary urethrolysis for BOO after a Burch colposuspension found a retropubic approach to be more successful in relieving urethral obstruction compared to the vaginal approach [8]. We speculate that most urethrolysis failures are either due to incomplete detachment of the urethra from the posterior aspect of the pubic bone or recurrent scar formation after the procedure. Although antidotal, in patients who have undergone a prior aggressive urethrolysis, we recommend placing a fat pad between the urethra and the pubic symphysis to prevent recurrent scar tissue from forming between these structures. This can be readily accomplished transvaginally with a Martius flap. An omental flap may be utilized if a retropubic approach is used. Most importantly, our results show that a repeat procedure can be successful in improving both obstructive and irritative voiding complaints. Patients with persistent or recurrent BOO after a prior failed urethrolysis should be managed with a repeat urethrolysis, and a transvaginal approach is very reasonable.

New onset SUI was an infrequent complication in this series of patients. This is similar to the reported rates in most other urethrolysis series (without concomitant resuspension), which range between 13 and 19% [5, 7, 14, 15], although a rate of 47.1% was recently recounted in one recently published study [19]. In the previously published series of patients undergoing repeat urethrolysis, a rate of 18% was found [16]. Even among patients undergoing circumferential mobilization of the urethra, it was uncommon. It is interesting to note that this complication did not occur more frequently in patients undergoing more aggressive urethral mobilization nor was it more common in patients having repeat procedures. Among the patients with preoperative SUI, their symptoms did not necessarily worsen after the procedure. Some physicians have recommended that a concomitant procedure be done at the time of urethrolysis to resuspend the bladder neck [13]. With this relatively low rate of de novo SUI and the concern of performing a procedure with a potential to promote scarring and recurrent obstruction, we do not recommend this with either primary or repeat procedures nor do we recommend it in patients with mild SUI symptoms preoperatively [7]. Furthermore, urethral mobilization and dissection during a urethrolysis should not be minimized under the assumption that this would result in less SUI postoperatively, especially when this concern prevents one from adequately freeing the urethral and retropubic attachments.

We conclude from this review that an aggressive TVU can be very successful in relieving urinary retention due to iatrogenic BOO, even when a prior urethrolysis has been unsuccessful. Irritative symptoms are less often completely resolved but can be improved after the procedure to the extent that anticholinergic medications will adequately control the symptoms. Lastly, SUI is an infrequent complication, even after a very aggressive primary or repeat urethrolysis, but if it occurs, it can usually be managed with injectable agents.