**2. Results**

A total of 155 patients underwent their first-time urethral BoNT-A injection for their voiding dysfunction refractory to medical therapies. The patients included 80 with NNVD (22 men and 58 women, aged 66.6 ± 16.9 years) and 75 with neurogenic voiding dysfunction (NVD, 34 men and 41 women, aged 55.5 ± 19.7 years). Successful outcomes were reported in 92 (59.4%) patients, and a failed outcome was noted in 63 (40.6%). Table 1 lists the baseline patient demographics according to their voiding dysfunction subtypes, bladder dysfunction, and urethral sphincter dysfunctions between successful and failed subgroups. We found that the treatment outcome was not significantly different among different voiding dysfunction subtypes.


**Table 1.** Treatment outcome according to patient characteristics at baseline.

VD, voiding dysfunction; DO, detrusor overactivity; DU, detrusor underactivity; DHIC, detrusor hyperreflexia and inadequate contractility; HSB, hypersensitive bladder; DV, dysfunctional voiding; DSD, detrusor sphincter dyssynergia; PRES, poor relaxation of external sphincter.

When we compared the videourodynamic study (VUDS) characteristics between patients with successful and failed outcomes, only the baseline maximum flow rate (Qmax) and an open bladder neck during voiding showed a significant difference between groups (Table 2). In the 92 patients with a successful outcome, 89 (96.7%) had an open bladder neck, whereas of the 63 with failed outcomes, 54 (85.7%) had a tight bladder neck (*p* < 0.001). Patients with a successful outcome had a significantly higher Qmax than those with failed outcome (*p* = 0.031). However, this fact was only observed in patients without DU (7.82 ± 4.97 versus 2.0 ± 2.65 mL/s, *p* = 0.004), but not in patients with DU (5.05 ± 5.15 versus 6.0 ± 5.66 mL/s, *p* = 0.804).

**Table 2.** The video-urodynamic (VUDS) characteristics between patients with successful and failed treatment outcomes.


DO, detrusor overactivity; DU, detrusor underactivity; BN, bladder neck.

The treatment outcome was also not related to the bladder dysfunction between NVD and NNVD patients (Table 3). Further analysis of the VUDS parameters revealed that in patients with NNVD and non-DU, patients with successful outcomes had significantly higher voiding detrusor pressure (Pdet) than those with failed outcomes (*p* = 0.013), but that was not found in patients with NNVD and DU (*p* = 0.456). In patients with NVD and non-DU, those with successful outcomes had significantly lower first sensation of filling (FSF, *p* = 0.050), smaller cystometric bladder capacity (CBC, *p* = 0.010), higher Qmax (*p* = 0.044), and smaller PVR (*p* = 0.011) than patients with failed outcomes. In patients with

NVD and DV, those with successful outcomes (*n* = 28) had a significantly lower FSF and full sensation (FS) and smaller CBC (*p* = 0.008) than those with failed outcomes (*n* = 24). Patients of NVD and DSD and a successful outcome had a higher Qmax (*p* = 0.042) and smaller PVR (*p* = 0.021) than those who failed the treatment. Interestingly, significantly more patients with a successful outcome had an open bladder neck (BN) during VUDS than those with a failed outcome, in any subgroup of the bladder or urethral sphincter dysfunction. (Table 4)

**Table 3.** The video-urodynamic (VUDS) characteristics of voiding dysfunctions between patients with successful and failed treatment outcomes.


DU, detrusor underactivity; DV, dysfunctional voiding; DSD, detrusor sphincter dyssynergia; PRES, poor relaxation of external sphincter.

After urethral BoNT-A injection for treatment of voiding dysfunction, increased urinary incontinence was reported in 12 (13%) patients with successful outcomes, mostly occurring during sleep. De novo urinary tract infection was only observed in four (2.6%) patients overall.


FSF, first sensationresidual; BN, bladder neck; DO, detrusor overactivity; NVD, neurogenic voiding dysfunction; NNVD, non-neurogenic voiding dysfunction; DU, detrusor underactivity; DV, dysfunctional voiding;DSD,detrusorsphincterdyssynergia;NR,non-relaxingexternalsphincter;S,successfuloutcome;F,failedoutcome.

#### **3. Discussion**

The results of this study revealed that BoNT-A urethral sphincter injection can improve voiding efficiency in about 60% of patients regardless of neurogenic or non-neurogenic etiology. Preoperative VUDS provides an important prognostic indicator for successful treatment. Patients with NNVD who had higher voiding detrusor pressure and smaller PVR might benefit more from urethral sphincter BoNT-A injection than those with lower voiding pressure. Patients who were found to have a tight bladder neck during VUDS and the DU patients with a very low Qmax might have less favorable therapeutic outcomes. In addition, patients with non-DU NVD and reduced bladder sensation might not benefit from urethral sphincter BoNT-A injection.

The application of BoNT-A in urology started from urethral sphincter injections for the treatment of DSD in patients with SCI and MS [4]. After that, the treatment was extended to treat DO and urinary incontinence in NVD and NNVD patients [8,15]. Double-blind placebo-controlled studies of therapeutic efficacy of BoNT-A urethral sphincter injection have also confirmed the validity and durability of this treatment in patients with SCI and DSD [4,16]. A 50% reduction of the occurrence of urinary tract infection after urethral sphincter BoNT-A injections for DSD has also been reported in a meta-analysis [17]. Previous studies have shown that urethral sphincter injections with 100 to 200 U of BoNT-A were effective in patients with voiding dysfunction due to MS, CVA, or SCI [8,18]. Patients with chronic CVA and chronic urinary retention might be able to get rid of clean intermittent catheterization (CIC) after the urethral injection of 100 U of BoNT-A [8,19].

Several small cohort studies also confirmed the successful therapeutic results in patients with NNVD and voiding dysfunction at doses of 100 U or 50 U of BoNT-A [7,20,21]. Some patients with NNVD and DU could also have recovery of detrusor contractility after urethral sphincter BoNT-A injections and long lasting therapeutic effects [22]. However, until now, urethral BoNT-A injection for voiding dysfunction remained an off-label treatment. Although urethral BoNT-A injection can result in relaxation of the striated urethral sphincter [12], patients with DU and a tight bladder outlet might not have a successful treatment outcome because the bladder neck cannot open on micturition [14]. The treatment results between patients with NVD and NNVD, or between patients with different bladder contractility and urethral sphincter tonicity have not been compared. In addition, although urethral sphincter BoNT-A injection is effective in improving voiding efficiency (VE) and decreasing PVR, incomplete emptying remains a problem to be solved, and postoperative urinary incontinence is still another de novo issue for women with SCI or MS [8,18,23,24]. Under these considerations, patients might not be completely satisfied with the treatment outcome of urethral sphincter BoNT-A injection [23,25]. Therefore, patient selection is important for a successful treatment.

Urination is a complex interaction with appropriate coordination among the central and peripheral neural controls, sustained detrusor contraction, adequate bladder neck relaxation, and complete relaxation of the external sphincter and pelvic floor muscles. With one or more defects of those micturition mechanisms, patients may develop NVD or NNVD. Patients with DSD or DV usually cannot achieve efficient voiding due to hypertonicity of the external sphincter. Patients with DU who use abdominal pressure to void might have voiding difficulty and incomplete bladder emptying due to a tight bladder neck or a non-relaxing urethral sphincter or pelvic floor. Patients with detrusor hyperreflexia (DHIC) might have significant PVR because of low detrusor contractility without any anatomical BOO. In the era of BoNT-A, although theoretically the urethral sphincter BoNT-A injection might produce benefits by reducing bladder outlet resistance in voiding dysfunction, the unrealistic expectations of BoNT-A urethral sphincter injections often results in failed treatment in patients not suitable for this treatment [23].

In this study, we found that an open bladder neck during VUDS is essential for successful BoNT-A injection in patients with NVD or NNVD. A tight bladder neck during voiding is an unfavorable prognostic factor for the successful outcome of BoNT-A treatment. If the bladder neck cannot be opened, urine output will be inhibited at this gate. On the contrary, after transurethral incision of the bladder neck, patients with DSD or DU usually can urinate by abdominal straining or detrusor

contraction after the urethral BoNT-A injection. However, urinary incontinence might be a bothersome problem. Before urethral sphincter BoNT-A injection, patients should be informed about this potential adverse event. This adverse event is also the reason why many NVD patients finally select detrusor BoNT-A injections and periodic CIC for the solution of their voiding problem [23].

BoNT-A urethral sphincter injection results in decreased urethral pressure, increased Qmax, decreased PVR, and a reduction of autonomic dysreflexia in NVD patients with DSD due to SCI or MS [4,5,12,25]. The external sphincter hypertonicity might have different severity; therefore, injection of 100 U of BoNT-A might not be adequate for an efficient voiding in high grade DSD. In addition, because patients with NVD and DSD also have uninhibited DO, an increased urinary incontinence grade might develop after effective urethral sphincter BoNT-A injections [26].

NNVD due to DV is difficult to treat because the actual pathophysiology has not been elucidated and the only known LUTD is dysregulated urethral function with spastic or a non-relaxing external urethral sphincter during voiding [27]. DV results in voiding symptoms: Slow stream and large PVR. Therefore, attempts to reduce the hypertonicity or hyperactivity of the urethral sphincter by medication, and resume spontaneous voiding, always results in failure. It is also postulated that the psychologic voiding dysfunction due to anxiety or depression might be a cause of a low detrusor contractility and non-relaxing urethral sphincter through inhibiting detrusor contraction [28]. BoNT-A urethral sphincter injection can reduce urethral resistance but does not have an effect on the psychological insult; therefore, the successful outcome was only observed in 60% of patients with DV. If the bladder neck is not open during voiding, patients with DV might have more difficulty urinating.

DU may result from neurogenic, myogenic, obstructive, or idiopathic etiology. DU patients need to use abdominal pressure to void or depend on CIC. A sustained abdominal pressure is necessary to efficiently empty the bladder [29]. If the bladder outlet resistance is high at the level of bladder neck or urethral sphincter, patients need higher abdominal pressure to overcome the resistance, and therefore, they might not able to void efficiently. If the bladder neck is not open during voiding, urethral BoNT-A injections to the urethral sphincter might not be successful. Therefore, if VUDS shows a tight bladder neck, patients with DU and voiding dysfunction should undergo transurethral incision of the bladder neck first, otherwise, BoNT-A urethral sphincter injections might fail. In addition to an open BN, an adequate abdominal pressure is necessary for patients with DU who wish to void spontaneously after urethral BoNT-A injection.

Normal bladder sensation is another important factor for an efficient urination. The sensory afferents from the bladder urothelium and detrusor are essential parts of the voiding reflex circuit. Patients with DU might have reduced bladder sensations of filling and fullness [30]. The deficit of the bladder sensation will cause difficult initiation of voiding, as well as inefficient bladder emptying when the bladder has not been completely emptied [31]. Therefore, although the urethral resistance has been reduced after the BoNT-A injection, patients with DU and reduced bladder sensation might still have a poor outcome after treatment, especially in patients with NVD [32]. For these patients timed voiding and instruction of correct usage of abdominal pressure to void are mandatory after urethral sphincter BoNT-A injection in patients who have DU and reduced bladder sensation.

A limitation of this study is the mixed patient cohort. We aimed to find predictive factors for patients with different bladder and urethral sphincter dysfunctions. Interestingly, the success rates were similar between NVD and NNVD, and between different bladder or urethral sphincter dysfunctions. Treatment of voiding dysfunction is not an easy task. The possible causes of treatment failure have several different aspects [27]. Identification of the underlying causes of failure may improve the success rate after urethral sphincter BoNT-A injection. In this regard, VUDS before urethral BoNT-A treatment is mandatory to assess the vesicourethral dysfunction in patients with NVD and NNVD. Careful evaluation of the bladder neck opening and a higher Qmax at baseline may provide predictive value for a successful BoNT-A treatment outcome. However, urinary incontinence might be a de novo adverse event after urethral sphincter BoNT-A injection. Patients who are planning to undergo

urethral sphincter BoNT-A injection for voiding dysfunction should be fully informed of the limited therapeutic efficacy and the possible adverse event of urinary incontinence before treatment.
