Does Reduction of Number of Intradetrusor Injection Sites of aboBoNTA (Dysport®) Impact Efficacy and Safety in a Rat Model of Neurogenic Detrusor Overactivity?

Intradetrusor injections of Botulinum toxin A—currently onabotulinumtoxinA—is registered as a second-line treatment to treat neurogenic detrusor overactivity (NDO). The common clinical practice is 30 × 1 mL injections in the detrusor; however, protocols remain variable and standardization is warranted. The effect of reducing the number of injection sites of Dysport® abobotulinumtoxinA (aboBoNTA) was assessed in the spinal cord-injured rat (SCI). Nineteen days post-spinalization, female rats received intradetrusor injections of saline or aboBoNTA 22.5 U distributed among four or eight sites. Two days after injection, continuous cystometry was performed in conscious rats. Efficacy of aboBoNTA 22.5 U was assessed versus aggregated saline groups on clinically-relevant parameters: maximal pressure, bladder capacity, compliance, voiding efficiency, as well as amplitude, frequency, and volume threshold for nonvoiding contractions (NVC). AboBoNTA 22.5 U significantly decreased maximal pressure, without affecting voiding efficiency. Injected in four sites, aboBoNTA significantly increased bladder capacity and compliance while only the latter when in eight sites. AboBoNTA significantly reduced NVC frequency and amplitude. This preclinical investigation showed similar inhibiting effects of aboBoNTA despite the number of sites reduction. Further studies are warranted to optimize dosing schemes to improve the risk-benefit ratio of BoNTA-based treatment modalities for NDO and further idiopathic overactive bladder.


Introduction
Intradetrusor injections of Botulinum toxin A (BoNTA), i.e., onabotulinumtoxin A (onaBoNTA), first tested in spinal cord-injured (SCI) patients in 2000 [1], is approved as of 2011 by the US Food and Drug Administration and European Union as a second-line treatment to treat neurogenic detrusor overactivity (NDO) in humans.
The majority of clinical published data relates to onaBoNTA (Botox ® ) and, to a lesser extent, to abobotulinumtoxinA (aboBoNTA, Dysport ® ). The potency of these two formulations is different and doses cannot be interchanged. Nonetheless, intradetrusor injections in NDO patients of either formulations result in an increase in maximal bladder capacity and compliance, associated with decreased maximum detrusor voiding and filling pressures, and attenuation of uninhibited detrusor contractions (UDC) [10][11][12][13][14][15][16][17].
Various intradetrusor injections protocols have been described differing in dosage, dilution, as well as in number and location of injection sites. It is now of importance to harmonize clinical procedures in order to optimize dosing schemes. Several questions remain to be answered, especially the influence of variation of injection sites on urodynamic outcomes. This study is the first preclinical investigation evaluating the efficacy of aboBoNTA intradetrusor injections when reducing the number of injection sites. This comparison was performed using clinically relevant urodynamic parameters assessed by continuous cystometry in the SCI rat model [18,19], displaying a voiding pattern comparable to that of neurogenic patients [20], that has already been validated in-house for investigations on BoNTA [21].

Effect of Intradetrusor Injections of aboBoNT-A 22.5U in Four or Eight Sites in Rats 19 Days after SCI-Induced NDO
As data were not significantly different for all urodynamic parameters, S-4 sites and S-8 sites groups were aggregated (ns, p > 0.05; Table 1). AboBoNTA 22.5 U administered in either four or eight sites significantly decreased maximal pressure (p < 0.001 for both sites number; Figure 1C,D and Table 1) when compared with aggregated controls ( Figure 1A,B and Table 1) without affecting voiding efficiency (p > 0.05 for both; Figure 1C,D and Table 1). When injected in four sites, aboBoNTA 22.5 U significantly increased the infused volume, an index of bladder capacity, and compliance (p < 0.01; Figure 1C,D and Table 1), this latter being also improved with injection in eight sites albeit not significantly (p = 0.0584; Figure 1D and Table 1).
AboBoNTA 22.5U significantly decreased the amplitude of NVC when injected in four and eight sites (p < 0.05 and p < 0.01 respectively; Figure 1C,D and Table 1). NVCs' frequency was also decreased, although not significantly for both four and eight sites (p = 0.0525 and p = 0.0772, respectively; Figure 1C,D and Table 1). No effect was observed on the volume threshold (p > 0.05; Figure 1C,D and Table 1).  AboBoNTA 22.5U significantly decreased the amplitude of NVC when injected in four and eight sites (p < 0.05 and p < 0.01 respectively; Figure 1C,D and Table 1). NVCs' frequency was also decreased, although not significantly for both four and eight sites (p = 0.0525 and p = 0.0772, respectively; Figure 1C,D and Table 1). No effect was observed on the volume threshold (p > 0.05; Figure 1C,D and Table 1).

Discussion
This is the first preclinical investigation comparing the effect of aboBoNTA intradetrusor injections when reducing the number of sites for injection in the SCI rat model. AboBoNTA 22.5 U significantly and similarly lowered VC maximal pressure and improved compliance of the bladder, albeit not significantly when injected in eight sites. Moreover, bladder capacity was increased with four injection sites. Whatever the number of sites, the amplitude of NVC was significantly decreased. In addition, frequency of NVC was decreased although not significantly when injected in four or eight sites. Finally, a similar BWL was observed post-intradetrusor injection whatever the number of injection sites.
As the bulk of available evidence on the efficacy of BoNTA to treat NDO increases, the need to determine an optimal dosing and administration scheme, both cost-effective and patient-oriented, becomes more and more meaningful. These experiments were performed in a preclinical model undergoing a pathophysiological process comparable to that occurring in SCI patients and characterized by the emergence of NDO, thereby mimicking at best the voiding pattern of neurogenic patients [20]. Indeed, the SCI rat displays increased intravesical maximal pressure and occurrence of NVC during the bladder filling phase, the latter ensuing from afferent C-fibers recruitment [22]. Main considerations for the treatment of NDO are maintenance of low bladder pressures to prevent development of upper urinary tract complications and suppression of UDC, in order to improve bladder capacity, as well as to decrease urinary frequency and urgency. Therefore, the study design fit as closely as possible to the clinical BoNTA injection procedure and urodynamic parameters were selected to best characterize those clinically relevant in neurogenic patients ( Table 1). Dosing of aboBoNTA was chosen according to the previous dose-escalating study i.e., at a ratio 3:1 of the 7.5 U minimal effective dose (MED) of onaBoNTA determined in the same experimental paradigm [21]. Indeed, although they are not strictly speaking interchangeable, a 3:1 to 6:1 ratio of aboBoNTA:onaBoNTA is commonly reported to yield similar therapeutic effects [23,24]. Hence, intradetrusor injections of aboBoNTA delivering 22.5 U in total were performed in four or eight sites (1.8 and 0.9 U/µL per site, respectively).
Similarly to what is reported in NDO patients, aboBoNTA 22.5 U intradetrusor injections in SCI rats demonstrated an effect on both efferent and afferent neural pathways. Indeed, BoNTA is known to cleave synaptosomal-associated protein-molecular weight 25 kilodalton (SNAP-25) protein of the soluble N-ethylmaleimide sensitive factor attachment protein receptor (SNARE) complex involved in ACh exocytosis at the neuro-muscular junction [25]. Inhibition of ACh release at the presynaptic nerve terminals accounts for the effect of BoNTA on bladder efferent signalling, with a reduction of VC maximal pressure, as already observed in this model [2]. This also matches the decrease of maximal pressure at contraction obtained in BoNTA-treated NDO patients [10][11][12]14,17]. This decrease can sometimes increase post-void residual volume [10], meaning urinary retention and subsequent risk of urinary tract infection in BoNTA-treated patients [26]. However, such an effect on the post-void residual volume was not found since voiding efficiency remained unchanged. Furthermore, apart from its direct action at the presynaptic level, BoNTA also modulates bladder afferents that also express SNAP-25 [27]. BoNTA has been demonstrated to affect the release of other neurotransmitters in bladder tissue such as ATP, substance P, calcitonine gene-relatide peptide (CGRP), and glutamate [3,6,27]. An additional impact of BoNTA on bladder afferent pathways is the decrease of sensory receptor expression levels like P2X3 or the transient receptor potential vanilloid subfamily 1 (TRPV1) [28], as well as NGF excretion [29,30]. Here, the attenuation of NVC, as well as the increased bladder compliance and capacity, are in accordance with the inhibition of bladder sensory afferent signalling by BoNTA in SCI rats [31]. Indeed, reduction of NVC following onaBoNTA intradetrusor injections has already been reported [2,19], as well as a decrease in frequency and amplitude of NVC following aboBoNTA treatment [21], although nature and concentration of injected compound were different from the present study. Similarly, BoNTA injections attenuate UDC in neurogenic patients, with their pressure at first involuntary contraction being decreased [14,30] and detrusor volume at first contraction increased [13,30]. Improvement of bladder storage capacity [10][11][12]14,17] and compliance [14] has also been reported in SCI patients after BoNTA treatment. Thus, all in all, the efficacy on clinically-relevant urodynamic parameters is similar for four and eight sites of injection at 22.5 U total dose of aboBoNTA in the SCI rat model.
From a clinical point of view, higher local diffusion of toxin is awaited with higher total volume and number of injection sites i.e., lower concentrated BoNTA at each site, as reported in the unique clinical study on toxin distribution by magnet resonance imaging after injection in the bladder wall [32]. Likewise, when functionally studying toxin diffusion in the bladder of guinea pigs, a single injection of BoNTA 2U in 20 µL of saline induces increased spread in bladder from the injection site than in 2 µL [33]. However, since the first BoNTA injection description for NDO treatment [1], injections from 10 to 50 sites were tested, and the only one direct comparative study reported for 30 versus 10 injection sites concluded that lowering the number of injection sites did not impair efficacy [34]. As in the present study, where the main clinically-relevant parameters (maximal pressure and NVC) were similarly impacted with BoNTA in four or eight injections sites. An additional improvement of infused volume (index of bladder capacity) and compliance was even observed with four injection sites. Thus, BoNTA diffusion, and associated efficacy on NDO, was not impaired in the rat bladder. Concordantly, Liao et al., recently observed that similar therapeutics effects were produced by injection at 10, 20, and 40 sites in patients with refractory detrusor overactivity [35]. Finally, study results of a phase IIa clinical trial with NDO patients (ClinicalTrials.gov Identifier: NCT01357980) revealed a reduction of urinary incontinence episodes, a main symptomatic features investigated in the clinic, as well as of maximum detrusor pressure, and an increase of maximum cystometric capacity with aboBoNTA 750 U injected in 15 or 30 sites.
Diminishing the number of sites is still a matter of debate as, despite allowing a faster and consequent less painful injection procedure, it implies a careful investigation of safety aspects to avoid deleterious consequences. Indeed, with fewer injected sites i.e., greater concentrations per site, the risk for greater distal and/or systemic effects is increased due to unwanted toxin spread out of the bladder wall. Indeed, adverse events (AE) are closely related to toxin propensity to spread, diffuse and migrate, along with dilution and total volume injected [36]. Investigations of the link between the injected concentration per site, BoNTA distribution beyond its original injected site, and undesired side-effects are scarce. The only direct clinical comparative study reported that reducing the number of injection sites from 30 to 10 did not impair safety [34]. However, injection of the same dosing with a lower number of injection sites, meaning a higher concentration per injection site as in the present study in SCI rat, could cause a higher toxin leakage out of the detrusor [32]. In the current study, both treatment regimen affected rats' body weight to the same extent, although one death was reported for the rats treated with aboBoNTA 22.5 U at four sites of injection. Systemic AE after BoNTA injection have already been measured by body weight changes in mice [37] and such an effect on body weight has already been reported with onaBoNTA in rats, although in a different model and with a different treatment paradigm [38]. Additionally, in clinical studies, decreasing concentration per site appears to play a role in the prevention of systemic side effects, with no more side-effects reported, either by reducing the total dose [39] or by increasing the number of injection sites [40]. Thus, an excess of toxin at each injection site might spread, thereby increasing AE [41]. Furthermore, a technical incident at one site could further increment deleterious consequences for patients. For example, one patient was reported with 40% of toxin found out of the detrusor due to penetration depth hitch for one out of 10 injections [32].
Drawing a parallel between clinical and preclinical results points out their similarities and further supports the use of the SCI model for evaluation of the effect of BoNTA intradetrusor injections [18,19,21]. One can argue that the increase in the NVC volume threshold previously observed with aboBoNTA 10U injected in eight sites was not reproduced, albeit using a higher dose. However, the nature of the product was different and the prior study was a dose-response-based investigation therefore built on a different statistical analysis [21]. The only limitation of the SCI rat model is that urgency and urinary incontinence, main symptomatic clinical features investigated, are not quantified in animal preclinical studies. Although the main aim of this study was to compare the therapeutic effect of aboBoNTA when reducing number of injection sites in the bladder, additional experimental endpoints could provide further indications as to the safety aspects. Furthermore, BWL might have been the consequence of a decrease in food and water intake due to a BoNTA-induced transient muscular weakness as it was already reported in patients [42]. Thus, monitoring those consumptions and evaluation of neuromuscular function by measurement of grip strength could also be informative, as well as harvesting of bladder, followed by histological assessment.

Conclusions
This preclinical study focused on a key aspect of the BoNTA injection procedure (i.e., concentration and sites number) and is a first step for standardization of clinical protocols. The main result of the present study is that reduction by half of the number of sites for aboBoNTA injection did not hamper urodynamic improvements awaited in the SCI rat model. However, a higher number of sites might be more conservative from a safety perspective. Thus, additional data on the safety aspects are warranted to determine whether decreasing the number of sites allows an improvement of the benefit-risk ratio.

Experimental Section
All procedures were performed in compliance with the legislation on the use of laboratory animals (NIH publication N 85-23, revised 1996) and Animal Care Regulations in force in France as of 1988 (authorization from competent French Ministry of Agriculture-Agreement No. A78-322-3, 2013).

Saline or aboBoNTA Intradetrusor Injections
At 19 days post-spinalization, i.e., when the NDO is well-established, bladders were exposed under isoflurane anesthesia (1.5%-2.0%) and emptied. Using a microscope, aboBoNTA or saline was injected into the detrusor sparing the trigone with a 30-gauge needle connected to a microdialysis pump (1 µL/min; CMA/102; Microdialysis AB, Solna, Sweden) [18,19]. A total volume of 12.5 µL was evenly distributed in four divided sites for the S-4 sites and aboBoNTA-4 sites (5.6 U in 3.125 µL i.e., 1.8 U/µL per site) groups and a total volume of 25 µL in eight divided sites for S-8 sites and aboBoNTA-8 sites (2.8 U in 3.125 µL i.e., 0.9 U/µL per site) groups (Scheme 1). A PE-50 catheter was inserted within the bladder dome, tunnelled subcutaneously, exteriorized at the back of the neck, and sutured between the scapulae. Postoperatively, rats were treated with 10 mg/kg gentamicin.

Cystometry Experiments in Conscious Spinal-Cord-Injured Rats
Cystometry was performed in conscious rats 48 h after intradetrusor injections, as previously described [21]. Clinically-relevant voiding and filling parameters were analyzed: maximal pressure of voiding contractions (VC) (millimetre of mercury, mmHg; 1 mmHg corresponding to 1.36 cm•H2O), infused volume (microliters, μL) as an index of bladder capacity, compliance (ml per mmHg) and voiding efficiency (%, ratio of voided volume to infused volume), as residual volume is not measurable during continuous cystometry. Clinically-relevant parameters characterizing bladder nonvoiding contractions (NVC), defined as contractions with an amplitude of ≥ 3 mmHg during the filling phase, were analyzed: amplitude of NVC (mmHg), frequency of NVC (number per min) and volume threshold to elicit NVC (% of total bladder-filling volume), which is the preclinical equivalent to the volume at first involuntary detrusor contraction in clinical urodynamic investigations [10,41].

Statistical Analysis
All data were expressed as mean plus or minus the standard error of the mean and averaged per treatment group. A Grubbs' test was used for exclusion of outliers. An unpaired Student's t-test was used for the comparison of body weight (GraphPad Prism ® 5.04, La Jolla, CA, USA, 2010). The area under curve (AUC) of BWL was compared between aboBoNTA-4 sites or aboBoNTA-8 sites groups for homogeneity with Fisher's test and for significance with Student's t-test (SAS ® USA version 9.2, Cary, NC, USA, 2009). For each cystometry parameter, an analysis of variance was performed to compare (1) the S-4 sites and S-8 sites saline groups to allow their aggregation, and (2) the aboBoNTA-4 sites or aboBoNTA-8 sites groups versus the aggregated saline groups (SAS ® ). p values < 0.05 were considered significant.

Drugs and Chemicals
AboBoNTA (Batch E194441) was provided by Ipsen Biopharm Ltd., Wrexham, UK. Antibiotics (except gentamicin) and anaesthetics were purchased from Centravet (Dinan, France) and Roche Pharma (Neuilly-sur-Seine, France). Gentamicin was purchased from Merck and Co. Inc. (Whitehouse Station, NJ, USA). All other drugs and chemicals were purchased from Sigma-Aldrich (Lyon, France).

Cystometry Experiments in Conscious Spinal-Cord-Injured Rats
Cystometry was performed in conscious rats 48 h after intradetrusor injections, as previously described [21]. Clinically-relevant voiding and filling parameters were analyzed: maximal pressure of voiding contractions (VC) (millimetre of mercury, mmHg; 1 mmHg corresponding to 1.36 cm¨H 2 O), infused volume (microliters, µL) as an index of bladder capacity, compliance (mL per mmHg) and voiding efficiency (%, ratio of voided volume to infused volume), as residual volume is not measurable during continuous cystometry. Clinically-relevant parameters characterizing bladder nonvoiding contractions (NVC), defined as contractions with an amplitude of ě 3 mmHg during the filling phase, were analyzed: amplitude of NVC (mmHg), frequency of NVC (number per min) and volume threshold to elicit NVC (% of total bladder-filling volume), which is the preclinical equivalent to the volume at first involuntary detrusor contraction in clinical urodynamic investigations [10,41].

Statistical Analysis
All data were expressed as mean plus or minus the standard error of the mean and averaged per treatment group. A Grubbs' test was used for exclusion of outliers. An unpaired Student's t-test was used for the comparison of body weight (GraphPad Prism ® 5.04, La Jolla, CA, USA, 2010). The area under curve (AUC) of BWL was compared between aboBoNTA-4 sites or aboBoNTA-8 sites groups for homogeneity with Fisher's test and for significance with Student's t-test (SAS ® USA version 9.2, Cary, NC, USA, 2009). For each cystometry parameter, an analysis of variance was performed to compare (1) the S-4 sites and S-8 sites saline groups to allow their aggregation, and (2) the aboBoNTA-4 sites or aboBoNTA-8 sites groups versus the aggregated saline groups (SAS ® ). p values < 0.05 were considered significant.

Drugs and Chemicals
AboBoNTA (Batch E194441) was provided by Ipsen Biopharm Ltd., Wrexham, UK. Antibiotics (except gentamicin) and anaesthetics were purchased from Centravet (Dinan, France) and Roche Pharma (Neuilly-sur-Seine, France). Gentamicin was purchased from Merck and Co. Inc. (Whitehouse Station, NJ, USA). All other drugs and chemicals were purchased from Sigma-Aldrich (Lyon, France).