Nerve sprouting and neurogenic inflammation characterize the neurogenic detrusor overactive bladder of patients no longer responsive to drug therapies

Abstract Urothelium and Lamina Propria (LP) are considered an integrate sensory system which is able to control the detrusor activity. Complete supra‐sacral spinal cord lesions cause Neurogenic Detrusor Overactivity (NDO) whose main symptoms are urgency and incontinence. NDO therapy at first consists in anti‐muscarinic drugs; secondly, in intra‐vesical injection of botulinum toxin. However, with time, all the patients become insensitive to the drugs and decide for cystoplastic surgery. With the aim to get deeper in both NDO and drug's efficacy lack pathogenesis, we investigated the innervation, muscular and connective changes in NDO bladders after surgery by using morphological and quantitative methodologies. Bladder innervation showed a significant global loss associated with an increase in the nerve endings located in the upper LP where a neurogenic inflammation was also present. Smooth muscle cells (SMC) anomalies and fibrosis were found in the detrusor. The increased innervation in the ULP is suggestive for a sprouting and could condition NDO evolution and drug efficacy length. Denervation might cause the SMC anomalies responsible for the detrusor altered contractile activity and intra‐cellular traffic and favour the appearance of fibrosis. Inflammation might accelerate these damages. From the clinical point of view, an early anti‐inflammatory treatment could positively influence the disease fate.


| 4077
TRAINI eT Al. guarantees the bladder filling by maintaining low internal pressure. 1 Once the bladder is adequately full and/or intraluminal pressure is sufficiently high, the micturition reflex initiates. 1,2 The voiding is due to parasympathetic nerve activation that induces detrusor contraction and upper urethral sphincter relaxation. The effective emptying of the bladder is under the voluntary control exerted by different nuclei located in the brain and spinal cord whose final responses are the relaxation of the lower striated urethral sphincter mediated by the pudendal nerves. 3,4 In the last decade, several studies on the bladder physiology have highlighted the importance of a sensory system located in the mucosa in the perception of the organ filling and voiding and in its modification. This sensory system consists of the urothelium (U), and of the nerve endings (NE) and specialized stromal cells (myofibroblasts, Myo, and telocytes, TC) resident in the lamina propria (LP). 5 The hypothesis is that the sensation of bladder fullness takes place in the mucosa where, during filling, the U releases chemical mediators that excite the afferent fibres. The TC and Myo of the LP respond both to the molecules released by the U and to the wall stretching, and, because of their vicinity to the nerve endings, might directly influence the afferent responses. 2,5 Complete supra-sacral spinal cord lesion of traumatic, inflammatory or degenerative origin causes neurogenic detrusor overactivity (NDO), a pathology characterized by high intra-vesical pressure, detrusor-sphincter dyssynergia, reduced bladder capacity, urinary frequency, urgency and incontinence, post voiding urine residual. 3 If untreated NDO may potentially lead to upper urinary tract damage and renal failure. 6 Up-to-date the treatment of NDO consists in the administration of oral anti-muscarinic drugs whose efficacy is due to the blockade of the efferent parasympathetic innervation in the detrusor thus preventing involuntary contractions and bladder emptying. 3 With time, these drugs lose their efficacy in many patients, or their side effects result too severe. The second-choice treatment is the intra-detrusor injection of the botulinum toxin whose mechanism(s) and site(s) of action are under debate. Unfortunately, it is also possible that a patient does not respond to the toxin, since the beginning. Nevertheless, earlier or later, the patients become insensitive to the pharmacological therapies and the remaining possibility is the bladder augmentation as effective long-term solution. 3 The spinal cord interruption causes the complete loss of the descending (voluntary) pathway and, consequently, of the functional and trophic control exerted by the central nerves on the peripheral innervation. Numerous studies in NDO patients, mainly done in the early phases of the disease, report quantitative and qualitative changes in the bladder innervation, 7,8 variable degree of detrusor fibrosis 6,12 and intense inflammatory signs 6,10,13,14 with an important involvement of the Myo and TC network. 14 Botulinum toxin treatment seems to ameliorate, at least transitorily, some of the innervation changes 11 and the fibrosis. 6 On the contrary, the toxin does not influence the inflammatory picture. 6,10,13,14 This study investigates for the first time the nervous, muscle and connective tissues in the bladder wall of NDO patients no longer responsive to botulinum toxin and subjected to cystoplastic surgery.
At these aims, immunohistochemistry and transmission electron microscopy (TEM) are used. The results are correlated with the urodynamic outcomes. With this work we intend to give a contribution in the understanding the pathogenic mechanism(s) responsible for the NDO evolution and for the loss of the drug efficacy.

| Subjects and sample collection
Full thickness bladder specimens of eight patients (four females, four males; mean age: 40.6 ± 2.9 years; mean years of treatment 14 ± 1.5) with a clinical diagnosis of NDO and subjected to partial/total cystectomy were collected. The NDO was due to supra-sacral spinal cord lesions. The surgical intervention choice based on the loss of anti-muscarinic drugs and botulinum toxin effectiveness and on the combination of one or more of the following elements: (i) the urodynamic parameters (see below); (ii) the presence of vesicoureteral reflux; (iii) personal reasons. Notably, one of the patients never responded to the botulinum toxin, thus he was separately evaluated and named non-responder. Full thickness bladder specimens of six patients (one female, five males; mean age: 72 ± 3 years) operated for bladder cancer represented the controls. During the surgical procedure, specimens were collected from the bladder lateral wall. In controls, attention was payed to collect the specimens far from the cancer lesions. Then, the specimens were processed for histological, histochemical, immunohistochemical and TEM investigations.

All the patients gave written informed consent and the local Ethical
Committee approved the study protocol.

| Routine histology and histochemistry
The specimens were fixed in 4% paraformaldehyde in 0. After washing with tap water, both staining were visualized by light microscopy (Reichert Technologies, Reichert Inc, Depew, USA).

| Immunohistochemistry
The sections were deparaffinized and rehydrated as above, then

| Quantitative analysis
The fibrosis quantitation in the detrusor was done on optical images acquired with 10× objective from Van Gieson stained slides (five images/patient). Overlapping between adjacent portions was accurately avoided. The analysis was performed by using Image J (NIH, Bethesda,

Host IHC Producer
Primary antibody TA B L E 1 List of primary and secondary antibodies ML, USA). Choosing the red in the gamma of the image colours, the connective tissue areas were selectively identified and their percentage respect to the total area was calculated. As reported in Comperat et al, 6 it was chosen a cut of 20% to distinguish between: (i) mild fibrosis, when less than 20% of muscle wall was affected, and (ii) consistent fibrosis, when more than 20% of muscle wall showed fibrosis.
The quantitative analysis of PGP9.5, nNOS and S100β immunoreactivity (IR) was done in optical images acquired with 40× objective. The quantitation of these markers was performed separately in four regions of interest (ROI) of the bladder wall, namely the upper lamina propria (ULP), the deep lamina propria (DLP), the detrusor and the adventitia. For each ROI, 10 images/patient were acquired and each one was analysed by using Image J. The photographs' threshold value was set to analyse the structures of interest. The labelling was converted to a grey scale image and the intensity of labelling and the area of IR were calculated. The results were expressed as mean optical density ±SEM. Unpaired student t test was used for statistical analysis and differences of P < 0.05 between the groups were considered as significant.

| RE SULTS
The urodynamic evaluation of the patients, showing NDO and reduced bladder compliance, and presence of incontinence and/or vesico-ureteral reflux are reported in Materials and Methods.
All the patients show important modifications of the muscular, connective and nervous tissues compared to controls. The Some SMC display clusters of αSMA-IR located close to the nucleus ( Figure 1I). Under the TEM, the clear perinuclear ring is devoid in contractile filaments, that are confined to the periphery ( Figure 1E) and contain dilated cisternae of the Golgi apparatus and vacuoli ( Figure 3D). In some cells, likely in those carrying αSMA-IR clusters, the dense bodies and the contractile filaments have a chaotic distribution ( Figure 1F).

| Muscarinic receptor type 2 (Mr2)-IR
In controls Mr2-IR appears as spots positioned along the SMC plasma membrane ( Figure 1G) while in all the NDO patients, the labelling is mainly cytoplasmic ( Figure 1H,I, Figure S1A,B). In those SMC where the αSMA-IR is clustered to the nucleus proximity, the Mr2-IR is adjacent to these clusters ( Figure 1I).

| Caveolin 1 (Cav1)-IR
In controls Cav1-IR appears as punctate spots regularly aligned along the SMC plasma membrane (Figure 2A); in the NDO patients the spots are confined to small portions of the plasmalemma of some SMC ( Figure 2B) while in others these spots have a distribution like in controls but more intensely labelled ( Figure 2C). In the non-responder patient, the Cav1 positive spots appear particularly thick and seemed to enter the SMC body ( Figure 2D). In this patient, under the TEM, the caveolae are packed in large groups and the deepest ones look devoid of any contact with the plasma membrane ( Figure 1E).

| PAS staining and TEM
In NDO patients, many SMC are rich in PAS-positive granules

| Nerve structures
The pan neuronal marker Protein Gene Product 9.5 (PGP9.5) antibody identifies both neurons and nerve fibres; the S100β antibody labels the glial cells and the neuronal Nitric Oxide Synthase (nNOS) antibody labels the nitrergic structures present in the entire bladder wall.
In the detrusor, all markers label thin nerve fibres ( Figure 5A-F) that appear reduced in the NDO patients ( Figure 5D-F) compared with controls ( Figure 5A-C). Under the TEM, both, nerve fibres and endings of the NDO patients, show normal features and maintain the contacts with the SMC (Figure 5G,H). Quantitation of the three markers demonstrates that all of them are significantly decreased in the patients respect to controls ( Figure 7A).
In the adventitia, the three markers label small ganglia and thick nerve bundles. Quantitation shows a PGP9.5-IR decrease in the NDO patients; however, this decrease reaches the significance in the non-responder only. On the contrary, the S100β-and nNOS-IR are significantly reduced in all patients, non-responder included ( Figure 7B).
In the ULP of controls, PGP9.5-, S100β-and nNOS-IRs detected  Figure 6D-I). The quantitation of PGP9.5-and S100β-IR demonstrates a significant increase in NDO patients in comparison with controls ( Figure 7C) while that of nNOS-IR is unchanged respect to controls ( Figure 7C). On the contrary, in the non-responder patient, all the three labelling are significantly decreased compared to controls and to the other NDO patients ( Figure 7C).
In the DLP, PGP9.5-, S100β-and nNOS-IRs identify thin nerve bundles and few varicose nerve fibres most of which close to blood vessels (data not shown). Quantitation of PGP9.5-, S100β-and nNOS-IRs does not show any difference between controls and NDO patients, non-responder one included ( Figure 7D). Connective tissue anomalies consist in a significant increase in the detrusor fibrosis compared to controls and a consistent F I G U R E 7 Bladder nerve structures. Quantitative analysis. In the detrusor (A) and adventitia (B), PGP9.5-, S100β-and nNOS-IR nerve structures are significantly decreased in the NDO patients (dark grey column), non-responder one included (light grey column), respect to the controls (white column). In the ULP (C), PGP9.5-and S100β-IR nerve structures are significantly increased in the NDO patients respect to the controls. On the contrary, both markers are significantly reduced in the non-responder patient.
The nNOS-IR nerve structures do not change in the NDO patients compared to the controls except for the non-responder one where they are significantly reduced. In the DLP (D), none of the markers investigated show significant changes amongst the groups. *P < 0.05; **P < 0.005; ***P < 0.001 thickening of the SMC basal lamina. While fibrosis represents "per se" a mechanical obstacle to contraction, the thicker basal lamina might affect contractility also by interfering with the SMC intra-extra cellular exchanges and number and functionality of the cell-to-cell contacts. Indeed, a significant decrease of SMC intermediate junctions was reported in the neurogenic bladder 12 and qualitative and quantitative anomalies of the collagen have been described in "non-compliant" bladders. 17,18 About fibrosis, a peculiar case is the non-responder patient who, treated only once with the botulinum toxin, presents the highest percentage of fibrosis.
Interestingly, this case supports the hypothesis that the fibrosis in NDO bladders is not related to the botulinum toxin treatment. 6,19 To synthesize, the numerous detrusor alterations reported well explain the muscle dysfunctionality progressively limiting the drug effectiveness.
Presently, and for the first time, the bladder innervation is This surprising and unexpected finding is suggestive of a nerve sprouting and needs to be discussed. In the last decade, a growing number of reports attribute to the urothelium and LP a main role in bladder functionality. These two layers would form an integrate sensory system working as a stretch receptor able to control and condition the detrusor responses. 3,21 Moreover, it has also been proposed that the drugs used for NDO treatment (anti-muscarinic and botulinum toxin) act (primarily) at this level. 2 In the detrusor of our patients, according with other reports, 6,13 the inflammation mainly consists of an eosinophil granulocytes infiltrate. The question arising from these data is whether and how inflammation plays a role in the NDO pathogenesis and in the loss of botulinum efficacy. The genesis of inflammation is multiple and the causes difficult to remove. Mechanical stimuli on the bladder wall such as the repeated catheterization, drug injections and numerous endoscopies have importance. Interestingly, a significant font of inflammation is the nervous system. Indeed, a neurogenic inflammation has been considered responsible for the maintenance and chronicity of this condition. 13 It has been hypothesized that an increased expression of excitatory sensory receptors in NDO is associated with an increase in histological inflammation. 12,13,28 Furthermore, the present report of a greater increase in S100β-IR compared to PGP9.5-IR (+60% vs +33%, respectively) in the ULP might be another sign in favour of a neurogenic inflammation since increased expression of S100β has been found in chronic inflammatory diseases. 29 Finally, signs of activated Schwann cells have been reported in NDO bladder. 12 In summary, our findings indicate that NDO results from a cascade of events triggered by the spinal cord lesion but whose evolution generates inside the bladder with some differences between the detrusor and the LP. Although the detrusor damages have importance, we consider the bladder mucosa as the main site where the NDO evolution takes place, the damage severity is determined, and the length of the drug efficacy is conditioned. Furthermore, we believe the neurogenic inflammation is pivotal in accelerating these events progression.
To conclude, although the NDO fate is in some ways unavoidable, the possibility to slow down its evolution towards the surgical intervention seems to reside in a careful and continuous monitoring of any signs of urinary inflammation and in their targeted treatment.

ACK N OWLED G EM ENTS
The authors wish to thank the surgeon Prof. S. Serni and his collaborator, Dr. I. Frizzi, for making available the control specimens.

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

AUTH O R S CO NTR I B UTI O N
CT: performed the histological and immunohistochemical experiments, planned the work's steps and analysed the collected data;