[1] 14w P2X3-Immunoreactive Nerve Fibres in Neurogenic Detrusor Overactivity and the Effect of Intravesical Resiniferatoxin $

[1] 255w Objectives: The ATP-gated purinergic receptor P2X3 is expressed by small diameter sensory neurons and has been identified in normal and neurogenic human bladder suburothelial fibres. Animal models have shown that ATP is released by the urothelium during bladder distension, suggesting a mechanosensory role for P2X3 receptors in normal bladder function. Successful treatment of spinal neurogenic detrusor overactivity (NDO) with intravesical resiniferatoxin (RTX), which partly acts on suburothelial C fibres, provides evidence for the emergence of a C fibremediated spinal reflex. The aim of this study was to investigate the possible role of P2X3-positive innervation in this pathological voiding reflex by comparing suburothelial P2X3 immunoreactivity of controls and in patients with NDO before and after intravesical RTX. Methods: Bladder biopsies were obtained from 8 controls and 20 patients with refractory NDO enrolled in a trial of intravesical RTX. P2X3 nerve fibre density and intensity were studied in the specimens by immunohistochemistry. Results: P2X3-IR nerve fibres were significantly increased in patients with NDO compared to controls ( p ¼ 0:014). Thirteen patients had pre-and post-RTX biopsies available for immunohistochemistry; 5 of them responded clinically and 8 were non-responders. In the 5 patients who responded to RTX, there was a significant decrease in P2X3-positive fibres ( p ¼ 0:032), whereas in non-responders, P2X3-IR nerve fibre density did not change significantly. Conclusions: In patients with NDO, the numbers of P2X3-IR nerve fibres were increased in the suburothelium. There was a significant decrease in P2X3 immunoreactivity in responders to RTX, indicating a potential pathophysiological role for the P2X3 expressing fibres.

[1] 82w The ATP-gated purinergic P2X receptors have already been investigated in normal and overactive human bladder biopsies. P2X1 and P2X2 receptors are expressed mostly in detrusor smooth muscle and have been suggested to play a role in normal bladder contraction [1,2] and to be upregulated in detrusor overactivity respectively [3]. P2X3 is expressed by small diameter sensory neurons [4][5][6] and has been identified in the suburothelium in normal (post-mortem) [7] and overactive human urinary bladders of both neurogenic [7] and idiopathic origin [1,3].

[2] 173w Studies in women with idiopathic DO showed colocalization of P2X3 with the pan-neuronal marker protein gene product 9.5 (PGP9.5) on suburothelial fibres [3]. Fewer P2X3-IR fibres were observed compared to those immunostained for PGP9.5 in the normal and NDO bladders, indicating that a subpopulation of suburothelial fibres express P2X3 [7]. Kennedy and Leff have proposed a nociceptive role for ATP [8] and Ferguson proposed that ATP is released from the urothelium as a sensory mediator for the degree of bladder distension [9]. As such it is probable that P2X3 receptors localized in the suburothelium have a sensory function. Complementing this hypothesis, P2X3-deficient (P2X3 À/À ) mice demonstrate increased cystometric capacity and decreased frequency of voiding [10]. Vlaskovska et al. confirmed P2X3 receptor immunoreactivity on afferent fibres in the mouse urinary bladder and demonstrated a significantly delayed and increased threshold of activation of these afferents by P2X agonists in P2X3 À/À mice [11]. Thus it appears that P2X3-positive afferents in the suburothelium play a role in mechanosensory signal transduction in the normal mouse bladder.

[3] 95w In animal models of spinal cord disease, a segmental sacral spinal reflex has been demonstrated that stimulates detrusor efferent activity. The afferent arm of this aberrant reflex pathway is composed of unmyelinated, small diameter nerve fibres that are capsaicin-sensitive and present in the suburothelium [12]. Intravesical therapy with the C fibre toxins capsaicin (CAP) and resiniferatoxin (RTX) can reduce urinary frequency and urge incontinence in patients with neurogenic detrusor overactivity (NDO) due to spinal cord lesions [13,14], providing indirect evidence for the activation of a similar pathological sacral spinal reflex in humans with spinal NDO.

[4] 169w This is an immunohistochemical study using biopsies obtained from patients recruited into a randomised, double blind, placebo-controlled clinical trial of RTX. The clinical trial was suspended, and later closed, after clinical responses to treatment were at variance with the dose-escalating study design. It was determined that adsorption of RTX to the giving sets and delivery system used in the trial accounted for this variation in clinical response reported by patients, despite the administration of escalating doses of RTX (see Sections 2 and 3). The results of the clinical trial have not been reported in detail, however responders to treatment were identified (Table 1) and biopsies from this group taken before and after treatment were compared to those of non-responders and controls. The aim of our study was to investigate the possible role of P2X3 receptors in this pathological voiding reflex by examining P2X3 immunoreactivity in the suburothelium of NDO and control bladders. We also wanted to evaluate the effect of treatment with RTX on P2X3 immunoreactive (IR) suburothelial fibres.

[1] 95w One, 5 or 10 ml of sterile stock solution (10 mM or 6.3 mg/ml RTX in 100% dehydrated alcohol stored at À10 to À20 8C) was diluted in 9, 5 or 0 ml of dehydrated alcohol and 90 ml sterile Sodium Chloride 0.9% to achieve dosing concentrations 62.9 ng/ ml (0.1 mM), 314.4 ng/ml (0.5 mM) or 628.7 ng/ml (1.0 mM) RTX. The solutions were prepared in PVC containing sterile containers, which were then refrigerated and stored for up to 24 hours prior to administration. The time of preparation and instillation of RTX were recorded.

[2] 30w Up to 100 ml RTX or placebo was instilled into an emptied bladder at a rate of approximately 50 ml/minute and retained in the bladder for up to 30 minutes.

[3] 80w As it was not known at the time of the study that RTX adsorbs to polypropylene and PVC present in the administration sets with 20% of the RTX bound by 30 minutes, 50% by 2 hours and 80% by 4 and up to 24 hours at room temperature or 4 8C across a range of doses, the dose of RTX delivered was retrospectively estimated according to the pre instillation incubation time (Table 2; personal communication, ICOS Corporation, Bothell, WA, USA).

[4] 138w Seventeen patients received the maximum dose of RTX (label claim 1 mM) and 6 (32%) of these responded clinically to RTX. In the 6 responders, bladder capacity increased from a mean AE SEM of 153:8 AE 32:7 ml to 277:4 AE 66:2 ml. Two of the patients went into retention for several months before detrusor function returned and Twenty patients with NDO had biopsies taken at baseline (pre treatment); 17 of these 20 patients progressed through the protocol to receive label claim 1 mM RTX. Of these 17 patients, 13 had biopsies available for analysis as 3 patients declined biopsies after treatment and in a further patient the post treatment biopsy was unsuitable for processing. These 13 patients with biopsies available for analysis were divided into responder (n ¼ 5) and non-responder groups (n ¼ 8; Table 1).

[5] 130w Urothelium was present in all specimens and muscularis propria in 54.3% of them. Pre-treatment biopsies from 2 responders showed mild urothelial hyperplasia. Findings of chronic and, less often, acute inflammation, were present in most biopsies (Table 3), and increased vascularity was found in only one nonresponder pre-treatment specimen, from a patient who had a UTI. Mean inflammation score was significantly higher in patients with NDO compared to controls (1:45 AE 0:18 versus 0:62 AE 0:18, p ¼ 0:02). Mean pre-treatment inflammation scores were not different between responders, non-responders and controls and did not change following RTX treatment in both the responder ( p ¼ 0:69) and the non-responder ( p ¼ 0:28) groups. Multiple cystic von Brunn nests were identified in the baseline and subsequent biopsies taken from one responder.

[6] 51w Similar numbers of sections per biopsy were analysed for the NDO (mean AE S:D: 6:25 AE 1:97, range 3-10) compared to the control (7:0 AE 1:93, range 5-11) group, and also for the responders (5:6 AE 2:3, range 3-8) compared to the non-responders (6:0 AE 1:3, range 3-8) and the controls.

[7] 91w P2X3-IR nerve fibres were significantly increased in patients with NDO compared to controls ( p ¼ 0:014; Fig. 1). In the 5 patients who responded to RTX, P2X3positive fibres showed a significant reduction following treatment to values found in control tissues ( p ¼ 0:032; Figs. 2 and 3a). No significant change in P2X3-IR nerve fibre density was observed in nonresponders ( p ¼ 0:14, Fig. 3b) or in patients treated with placebo ( p ¼ 0:89). There was no difference in baseline values of P2X3-IR fibres between responders and non-responders.

[1] 263w The therapeutic efficacy of intravesical RTX in refractory spinal NDO has been demonstrated in up to 75% of patients [14,17] and in contrast to capsaicin, the instillation of RTX causes minimal or no sensitisation effects. However as our clinical trial of RTX progressed, several patients reported sensitisation effects at variance with the ''escalating dose'' study design. In addition, whilst the maximum dose used in this study was relatively high compared to other trials [14,17,18], surprisingly few patients had a clinical response (32% of those who received a label claim of 1 mM RTX). It was later determined that RTX adsorbs to polypropylene and PVC in the delivery system in a time-dependent manner. To exclude a dose-related bias of the clinical and immunohistochemical results and a possible cumulative effect of repeated instillations with escalating doses of the drug, we extrapolated the dose received by each patient based on the pre-instillation incubation time and compared the mean total number of treatments received by each patient group up to and beyond the point of biopsy. The means of these treatment parameters between the responder and non-responder groups were similar (Table 2). It has been reported that patients with multiple sclerosis and significant lower limb dysfunction, with a greater burden of spinal cord disease, are less likely to respond to intravesical vanilloid therapy compared to those who are less disabled [13]. In the current study however, neither disease type nor duration correlated with response to treatment and in patients with MS, clinical response to RTX therapy was independent of lower limb function, although the numbers are small.

[2] 215w RTX treatment was not associated with inflammatory or dysplastic changes in the tissues examined. Inflammatory changes, mainly chronic, were already present in 89.4% of our baseline NDO biopsies, a finding consistent with previous histopathological stu-dies of bladder biopsies from SCI patients [19,20], but also in 63% of our control biopsies (Table 3). Although a high percentage of the controls had some degree of inflammation on microscopy (mean inflammation score ¼ 0.6) the biopsies were taken from macroscopically normal bladders; we are not aware of any published data on the incidence of inflammation in such biopsies. Mean inflammation scores were significantly higher in the baseline biopsies from patients with NDO compared to controls. Although the means of the baseline responder and non-responder inflammation scores appeared to be higher than the controls, this did not reach statistical significance. This discrepancy is related to the differences in the number of patients in each group, with 20 patients in the pretreatment (baseline) group and 13 patients in the post treatment group (responder and non-responders). Mean baseline inflammation scores in biopsies from responders, non-responders did not change post-treatment in either the responder ( p ¼ 0:69) or the non-responder ( p ¼ 0:28) group. Thus it appears that the presence of inflammation is unrelated to treatment outcome following RTX administration.

[3] 58w Evidence of infection was present in baseline biopsies from 5 patients, with varying degrees of inflammation (severe n ¼ 2, moderate n ¼ 2 and mild n ¼ 1). These biopsies were included in the study, as nerve density did not change significantly in any of the specimens after repeat baseline assessments following antibiotic treatment for the UTI.

[4] 200w P2X3 is expressed on small diameter sensory neurons [6,21]. The neuronal expression of P2X3 in the suburothelium has been demonstrated in both animal [11] and human [3] studies co-localizing P2X3 with the pan-neuronal marker protein gene product 9.5. We have also previously reported the presence of P2X3immunoreactivity in the suburothelium of NDO bladders [7]. This is the first study to demonstrate that the expression of P2X3-IR nerve fibres in the suburothelium is increased in patients with NDO compared to control subjects. In the same series of biopsies, we have shown that the density of suburothelial fibres immunoreactive for PGP9.5 is increased [22]. Our results indicate that the increased suburothelial innervation observed in patients with NDO may be at least partly due to an increase of P2X3-positive suburothelial nerve fibres. We also observed a significant reduction of P2X3-IR suburothelial nerve fibres in patients with NDO who responded to intravesical RTX. Despite the small number of patients, these findings suggest that an increase of suburothelial nerve fibres expressing the purinergic sensory receptor P2X3 may be a causative factor in spinal NDO, and complements earlier studies showing that the number of suburothelial sensory nerves is increased in patients with idiopathic detrusor overactivity [23,24].

[5] 196w In the same series of biopsies, we have also reported an increase of TRPV1-IR suburothelial fibres followed by a significant decrease in the number of PGP9.5 and TRPV1-IR suburothelial nerve fibres in those patients who responded to RTX [22,25]. Since RTX targets TRPV1 receptors specifically, the decrease of P2X3-IR fibres in the same patients indicates that P2X3 and TRPV1 receptors are probably co-expressed on suburothelial afferents. This hypothesis is supported by the findings of Guo, who reported that 75% of TRPV1positive dorsal root ganglion (DRG) neurons and trigeminal ganglia also expressed P2X3 subunits in colocalisation studies [26]. Furthermore, in whole-cell patch-clamp electrophysiological recordings of DRG neuronal somata, Piper has shown that 45% of ATPsensitive rat DRG neurones respond to capsaicin and 70% of capsaicin-sensitive neurones respond to ATP. In addition the same author reported that extracellular Ca 2þ -mediated responses to capsaicin were not affected by prior exposure to ATP, but exposure of cells to capsaicin before application of ATP significantly increased the desensitization rate of ATP responses [27]. These findings complement studies in P2X3-deficient mice, in which pre-treatment with intravesical CAP resulted in attenuated afferent response to distension and intravesical administration of P2X agonists [11].

[6] 203w It is unknown why P2X3 suburothelial immunoreactivity should be increased in patients with NDO, but may be related to increased levels of neurotrophic factors such as nerve growth factor (NGF). A regulatory role for NGF in P2X3 expression has been suggested, as intrathecal administration of NGF induced P2X3 expression in sensory neurons in normal rats [28]. Furthermore, intravesical NGF has also been shown to induce detrusor overactivity through sensitization of bladder afferent terminals [29]. NGF levels in the bladder are increased after both inflammation [30,31] and chronic partial outlet obstruction [32]. The majority of our patients had urodynamic evidence of obstruction, most likely due to DSD, and had histological evidence of chronic bladder inflammation (Tables 1 and 3 respectively). Additionally, inflammatory changes as graded with a random score appeared to be more severe in the NDO patient group compared to controls ( p ¼ 0:02). Either of these factors could therefore account for elevated P2X3-IR nerve density observed in our patients compared to controls. However, this hypothesis cannot explain the differences in treatment outcome, as both mean inflammation scores and urodynamic evidence of DSD were not different between the responder and non-responder groups at baseline assessments and were not affected significantly by RTX.

[7] 141w Suburothelial unmyelinated afferent nerve fibres contain a wide variety of neuroactive substances such as ATP (4), which are believed to act in an antidromic manner [5,6,21]. However, ATP is also released from the urothelium in response to bladder distension and may bind to P2X3 expressing fibres in the suburothelium. P2X3 knockout mice have been found to have an increased bladder capacity but normal stretch-evoked ATP release by the urothelium [10,11]. In contrast, TRPV1 knockout mice release significantly less ATP from the urothelium in response to bladder distension [33]. Together with recent descriptions of myofibroblasts (or interstitial cells) in the bladder lamina propria, which were found to express the gap-junction protein connexin 43 [34] or the vanilloid receptor TRPV1 [35], our findings provide further evidence for the role of the bladder suburothelium as a mechanosensory stretch organ in normal and pathological conditions.

[1] 98w The increased numbers of P2X3-IR suburothelial nerve fibres in patients with NDO, together with the significant decrease of these fibres in patients responding to RTX, indicate a potential role for the P2X3-immunoreactive suburothelial afferents in NDO pathophysiology. As TRPV1 fibres were also markedly decreased in these responder but not in nonresponder tissues (our observations), the decreased numbers of P2X3 fibres in responders suggests that they may be co-expressed in at least some fibres with the RTX target receptor TRPV1. Further quantitative studies are necessary to characterise sub-populations of sensory neurons that may express P2X3 in human bladder dysfunction.

[1] 328w Between 3 and 5 flexible cystoscopic ''cold cup'' bladder biopsies taken from a consistent site (2 cm above and lateral to the ureteric orifice) were obtained from 20 patients (mean age AE standard deviation: 46:2 AE 10:6 years) with refractory spinal NDO due to multiple sclerosis (n ¼ 13), spinal cord vascular accidents (n ¼ 3), transverse myelitis (n ¼ 2), tropical spastic paraparesis (n ¼ 1) and cervical (C3/4) and lumbosacral (L5/S1) intervertebral disc herniation (n ¼ 1). Biopsies were also taken from 8 control subjects (mean age AE standard deviation 50:6 AE 21:5 years) being investigated for asymptomatic microhaematuria. The control group were comparable in terms of age to the responder and non-responder groups ( p ¼ 0:72 and 0.96 respectively). Controls were questioned specifically to ensure that they had no lower urinary tract symptoms. Biopsies were taken only from macroscopically normal urothelium. Clinical assessments included midstream urine specimen (MSU), followed by frequency-volume charts (FVCs) and cystometry (CMG) after withdrawal of anticholinergic medication for 10 days. In the patients with NDO, biopsies were taken at baseline and 4 weeks after the first and each subsequent instillation, and at the time of continuing clinical response in those who had improved symptomatically following treatment. Clinical response to treatment was defined as a >25% improvement in 2 or more of the following parameters, sustained for at least 6 weeks: number of micturition episodes per 24 hours, number of episodes of micturition associated with urgency per 24 hours, number of episodes of incontinence, bladder capacity, as measured by voided and catheterised volumes or maximum cystometric capacity. Patients who responded to RTX were retreated with the next highest dose when symptoms returned to baseline. Similarly, non-responders were retreated until a response was obtained or a maximum of 2 treatments of 1 mM RTX had been given. Assessments of clinical efficacy were made at weeks 4, 8, and 12 and monthly thereafter, or until a return towards pretreatment bladder function was observed.

[2] 16w Approval by the local ethics committee and informed consent were obtained from patients and control subjects.

[3] 61w Each specimen was immediately fixed in freshly prepared 4% paraformaldehyde w/v (PFA) in phosphate-buffered saline (PBS, 0.05 M phosphate; 0.9% w/v saline; pH 7.2) for 30-60 minutes and then transferred to 0.45 M sucrose in PBS and refrigerated overnight. The samples were embedded in OCT medium and 10 mM sections were cut using a cryostat and stained with haematoxylin and eosin.

[4] 96w The specimens were immediately fixed as described above and stored at 4 8C. Frozen sections of 4-5 mm thickness were collected onto clean glass superfrost slides. After washing in PBS, endogenous peroxidase was blocked by incubation with 0.3% w/v hydrogen peroxide in methanol. After a further wash in PBS, the tissue sections were incubated with rabbit antibody to P2X3 (Roche Bioscience #24485, 1:10,000) overnight. Specificity of the antibody has been previously described [15]. Sites of antibody attachment were revealed using a nickel enhanced ABC (peroxidase) method [16]. Nuclei were counterstained with 0.1% w/v aqueous neutral red.

[5] 72w Two pathologists blinded to both the origin of the tissue (controls or NDO patients; pre or post RTX or placebo) and the clinical response to treatment assessed each section for inflammatory changes, site of inflammation, vascularity and urothelial changes (hyperplasia and dysplasia). Where inflammation (acute or chronic) was present, a subjective score was used that ranged from 1 (mild) to 3 (severe). The histological findings were correlated with the results of MSU.

[6] 15w The Mann-Whitney test was used for statistical analysis (p values <0.05 were considered statistically significant).

[1] 79w A qualitative method of analysis was adopted for the evaluation of P2X3 staining, with a random grading scale of nerve fibre density and intensity ranging between 0-3, increasing by increments of 0.5, each increment being equivalent to up to 5 additional immunoreactive fibres per section. At least 3 and up to 11 sections were analysed per each evaluated biopsy and results were expressed as means of two independent observers (A.A. and Y.Y.), that were blinded to each other's results.