PATTERNED SACRAL AFFERENT STIMULATION FOR SUPPRESSION OF URETHRAL REFLEXES AND RESTORATION OF BLADDER VOIDING AFTER CHRONIC SPINAL CORD INJURY
by
JAIME LEE MCCOIN
Submitted in partial fulfillment of the requirements
for the degree of Doctor of Philosophy
Dissertation Adviser: Dr. Kenneth J. Gustafson
Department of Biomedical Engineering
CASE WESTERN RESERVE UNIVERSITY
August, 2013 CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the thesis/dissertation of
______Jaime Lee McCoin______
candidate for the ______Doctor of Philosophy______degree *.
__ Kenneth J. Gustafson, Ph.D.__
(chair of the committee)
__Kenneth J. Gustafson, Ph.D.__
__Robert F. Kirsch, Ph.D.__
__Dustin Tyler, Ph.D.__
__Thomas Chelimsky, M.D.__
___Narendra Bhadra, M.D. Ph.D.__
(date) _____6/6/2013______
*We also certify that written approval has been obtained for any
proprietary material contained therein.
ii
DEDICATION
To my parents, Jeff and Carol, for making me believe I really could change the world and always pushing me to be better. To my husband and best friend, Matthew Fulgo, without
whose bottomless love, patience and encouragement this dissertation would not have
been possible. And to all the friends who have shaped my life on the road through
college, graduate school and beyond.
iii
Table of Contents
TITLE PAGE ...... I COMMITTEE SIGNATURE PAGE...... II DEDICATION ...... III LIST OF TABLES ...... X LIST OF FIGURES ...... XI ACKNOWLEDGEMENTS ...... XII LIST OF ABBREVIATIONS ...... XIII
DISSERTATION ABSTRACT…………………………………………………………1
CHAPTER 1: INTRODUCTION AND LITERATURE REVIEW PROBLEM OVERVIEW ...... 4 THE LOWER URINARY TRACT ...... 5 Anatomy and Function ...... 5 Neurophysiology ...... 5 BLADDER DYSFUNCTION AFTER SPINAL CORD INJURY ...... 7 EXISTING THERAPIES FOR NEUROGENIC BLADDER ...... 9 Standard Clinical Treatments ...... 9 Pharmaceuticals ...... 9 Catheterization ...... 9 Surgical Interventions ...... 10 Neuroprosthetic Approaches to LUT Dysfunction ...... 11 Extradural mixed sacral roots: Brindley Vocare System ...... 11 Sensory Stimulation for LUT Neuromodulation Neuroprosthetics ...... 12 Sacral Root Stimulation ...... 12 Dorsal Penile Nerve Stimulation ...... 13 Posterior Tibial Nerve Stimulation ...... 14 Sensory Stimulation for Urethral Reflex Suppression ...... 14
iv
HYPOTHESES AND AIMS ...... 16 Aim 1: Demonstrate suppression of urethral reflexes using patterned sacral afferent stimulation in animals with chronic SCI ...... 16 Variable Hierarchy ...... 16 Aim 2: Demonstrate improved voiding efficiency using patterned sacral afferent stimulation in chronic SCI animals ...... 18 Aim 3: Determine if patterned sacral afferent stimulation can reduce abnormal urethral reflexes in human subjects with chronic SCI ...... 18 IMPACT ...... 19
CHAPTER 2: ELECTRICAL STIMULATION OF SACRAL DERMATOMES SUPPRESSES ABERRANT URETHRAL REFLEXES AFTER CHRONIC SPINAL CORD INJURY ABSTRACT ...... 25 INTRODUCTION ...... 26 Clinical Impact of SCI on Micturition ...... 26 Electrical Stimulation for Control of LUT Function ...... 26 MATERIALS & METHODS ...... 27 Chronic SCI animal model ...... 27 Terminal Procedure ...... 28 Data Collection: Stimulation Protocol ...... 30 Data Analysis ...... 31 RESULTS ...... 32 Reflex Activity Following Chronic SCI ...... 32 Effect of Dermatome Stimulation on Urethral Reflex Suppression ...... 32 DISCUSSION ...... 35 Reduction of Urethral Spasm after Chronic SCI ...... 35 Non-Responder Animals ...... 36 CONCLUSIONS ...... 37
v
CHAPTER 3: PATTERNED SACRAL AFFERENT STIMULATION PRODUCES CLINICALLY EFFECTIVE VOIDING IN CHRONIC SCI FELINES ABSTRACT ...... 46 INTRODUCTION ...... 47 MATERIALS & METHODS ...... 48 Chronic SCI animal model ...... 48 Urethral Reflex Suppression ...... 49 Voiding Protocol: Voiding Under Anesthesia ...... 50 Awake Voiding ...... 50 Daily Maintenance Voiding ...... 51 Subcutaneous Stimulation and Cutaneous Anesthesia ...... 51 Data Analysis ...... 52 RESULTS ...... 52 Suppression of EUS Reflex activity ...... 52 Voiding Under Anesthesia ...... 53 Awake and Daily Maintenance Voiding ...... 54 Subcutaneous Stimulation and Cutaneous Anesthesia ...... 55 DISCUSSION ...... 55 Sacral afferent suppression of urethral reflexes ...... 56 Sacral afferent stimulation improves bladder voiding in awake animals ...... 57 Bladder maintenance of animals using only electric stimulation ...... 57 Comments on Neurophysiology ...... 58 CONCLUSIONS ...... 58 ACKNOWLEDGEMENTS ...... 59
vi
CHAPTER 4: CHAPTER 4: PATTERNED AFFERENT STIMULATION OF SACRAL DERMATOMES FOR SUPPRESSION OF URETHRAL SPHINCTER REFLEXES IN HUMAN SUBJECTS ABSTRACT……………………………………………………………………………..63 INTRODUCTION ...... 67 MATERIALS AND METHODS ...... 68 Subject selection ...... 68 Experiment Overview ...... 69 Equipment and Set-up ...... 69 Sacral Dermatome Stimulation ...... 70 Data Analysis ...... 71 RESULTS ...... 72 Subject Response to Urodynamics ...... 72 Sacral Dermatome Stimulation ...... 72 Adverse Events ...... 73 DISCUSSION ...... 73 CONCLUSIONS ...... 74 ACKNOWLEDGEMENTS ...... 75
CHAPTER 5: DISCUSSION OF RESULTS SUMMARY OF KEY CONTRIBUTIONS ...... 82 DISCUSSION OF SPECIFIC AIMS ...... 85 Aim 1: Reduction of abnormal urethral reflex activity using patterned sacral afferent stimulation in chronic spinal-injured animals...... 85 Aim 2: Patterned Afferent Stimulation for Bladder Voiding in Chronic SCI animals 86 Aim 3: Application of Patterned Afferent Stimulation to Human Subjects with SCI . 87 HYPOTHESIS OF MECHANISM ...... 88 CONTINUING-RESEARCH PROPOSAL ...... 96 Remaining Challenges ...... 96 Successful demonstration of reflex suppression in human subjects ...... 96 Understanding of the neurophysiologic mechanism underlying afferent urethral reflex suppression ...... 97
vii
Development and deployment of a neuroprosthetic device for clinical trials ...... 97 Future Research Aims ...... 98 Aim 1: Improve understanding of the neurophysiologic mechanism ...... 99 Aim 1.1: Identify nerve pathways connecting cutaneous afferents to urethral sphincter motoneurons ...... 99 Aim 3: Demonstration of reflex suppression in humans with patterned sacral afferent stimulation ...... 101 Aim 3: Develop a neuroprosthetic device for future clinical trials ...... 104 BROADER IMPLICATIONS OF THIS WORK ...... 108 Treating detrusor-sphincter-dyssynergia in other neurologic disorders ...... 108 Application of patterned afferent stimulation to other aberrant reflexes ...... 108 CONCLUSIONS ...... 109
APPENDIX I: ADDITIONAL ANIMAL STUDY DETAILS NOT PREVIOUSLY INCLUDED CHRONIC SCI ANIMAL MODEL ...... 110 Electrode Implant Phase ...... 110 Spinal Transection ...... 111 Chronic SCI Maintenance ...... 112
APPENDIX II: ANIMAL AND HUMAN TESTING SETUP DOCUMENTS IMPORTANT POINTS TO AVOID CURRENT GROUND LOOPS ...... 114
APPENDIX III: COMPARISON OF ACUTE AND CHRONIC FELINE SCI DATA PREVIOUS ACUTE STUDY ...... 117 MATERIALS AND METHODS: COMPARISON OF ACUTE AND CHRONIC PREPARATIONS ...... 117 COMPARISON OF RESULTS ...... 119 DISCUSSION ...... 121 ACKNOWLEDGEMENTS ...... 122
viii
APPENDIX IV: ALTERNATE ANALYSES OF URETHRAL REFLEX DATA METHODOLOGY ...... 125 Frequency Content Exploration – Spectrograms ...... 125 Signal Power Exploration – Total (Mean) Power ...... 126 RESULTS ...... 127 DISCUSSION ...... 127
BIBLIOGRAPHY ...... 131
ix
LIST OF TABLES
Table 2-I: Control and Summarized dermatome data from individual animals…..…....38
Table 4-I: Summarized clinical information for all subjects…………………………..78
x
LIST OF FIGURES
Figure 1.1: Neurophysiology of the lower urinary system ……………………………...20
Figure 1.2: Bladder and sphincter traces for an infant, normal adult, and adult with detrusor-sphincter-dyssynergia…………………………………………………………..21
Figure 1.3: Electrode interface locations for controlling the lower urinary tract………..23
Figure 2.1: Surface electrode placement and patterned stimulus waveform...... 39
Figure 2.2: Suppression of aberrant urethral reflexes after chronic SCI ...... 40
Figure 2.3: Spatial selectivity of stimulus location ...... 41
Figure 2.4: Stimulus pattern affects reflex suppression ...... 43
Figure 3.1: Sample voiding traces with and without afferent stimulation...... 60
Figure 3.2: Voiding under light anesthesia...... 61
Figure 3.3: Awake voiding compared to manual expression...... 62
Figure 3.4: Daily maintenance voiding over time...... 63
Figure 4.1: Human System Diagram……………………...……………………………..76
Figure 4.2: Sample pressure trace showing periods of afferent stimulation……………..77
Figure 4.3: Sphincter Reflex Pressures With and Without Afferent Stimulat…………..79
Figure AI.1: General Instrumentation Setup- Chronic SCI Cats……………..………...113
Figure AI.2: General Instrumentation Setup- Human Subjects…………………..….....116
Figure AIII.1: Representative trials from two acute SCI animals……………………...123
Figure AIII.2: Direct comparison of acute to chronic SCI key metric data…………….124
Figure AIV.1: Spectrogram showing periods of suppression and control……………..128
Figure AIV.2: Sample Trace of Frequency Power Bins……………………………..…129
Figure AIV.2: Sample Trace of Frequency Power Bins………………………………..130 xi
ACKNOWLEDGEMENTS
The work reported herein was supported by the National Institute of Health award numbers DK077089 and T32 EB004314; Department of Veterans Affairs RR&D668; and the Cleveland FES Center. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or VA.
This work would not have been possible without a great deal of help from my colleagues in the Neural Engineering Center. To the graduate students who gave me guidance in my early years: Tim Mariano, Timothy Bruns, and especially Adam Boger; to the vet techs
Tina Emancipator and Jen Mikulan, for teaching me about veterinary medicine and working so hard with every one of my animals; to my labmate Mafred Franke for keeping the electronics running and sharing my burdens in the wee hours of all-night experiments; to my amazing undergraduates Julie Murphy and Daniel Young, for writing superior
Matlab code and helping to processing multiple gigabytes of data; to my mentor and sounding-board, Dr. Narendra Bhadra, for investing so much time and care in training me to be a passable surgeon and a budding scientist; and especially to my advisor, Ken
Gustafson, for five crazy, challenging, growth-inducing years—you never let me quit, and I have grown in ways I never imagined. I cannot thank any of you enough, nor express what you have done for me. I only hope I can continue to make you proud.
xii
LIST OF ABBREVIATIONS
SCI— Spinal cord injury
DSD— Detrusor-sphincter-dyssynergia
EUS—External urethral sphincter
S-, e.g. S2—Sacral root level 2
IACUC- Institutional Animal Care and Use Committee
IRB- Institutional Review Board
PSAS—Patterned Sacral Afferent Stimulation
PFCN—Posterior Femoral Cutaneous Nerve
xiii
Patterned Sacral Afferent Stimulation for Suppression of Urethral Reflexes and Restoration of Bladder Voiding After Chronic Spinal Cord Injury
Abstract
By
JAIME LEE MCCOIN
Trauma to the central nervous system often leads to a disruption in voluntary control
over the muscles and organs below the level of injury. Individuals with spinal cord injuries
(SCI) subsequently experience loss of bladder control and the emergence of dyssynergic reflex patterns. Reflex contractions of the external urethral sphincter (EUS) become uncoordinated from bladder contractions. This results in high bladder pressures, uterovesical reflux, and poor bladder emptying leading to serious medical complications.
Neuromodulation of sacral spinal circuits provides a novel way to reduce bladder and
EUS reflexes and improve bladder control. Existing neuromodulation techniques can control
overactive bladder reflexes but do not address EUS spasms. Complete bladder emptying
requires regular catheterization. This dissertation demonstrates a novel technique for the
targeted suppression of sphincter reflexes and expands on work previously done in animals
with acute spinal injuries.
Sacral dermatomes were mapped and a select set of stimulus parameters (duty cycle,
cycle time, and base frequency) were investigated in cats with chronic SCI. Patterned sacral
afferent stimulation (PSAS) was found to selectively suppress aberrant EUS reflexes in 5 of 8
1 chronic felines. Successful suppression was highly dependent on the specific dermatome location and stimulus pattern delivered; only the S1/S2 dermatome with (.75 ON .25 OFF, 20
Hz; 20 Hz continuous) stimulation was effective across all animals.
Reflex suppression was translated into improved bladder voiding in 3 animals. Two animals received daily stimulation without anesthesia; intermittent sacral root stimulation with PSAS replaced manual expression, the clinical standard of care, as the primary method of bladder maintenance in both animals. These are the first results demonstrating functionally meaningful outcome measures with sacral dermatome stimulation.
The stimulus patterns and locations found effective in cats were tested in ten human subjects to evaluate if PSAS would produce comparable reflex suppression. None of the human subjects experienced a significant decrease in sphincter activity. It is expected that a more expansive investigation of the stimulus parameter space will be necessary for effective translation in humans. If such stimuli are found, PSAS may provide SCI individuals with a non-invasive alternative for bladder voiding.
2
CHAPTER 1: Introduction and Literature Review
3
PROBLEM OVERVIEW
Urinary bladder function becomes disordered after spinal cord injury due to the loss
of supraspinal input and conscious control over the micturition reflex. Spinal reflexes
become hyper-responsive to sensory input from the bladder and other pelvic nerves, and
these reflexes prevent coordinated bladder-sphincter contraction which leads to
inefficient bladder voiding.
Renal failure secondary to bladder dysfunction was the leading cause of death in SCI
individuals up until the mid-1970s, when intermittent catheterization became standard
clinical practice for bladder maintenance. Catheterization and anticholinergic
pharmaceuticals are still the standard of treatment 40 years later, but new therapies are
being developed to restore more natural voiding function. Electrical stimulation of pelvic organs and sacral spinal nerves can suppress bladder reflexes through neuromodulation— changing the activity of the nervous system through sensory input to the sacral spinal circuits.
Bladder overactivity has been widely addressed by neuromodulation therapies, but sphincter spasms remain a problem. Continence is an improvement, but without the ability to selectively inhibit sphincter reflex contractions it is impossible to produce normal bladder voiding. This necessitates the continued use of catheters or other suboptimal options.
The work presented in this dissertation details a new approach to targeting overactive urethral sphincter reflexes, called Patterned Sacral Afferent Stimulation, or PSAS.
4
THE LOWER URINARY TRACT
Anatomy and Function
The principal function of the lower urinary tract (LUT) is to collect and store urine from the kidneys and periodically expel it from the body in a socially acceptable time and place. The LUT consists of two mechanically-simple components: the urinary bladder, which stores urine at low pressures; and the urethra, which provides the conduit for directing urine out of the body. Within the urethra there are two release-values: the internal urethral sphincter (IUS) and the external urethral sphincter (EUS). The IUS is located immediately distal to the bladder neck and is under involuntary autonomic control. The IUS is composed of smooth muscle and is a direct continuation of the bladder muscle. The EUS consists of somatic muscle and is under voluntary control. In men the external sphincter is located immediately distal to the prostate. The EUS in women is located mid-way along the length of the urethra(1).
In order to maintain low-pressure storage (micturition), the detrusor maintains a large compliance without contracting over a wide volume range. Standard bladder capacity ranges from 500 to 1000 ccs in able-bodied adults(2). The internal and external urethral sphincters increase their activity as the bladder fills, in order to prevent leakage and maintain continence(1). When bladder emptying (voiding) is desired, the urethral sphincters relax while the detrusor simultaneously contracts. The synergic relationship between the bladder and EUS allows for low pressure, complete bladder emptying.
Neurophysiology
The lower urinary tract is mechanically simple, but has a complex neurologic innervation comprising both autonomic and somatic components (Figure 1.1). The
5 bladder is under involuntary autonomic control and is innervated by both parasympathetic and sympathetic nerve bundles. The pelvic nerves provide parasympathetic drive to the bladder and arise from the S2 to S4 spinal levels. The hypogastric nerves provide sympathetic inhibition to the bladder and increased tone to the internal urethral sphincter and corresponding urethral smooth muscles. The hypogastric nerves originate from the T-11 to L-2 spinal thoracolumbar spinal levels. The external urethral sphincter is innervated by somatic fibers in the pudendal nerve, originating from the S1 through S4 spinal levels, and is connected to higher order brain centers through the brainstem. Somatic innervation of the pelvic floor muscles additionally contribute to urinary continence, particularly in women.
Stretch receptors in the detrusor wall increase in firing rate as the bladder fills, providing information to the spinal cord through both pelvic and hypogastric nerves.
This information is relayed through local spinal circuits and up the spinal cord to the micturition center (also known as Barrington’s nucleus) in the pons of the brainstem.
The micturition center coordinates incoming sensory information from the spine with the higher order cognitive control in the forebrain. As long as the bladder has not reached its structural capacity, the forebrain is able to block voiding through descending parasympathetic inhibition from the micturition center.
The micturition reflex is initiated when the micturition center relaxes the external urethral sphincter and stimulates parasympathetic neurons in the sacral spinal cord. This incites a bladder contraction with simultaneous sphincter relaxation, producing coordinated low pressure, low residual bladder voiding. Afferent feedback from sensory
6
neurons within the smooth muscle of the urethral conveys information to the sacral spinal
cord about urine flow, and helps to further enhance bladder contraction.
The system response to urethral afferent firing is different depending on the “state” of
the system at the time. During intentional micturition, urine flow serves to enhance
parasympathetic contraction of the bladder and relaxation of the external urethral
sphincter. However, when the system is in a state of continence, urine flow within the
urethra is interpreted as an unwanted leak and sympathetic activation of the internal and
external sphincters is increased. This is known as the “guarding reflex”.
The normal coordinated actions of the urinary bladder and sphincter are dependent on
a highly complex network of spinal and supraspinal control networks. The breakdown of
any one of the control centers can produce a dysfunctional cascade throughout the entire
system.
BLADDER DYSFUNCTION AFTER SPINAL CORD INJURY
There are approximately 300,000 individuals with spinal cord injuries living in the
United States, with an 80/20 ratio of men to women (3). An estimated 85-90 % of SCI
individuals experience some form of lower urinary tract dysfunction (4). It is important to distinguish LUT dysfunction that is caused by neurogenic damages versus dysfunction that is primarily mechanical in origin. Non-neurogenic bladder disorders are better
treated with surgical procedures such as pelvic slings or prostatectomies(5). Although
neurogenic bladder dysfunction can occur in individuals without any obvious central
nervous system damage (e.g. women with neurogenic detrusor over-activity), the work
presented here focuses solely on neurogenic dysfunction following spinal cord injury.
7
The presentation of neurogenic bladder dysfunction after SCI depends largely on the spinal level where damage occurred. Lesions of the sacral spinal cord or below generally lead to an areflexive bladder and underactive sphincter; these are termed lower motor- neuron (LMN) injuries. Upper motor-neuron (UMN) injuries occur at the lower thoracic spinal level and above, preserving the spinal reflex circuits that drive bladder and sphincter activity(6).
The restoration of pelvic organ functions is a highly desired priority for individuals with SCI (7,8). Bladder and bowel dysfunction negatively impact individuals’ quality of life (QOL). Urinary tract infection secondary to bladder management complications is one of the leading causes of re-hospitalization amongst SCI patients, contributing to an average annual cost of care from $16-20k, in 2005 dollars (9). Urinary and fecal incontinence are leading causes for patients to be institutionalized (10) as it puts a significant burden on patient care givers. Individuals with reduced hand function are especially susceptible to a loss of independence but bladder and bowel care are time consuming and socially-limiting for all injury levels(11).
Detrusor-sphincter dyssynergia (Figure 1.2) is a leading cause of morbidity in SCI patients, and if left untreated can cause severe medical complications. Reflex bladder contractions against the spastic sphincter outlet produce dangerously high intra-vesicular pressures, leading to ureteral reflux and kidney damage (hydronephrosis). Poor bladder emptying contributes to recurrent urinary tract infections and increased risk of septicemia. The prevalence of detrusor-sphincter dyssynergia (DSD) in individuals with bladder dysfunction has been reported at over 90% (12), with 2 different functional presentations (12,13).
8
EXISTING THERAPIES FOR NEUROGENIC BLADDER
Standard Clinical Treatments
Current treatments for neurogenic bladder dysfunction and DSD encompass a
combination of pharmaceutical, mechanical, and surgical options. Treatment focuses
primarily on the suppression of overactive detrusor contractions to improve bladder
capacity, reduce leakage and incontinence episodes, and reduce the risk of autonomic
dysreflexia(14).
Pharmaceuticals
Detrusor contractions can be limited by anticholinergic drugs, which block acetylcholine from binding to muscarinic receptors in the bladder(15,16).
Anticholinergics are widely used and are generally successful at reducing bladder overactivity, decreasing urine leakage and high bladder pressures. However oral delivery of anticholinergics produces systemic side effects such as dry mouth, constipation, headache and cognitive impairment that limit patient compliance(17). Anticholinergics are not effective for relieving urethral sphincter spasms, and therefore only address the storage phase of micturition. Additional interventions are required for bladder emptying.
Catheterization
The gold standard of bladder emptying is intermittent catheterization, in which a stiff rubber tube is inserted into the urethra past the urethral sphincter and used to directly drain urine from the bladder. Intermittent catheterization must be performed every 4-6 hours in order to maintain safe bladder volumes; frequency depends on fluid consumption as well as individual bladder capacity. Maintaining sterility during catheterization is 9
challenging even under ideal conditions; attempting sterile catheterization while in public
restrooms is especially difficult and restricts patients’ social mobility.
When intermittent catheterization is inconvenient or impossible (due to lack of hand function, care-giver support, or urethral strictures) a permanent catheter line can be placed. Foley catheters remain in the urethra; suprapubic catheters enter directly into the bladder through the abdomen. Both types of permanent catheters are associated with persistent urinary tract infections (UTIs) and decreased bladder capacity and compliance when the bladder continuously drains. Severe UTIs require repeat hospitalization and intravenous antibiotics, and are a significant cause of patient morbidity.
Surgical Interventions
Surgical interventions to allow bladder voiding include physically cutting the urethral
sphincter (sphincterotomy) or transection of the pudendal nerve or sacral dorsal spinal
roots. These effectively eliminate reflexive sphincter contractions. However, they also
eliminate the patient's ability to remain continent, producing continuous urine leakage
that must be captured through condom catheters or diapers. Skin breakdown and social
discomfort are frequent side effects(18–20) . Sphincterotomies are not always effective
long-term and follow-up procedures are required. The surgical procedure itself carries
significant risks including hemorrhage and erectile dysfunction(19).
Sacral root transection surgeries are highly invasive, irreversible, and becoming
progressively less desirable as hopes of a regenerative medicine “cure” for spinal cord
injury spread. Chemical blocks of the pudendal nerve are an alternative to permanent
transection, but suffer many of the same drawbacks as sphincterotomies(21).
10
Neuroprosthetic Approaches to LUT Dysfunction
The techniques listed above are widely, clinically available but none are capable of restoring any semblance of the natural continence/ voiding dynamic. Neuroprosthetics are a niche field not widely known among urologists, but they provide the best opportunity for reproducing physiologically normal LUT function. A wide array of
neuroprosthetic techniques have been investigated with interfaces at nearly every location
in the pelvis (Figure 1.3). Only those interfaces related to this dissertation are discussed
in detail below.
Extradural mixed sacral roots: Brindley Vocare System
The Brindley-Finetech neuroprosthesis, marketed as the Vocare System, gives SCI
patients greater control over their bladder and bowel emptying. Electrodes are placed on
the S2-S4 extradural (combined) sacral roots to produce “on-demand” bladder
contractions for bladder emptying. The mixed innervation of bladder and sphincter
efferent fibers across spinal levels necessitates the use of intermittent stimulation, which
exploits the difference in relaxation time between the slow-twitch bladder fibers and fast-
twitch sphincter fibers. This produces “post-stimulus” bladder voiding(21), which can
still completely empty the bladder as long as dyssynergic sphincter contractions do not
occur(22,23). However reflexive sphincter contractions are almost always present as
long as the afferent pathway connecting urethral stretch receptors to the sacral cord is
intact.
Stimulation of the sacral anterior nerve roots without dorsal rhizotomy has been tried
in humans with limited success(24). Therefore successful voiding with the Vocare
11
system requires permanent rhizotomy of the S1-S4 dorsal fibers to eliminate reflexive
sphincter reflexes. Rhizotomy is beneficial in that it eliminates overactive bladder
reflexes, but it also eliminates any residual pelvic sensation or desirable reflex function
and is a major deterrent for potential implant candidates.
Sensory Stimulation for LUT Neuromodulation Neuroprosthetics
Sensory neuromodulation offers a novel method to control the spinal reflexes that
produce unwanted reflexes after spinal cord injury. The spinal circuits controlling
micturition rely entirely on sensory feedback from pelvic afferents. The nature of these
reflexes is that they are highly dependent on/ responsive to numerous sensory inputs, the
majority of which coalesce in the sacral spinal cord. Therefore, neuromodulation of these sensory inputs offers a novel way to interrupt the aberrant reflexes which can be turned on or off with greater precision than existing technologies to restore some degree of normal neurologic function.
There are significant differences between the afferent and efferent response to electrical stimulation. Electrical stimulation of efferent fibers has been used to restore motor function since the early 1970s, and the patterns required for muscle activation are now well understood(25). Afferent stimulation is a much newer field and researchers are still learning which input parameters are most important.
Sacral Root Stimulation
Sacral root neuromodulation has been clinically demonstrated to reduce symptoms of urinary frequency, nocturia and incontinence in individuals with neurogenic bladder
12 dysfunction(26–28). Some studies report improvement in non-obstructive urinary retention, but others report no improvement(29).
The Medtronic Interstim medical device, which as of 2013 has a reported over 40,000 of implants worldwide, has become the most widely available form of neuromodulation for urinary dysfunction(30). The device benefits from the use of a S3 foraminal electrode that can be placed in a short outpatient procedure without any additional surgery to access the sacral roots. Patients typically undergo a trial period before committing to have the full pulse generator implanted; the percentage of patients who opt not to continue with the therapy after the trial period is reported anywhere from 50 to 80% (31).
One of the leading complications of sacral root neuromodulation is lead migration, leading to a loss of effectiveness and requiring a more complication re-operative procedure to correctly re-place the lead. Trauma to the roots and lead infection are less common, but more serious concerns.
Despite widespread use and numerous clinical studies, it is still unclear what the mechanism of action of sacral root neuromodulation is. It is widely reported that Melzac and Wall’s original gate theory of pain plays some role in sacral root stimulation, but the specifics are still being determined.
Dorsal Penile Nerve Stimulation
The dorsal penile (clitoral) nerve is a superficial branch of the pudendal nerve that innervates the skin of the genitals. Dorsal penile nerve (DPN) stimulation can inhibit reflexive detrusor contractions, either through mechanical stimulation such a penile squeeze(32) or electrical stimulation(33,34). DPN stimulation is a non-invasive technique that can increase bladder capacity and decrease the incidence of urine leakage,
13
though it requires maintenance of surface electrodes on the penis and is difficult to apply in women. The specific stimulus pattern was not a crucial factor for suppression of the detrusor reflexes.
Posterior Tibial Nerve Stimulation
The posterior tibial nerve is a mixed sensorimotor branch of the sciatic nerve that runs from the fourth lumbar to second sacral (L4-S2) spinal levels down to the inner ankle.
Biweekly posterior tibial nerve stimulation (PTNS) with percutaneous needle electrodes is able to reduce symptoms of frequency, urgency and urge incontinence in women with neurogenic bladder not associated with central neurologic injuries. It has not been shown effective in SCI patients with overactive bladder(6,35–37).
The mechanism by which PTNS produces lasting suppression of detrusor overactivity is unknown, but it is worth noting that the stimulation targets sensory fibers at the ankle— a target so distant from the pelvis that it seems surprising it could have such a strong impact specifically on the bladder. The key to PTNS lies in the spinal levels where it ultimately connects to,
Sensory Stimulation for Urethral Reflex Suppression
All sensory neuromodulation techniques to date have focused on the suppression of overactive bladder activity or reflex bladder excitation. While this is an important component of healthy bladder management after SCI, it only addresses the storage component of normal lower urinary tract function. Restoration of bladder emptying requires a way to selectively address urethral sphincter dyssynergia. Therefore there is a need to develop a way to modulate sphincter reflex activity that is comparable to existing bladder neuromodulation techniques.
14
In the late 1990s Bhadra and Grunewald were investigating quasitrapezoidal stimulus
waveforms for the selective activation of small bladder motor fibers on mixed bladder-
sphincter ventral roots, though resulting voiding was usually poor(38). During these
experiments it was observed that the spinal intact dogs would sometimes void without
having undergone a dorsal rhizotomy(35, unpublished data) . This suggested that sensory
feedback being delivered through the afferent roots modified the sacral spinal micturition
circuits so as to produce coordinated bladder emptying. This was the first empirical
evidence that afferent feedback to the sacral cord could produce urethral sphincter
neuromodulation for a functionally relevant outcome.
As previously noted, sacral root stimulation is challenging due to the invasive
surgical access required—a factor that can severely limit early translation studies in
humans. Evidence of bladder reflex suppression from peripheral afferent nerves (deep
perineal nerve, posterior tibial nerve) suggested that a similar superficial afferent pathway might exist to target urethral sphincter reflexes. Data showing that cutaneous afferents in the perianal skin of kittens play a role in initiating the micturition reflex (including urethral sphincter relaxation)(40) gave further reason to examine surface stimulation as a potential proxy for sacral root neuromodulation.
Research in cats with acute SCI showed that surface stimulation over the sacral dermatomes (L7-S3) can significantly reduce measures of sphincter spasticity, namely pressure spike amplitude and spike rate(41). However, the acute SCI animal model does not adequately represent the significant neurological reorganization changes that take
place in the days and weeks following spinal cord injury. For example, these animals do
not display distention-evoked bladder reflexes, which arise only in a chronic preparation.
15
Additionally not all animals responded to surface stimulation—some showed an increase
or no change in response to stimulation.
Variability in response to sensory stimulation is not unusual, but sensory stimulation
can evoke different physiologic responses with seemingly small changes to the stimulus pattern(42–45). This raised questions whether a broader investigation of the (very large)
potential stimulus parameter space would improve the response rate in chronic SCI
animals.
HYPOTHESES AND AIMS
Aim 1: Demonstrate suppression of urethral reflexes using patterned sacral afferent stimulation in animals with chronic SCI
The goal of Aim 1 is to show that the spinal reorganizational changes known to happen after chronic SCI do not eliminate the urethral suppression response seen in the
acute SCI animal preparation. It is a significant jump from acute to chronic SCI animals,
both in terms of impact and experimental difficulty. Showing that patterned sacral
afferent stimulation can successfully suppress urethral reflexes in a large animal chronic
SCI model significantly increases our confidence that the technique can be translated to
humans with SCI.
Variable Hierarchy
Previous application of sacral afferent stimulation in acute SCI cats did not examine
the impact of several other parameter variables that we have identified as likely being
important. These parameters can broadly be grouped into the following categories:
16 stimulus location, stimulus pattern, and “additional state variables”. A more detailed analysis of the key variables to cover while testing chronic SCI animals is listed below.
1. Sensory input location
1.1. Dermatome (skin patch corresponding to the target sacral spinal level); expected
to be between L7 and S3
1.2. Unilateral or bilateral stimulation: though only one electrode was used to
suppress urethral reflexes in acute SCI cats, cutaneous input from only one side
of the body may not be sufficient and bilateral stimulation may be necessary
2. Stimulus parameters
2.1. Stimulus pattern features: Duty cycle, cycle time, base frequency, bursting
features
2.2. Electrode area: large electrode covering many dermatomes; small electrode with
very specific target; subcutaneous wire
2.3. Stimulus amplitude: confirm selective activation of cutaneous afferents rather
than muscle spindles or deeper afferent nerves (pudendal, pelvic)
3. Additional sacral sensory inputs
3.1. Bladder volume: suppression effect may only work for a limited range of bladder
volumes, depending on the competing afferent input from bladder stretch
receptors (especially when overactive detrusor reflexes are present)
3.2. Other afferent inputs that may be difficult to control: lower/ hind limb position,
other cutaneous inputs to skin, temperature, residual fiber tracts descending from
supraspinal input
17
Aim 2: Demonstrate improved voiding efficiency using patterned sacral afferent stimulation in chronic SCI animals
Aim 2 is dependent on the success of aim 1, but logically follows that if afferent
stimulation reduces urethral reflex pressures, bladder outlet pressure will decrease and voiding will improve. There is a possibility that the isovolumetric conditions tested in aim 1 will be not match voiding tests in aim 2, due to the addition of fluid flow in the urethra adds an additional sensory input that was not previously present. However, the presence of the catheter in the urethra in aim 1 stimulates urethral afferents and it is unlikely that the response to fluid movement will be significantly different.
Once bladder voiding has been demonstrated in anesthetized animals, voiding tests can be conducted in awake-behaving animals to verify the success of patterned sacral afferent stimulation under “real world” conditions. If acute, awake voiding tests are well tolerated and afferent stimulation produces clinically acceptable voiding, we will move to electrically void the animals over a continuous multi-day period. Sustained bladder maintenance using patterned sacral afferent stimulation in awake-behaving animals will provide strong evidence that PSAS can be used as a valuable neuroprosthetic technique similar to what would be used in humans.
Aim 3: Determine if patterned sacral afferent stimulation can reduce abnormal urethral reflexes in human subjects with chronic SCI
It is necessary to demonstrate EUS reflex suppression in at least one individual as a
“proof of concept” that the appropriate afferent mediated pathways exist in humans as well as cats. Although we do not expect that humans and cats will respond identically to
18
patterned sacral afferent stimulation, we can leverage knowledge gained in animal
experiments to design appropriate stimulus parameters and locations to being testing in
humans.
We will not have the luxury of implanting electrodes for bladder drive into our human
subjects. Therefore we will need to modify standard urodynamics tests to generate LUT
reflexes and simultaneously apply PSAS to measure the effect of stimulus parameter and
location on bladder and urethral pressures in SCI individuals.
IMPACT
Afferent stimulation of the sacral dermatomes offers a novel method for non-invasive,
non-destructive restoration of bladder emptying. Demonstration that afferent stimulation
produces functional bladder voiding in chronic SCI animals would justify additional
studies investigating the mechanism of action and work to develop a neuroprosthetic
delivery device.
If successfully translated into human subjects it could increase the number of people
who benefit from neuroprosthetic bladder control. There is a potential for cost savings
with a one-time device purchase over the ongoing cost of catheters, but the biggest
savings would come from a reduction in urinary tract infections and associated hospital
admissions. A neuroprosthetic device that frees SCI individuals from catheterization
(which makes bladder maintenance in public places difficult and limits their ability to travel) could contribute to feelings of independence and greater quality of life.
19
Figure 1.4 Neurophysiology of the lower urinary system
Simplified neural network diagrams showing the primary connections controlling continence (a) and voiding (b). The Pontine storage and micturition centers are both located in the pons of the brain stem.
Figure reproduced from: N. Yoshimura, W.C. de Groat Neural control of the lower urinary tract. Int. J. Urol., 4 (1997), pp. 111–125. Republished with permission from John
Wiley and Sons, licence # 3142800818370
20
Figure 1.5 Bladder and sphincter traces for an infant, normal adult, and adult with
detrusor-sphincter-dyssynergia
Bladder and external urethral sphincter contractions are reciprocal in spinally
intact individuals (a,b). Infants have underdeveloped supraspinal control over
micturition, and bladder emptying is primarily driven by spinal reflexes, yet the
synergic relationship between bladder and sphincter is maintained (a). Adult
individuals have an additional level of cognitive control over micturition (b).
Individuals with supra-spinal injuries revert to a reflexively controlled system, but 21 the relationship between bladder and sphincter becomes disordered and simultaneous contractions emerge (detrusor-sphincter dyssynergia, c). Each panel shows the fluid detrusor pressure with a corresponding electromyographic recording of the external urethral sphincter muscle.
Figure reproduced from: De Groat WC. Integrative control of the lower urinary tract: preclinical perspective. British journal of pharmacology. 2006 Feb;147
Suppl S25–40. Republished with permission from John Wiley and Sons, License number #3142820177447
22
Figure 1.6: Electrode interface locations for controlling the lower urinary tract
(A) Intravesical, (B) Bladder wall, (C) Thigh, (D) Pelvic floor, (E) Dorsal penile nerve,
(F) Posterior tibial nerve, (G) Pelvic nerve, (H) Intradural sacral anterior root, (I)
Extradural mixed sacral root, (J) Intradural sacral posterior root, (K) Sacral nerve, (L)
Spinal cord, (M) Intraurethral, (N) Pudendal nerve, (O) Sacrum.
Figure reproduced from: Gaunt RA, Prochazka A. Control of urinary bladder function with devices: successes and failures. Progress in brain research. 2006;152:163–94.
Republished with permission from John Wiley and Sons, license # 3142860811653
23
CHAPTER 2: Electrical Stimulation of Sacral Dermatomes Suppresses Aberrant Urethral Reflexes After Chronic Spinal Cord Injury
This chapter has been published:
McCoin, J. L., Bhadra, N. and Gustafson, K. J., Electrical stimulation of sacral
dermatomes can suppress aberrant urethral reflexes in felines with chronic spinal cord
injury. Neurourol Urodyn 2013 Jan; 32(1): 92–97. doi: 10.1002/nau.22276
Reprinted here with permission from John Wiley & Sons, license # 3134021198540
24
ABSTRACT
Aims: Uncoordinated reflex contractions of the external urethral sphincter (EUS) are a
major component of voiding dysfunction after neurologic injury. Patterned stimulation of
sacral afferent pathways can reduce abnormal EUS reflexes after acute SCI; however, effectiveness following chronic SCI is unknown.
Methods: Four adult male cats were implanted with bilateral extradural sacral root electrodes to allow bladder activation and underwent subsequent spinal transection (T10-
12). Nine weeks after SCI urethral and bladder pressures were recorded with and without sacral afferent stimulation. Surface electrodes were applied to sacral and lumbar dermatomes and stimulus amplitude set below the muscle fasciculation threshold. The stimulation pattern was varied by on/off times of fixed frequency at each location.
Results: Reflexive EUS contractions were observed in all animals after chronic SCI.
Patterned sacral dermatome stimulation reduced EUS reflex rate and amplitude in 2 of 4
cats. Suppression was dependent on both the stimulus location and pattern. Sacral
locations and a stimulation pattern of (0.75 seconds on, 0.25 seconds off, 20 Hz) were
effective in both responder animals.
Conclusions: Patterned sacral dermatome stimulation can reduce urethral abnormal reflexes following chronic SCI. Reflex suppression is dependent on both the stimulation
location and stimulus pattern. Reduction of reflexive EUS activity after chronic SCI with
this non-destructive and non-invasive approach may provide an advance for the treatment of detrusor-sphincter-dyssynergia.
25
INTRODUCTION
Clinical Impact of SCI on Micturition
Bladder hyperreflexia and detrusor-sphincter-dyssynergia (DSD) affect over 90% of individuals with spinal cord injuries (46). In the urinary tract, fluid flow in the urethra incites a “guarding reflex” that increases urethral sphincter tone to prevent leakage. This reflex is normally suppressed during voluntary micturition, however is disorganized after
SCI (12). Abnormal external urethral sphincter (EUS) reflexes can obstruct voiding. The resulting high-pressure or high-residual voiding reduces patient quality of life, incurs high medical costs, and can produce long-term damage to the bladder and kidneys if not managed effectively (46). Present clinical standards of bladder maintenance include clean intermittent catheterization, anticholinergic pharmacotheraputics, sphincterotomy, urethral stents, or pudendal nerve block (47,48). Such management methods are limited by factors such as severe spasticity, poor upper limb function, intolerable systemic effects or urinary incontinence (48).
Electrical Stimulation for Control of LUT Function
The advent of sacral anterior root stimulation combined with dorsal rhizotomy in the
1970s provided a means of abolishing reflexes and producing bladder evacuation (22,23).
It has resulted in an improved quality of life for many; however, rhizotomy that eliminates residual bowel and sexual functions is undesirable and limits wider patient acceptance (23,49). A means of obtaining effective bladder evacuation without interfering with any existing neural capacities would improve Lower Urinary Tract
(LUT) management after neurological injury.
26
Electrically stimulating afferent pathways can interrupt or excite LUT reflexes, such
as inhibition of bladder hyperreflexia. Sensory stimulation of the dorsal genital nerve
(34,50), posterior tibial nerve (51), and sacral spinal nerves (24,52) all provide inputs to
the lumbosacral spinal cord and produce functional improvements in bladder capacity
and low pressure urine storage. Reflex bladder contractions can be activated using
stimulation of pudendal afferent fibers (53,54). The amplitude of the evoked bladder
contractions is dependent on the stimulus frequency (44,55). In addition, different
stimulus patterns are more effective at different locations within the urethra (43,56).
It is unknown whether afferent stimulation can specifically inhibit urethral sphincter
reflexes. Our group previously demonstrated that patterned dermatome stimulation
reduces spastic urethral spike-activity in cats after acute spinal transection (41). Surface stimulation of sacral dermatomes provides a reversible, non-invasive technique that may be able to interrupt the aberrant urethral reflex pathway while retaining desirable bowel and sexual reflexes that are lost with rhizotomy. To be of clinical relevance, sacral surface stimulation must be able to suppress urethral sphincter spasms after chronic SCI, with the resultant changes in the spinal circuits (57). This study sought to determine if afferent reflex suppression can be achieved following neural reorganization in cats after chronic SCI.
MATERIALS & METHODS
Chronic SCI animal model
Four sexually intact adult male cats had extradural sacral electrodes implanted for on-
demand bladder drive; a second surgery to transect the spinal cord was carried out after
27 the electrodes had stabilized and animals were deemed behaviorally suitable for chronic
SCI maintenance. All procedures were carried out under general anesthesia (ketamine induction, isoflurane maintenance), with prior approval from the Case Western Reserve
University IACUC. Animals first underwent a laminectomy at the L7-S2 level. Tripolar spiral cuff electrodes (Ardiem Medical, Indiana, PA) of 1.25 mm diameter were implanted on the extradural sacral roots eliciting the greatest bladder pressures intraoperatively (verified S2 roots post mortem). Electrode leads were tunneled subcutaneously to exit in the inter-scapular region. Animals were fitted with jackets
(Lomir Biomedical Inc., Malone, NY) to protect the lead exit sites.
Animals underwent surgical spinal transection at the T10-T12 vertebral level (after 17 weeks in animals 1-2, 6 weeks in animals 3-4). The dura was exposed through a laminectomy, and the cord cut through a small incision made in the dura with local application of intradural Marcaine .25% (Hospira Inc., Lake Forest, IL). Completeness of transection was visually confirmed prior to closure. Following recovery, animals received manual bladder expression 2-3 times daily; reflex defecation circumvented the need for assistive bowel care. A 9 week survival period established a chronic model of abnormal LUT reflexes.
Post mortem dissection was carried out to confirm the spinal root levels, identify any damage to the sacral root or electrode implants, and verify completeness of spinal transection.
Terminal Procedure
Nine weeks post-SCI animals were anesthetized with an IV infusion of alpha- chloralose (75 mg/kg induction, 19 mg/kg supplemental maintenance as needed) (Sigma
28
Alderitch, St. Louis, MO). Sub-cutaneous buprenorphine (.01 mg/kg) was given every
12 hours. Each animal was instrumented with a suprapubic bladder catheter for bladder
filling and draining, and measuring bladder pressure. External sphincter and proximal
urethral pressures were measured using a 3.5 French catheter (Gaeltec, Isle of Skye,
Scotland) mounted with two microtransducer pressure sensors placed into the urethra.
The transducer was zeroed to atmospheric pressure prior to use. The gain of the
transducer was calculated using a mercury manometer and two-point linear slope. Active
urethral pressure profiles (UPP) were conducted to determine the position of the EUS
from the external meatus.
Sacral dermatome levels more effectively deliver sensory stimulation to the spinal
circuits generating unwanted urethral reflexes (41,58). To investigate surface localization
effects, the L4 through S4 dermatomes were shaved and prepped with a medical depilatory prior to application of surface electrode patches. Dermatome locations were estimated from published dermatome maps (59,60). “Large” surface electrodes
(4cmx4cm square, Re-Ply Unipatch; Covidien, San Francisco CA) were used for gross dermatome localization effects (lumbar vs. sacral dermatomes). “Small” electrodes (2
cm diameter round, Cardinal Health) were used for finer resolution within lumbar and
sacral dermatomes. Figure 1 shows placement of large and small electrodes for spatial
differentiation and the general pattern of electrical stimulation.
29
Data Collection: Stimulation Protocol
Bladder contractions were evoked using 5-10 seconds of 20 Hz stimulation on sacral
root electrodes. Control trials consisting of bladder drive without any dermatome
stimulation were used to consistently evoke EUS reflex activity. Baseline EUS spasticity
was defined as the pressure spikes following root stimulation.
Dermatome stimulation consisted of monophasic constant-current pulses of 100 µs
(DS7A, Digitimer, Hertfordshire, England). Afferent neurons typically utilize bursts of
action potentials with varying burst durations, frequency and inter-burst intervals
suggesting that particular stimulation patterns will provide more effective reflex
modulation than others. We used cycle time (s), duty cycle (%), and base frequency (Hz) to generate patterned stimuli. Cycle time was fixed at 1 s for all patterns. Duty cycle was varied (25%, 50%, 75%, 100%) for a fixed base frequency (20 Hz). Cycle time and duty cycle are simultaneously described though stimulation ON time and OFF time (Figure 1).
At 100% duty cycle, representing continuous stimulation, base frequency was varied between 10, 20 and 40 Hz for all animals (except animal 2, which did not respond to 20
Hz). Stimulation amplitude was determined by visible muscle fasciculation; all trials reported were conducted at 90% of fasciculation threshold.
Voiding improvement was investigated by adding (.75 ON .25 OFF, 20 Hz) dermatome stimulation to intermittent sacral root stimulation (2.0 ON 4.0 OFF, 20 Hz) in one animal (#2). Surface stimulation was limited to the “middle” 2cm round surface electrode (electrode [5], Figure 2.1). Voiding was limited to 4 paired, bladder volume matched control/ dermatome runs. Voiding percentages were calculated for each run; voiding to completion was not attempted.
30
In the same animal a 22 gauge stainless steel needle was inserted through the same
location and stimulated with (.75 ON .25 OFF, 20 Hz) to test the effect of on urethral reflexes. All data was recorded in a custom designed data-acquisition program (Labview,
National Instruments, Austin, TX).
Data Analysis
Urethral sphincter spasms were observed as EUS pressure “spikes” in all animals, due
to high-fidelity microtransducer recording (Figure 2.2). Spike rate and spike amplitude
were used as the primary measures of reflex EUS activity (41). Spikes were defined as
pressure increases 1 standard deviation above a 0.5 second moving average, quantified in
Matlab (Mathworks, Natick, MA). Absolute pressure was used rather than pressure
evoked above baseline. Baseline urethral pressure was consistent at ~15 cmH2O.
Variables reduced from identified spikes include: Pur= Average urethral spike pressure amplitude (cmH2O) and Rur= Average urethral spike rate (number/sec) calculated over the period of dermatome stimulation (typically 60 seconds), with a corresponding length chosen for comparative control trials.
For each animal one-way ANOVAs were used to analyze the effect of dermatome stimulation on reflex activity when compared to control trials. Additionally ANOVAs were calculated for stimulus patterns at each electrode location, and across locations for each pattern. The Tukey-Kramer method was used to determine which parameter combinations differed significantly; only reductions from baseline which reached (P<.05) significance were labeled “suppression”.
31
RESULTS
Reflex Activity Following Chronic SCI
Prior to SCI, no animals displayed urethral reflex activity under light isoflurane anesthesia. Neither EUS nor bladder activity was evoked when applying dermatome stimulation at sub-fasciculation (4.5-12.5 mA) threshold levels, and current amplitudes high enough to produce urethral pressure changes were inseparable from motion artifact due to hind limb muscle contractions.
The bulbocavernosus reflex reappeared in all animals within 12 hours after SCI and
LUT reflexes appeared within 1 week. All animals displayed distention evoked bladder reflexes and the progressive development of spastic urethral sphincter reflexes that were observed under light isoflurane or alpha-chloralose. These EUS reflexes were consistently evoked following sacral root stimulation, however they also occurred spontaneously or with urethral catheter movement. Baseline reflex activity with effective and ineffective patterned surface stimulation are shown in Figure 2.
Three animals had confirmed complete transection of the spinal cord. Animal 1 had partial voluntary stepping and visceral sensation caudal to transection, suggesting incomplete injury. All animals had at least one working sacral root electrode capable of producing greater than 50 cm H2O intra-vesicular pressure and greater than 70 cm H2O urethral sphincter pressure. The right S2 root was damaged in animal 2 during the week after spinal transection.
Effect of Dermatome Stimulation on Urethral Reflex Suppression
Significant, reproducible suppression of urethral reflexes was achieved in 2 of 4 animals when effective stimulus parameters were used (Figure 2.2). Reflex reduction
32
was observed within 2 seconds of the start of dermatome stimulation, and was able to be
maintained for over 60 seconds. Statistically significant suppression was only achieved
for a limited combination of electrode locations and stimulus patterns. Urethral reflexes
were reduced by dermatome stimulation at all locations tested in responder animals (1
and 2), but only sacral dermatome locations significantly reduced urethral activity
compared to control and were classified as suppression (Figure 3).
Rur was reduced 92.6% and Pur reduced 61.4% (P<.001) using the large electrode
location [S] (animal 1, 20 Hz continuous). Sacral [S] stimulation with (.75 ON .25 OFF,
20 Hz) produced comparable suppression to 20 Hz continuous stimulation (see Figure 4);
20 Hz continuous data are presented as lumbar stimulation [L] was not attempted using
(.75 ON .25 OFF, 20 Hz).
Finer resolution of the sacral skin beneath [S] using smaller electrodes [1-5] (animal
2) demonstrated finer spatial resolution within the sacral dermatomes. Locations [4] and
[5] both significantly suppressed urethral reflexes. Location [5] reduced Pur 80.8% and
Rur 86.1% (P <.001) using (.75 ON .25 OFF, 20 Hz). Suppression was not achieved at electrodes [1-4] for any other patterns tested (data not shown).
Data shown in Figures 2.3 and 2.4 are normalized to the control for each animal. The
Pur and Rur, control values for all animals are shown in Table I. The pooled values for
all dermatome stimulation trials are also shown for animals 3 and 4, where no differences
were observed between surface stimulation and control for any stimulus locations or
patterns tested. Only (.75 ON .25 OFF, 20 Hz) patterned stimulation produced
suppression in both responder animals, when applied at effective locations. Increasing
duty cycle from 25% to 75% led to progressive reductions in spike rate Rur and
33
amplitude Pur, however only 75% produced suppression compared to control and were
classified as suppression. Figure 4 shows patterned stimuli tested at electrode locations
[S] (animal 1) and [5] (animal 2). Data are normalized to the control for each animal.
Continuous (20 Hz) stimulation produced suppression in animal 1, but not animal 2.
In animal 1, 10 Hz continuous reduced Rur by 82.4% (P<.001) and Pur by 28.2%, (P=
.053) and 40 Hz continuous reduced Pur 38.1% (P=.289) and Rur 60.1% (P= .084).
Bladder emptying was improved from 30.3%±3.8% (8.95± 2.2mL) without dermatome stimulation to 41.93% ±6.6% (12.43 ± 2.9 mL) with stimulation (P<.05) in single bladder drive-limited stimulus runs. When normalized to control voided volumes, this represents a 38.3% ±21.7% increase with the addition of effective surface stimulation. Bladder volume for all trials was 29.3 mL ±3.5 mL.
Subcutaneous stimulation reduced Rur (P<.01) and Pur (P<.001) when compared with
control trials. Subcutaneous stimulation was not significantly different from surface
stimulation using the same (.75 ON .25 OFF, 20 Hz) pattern (Rur (P= .362), Pur (P=
0.771)). In 7 subcutaneous trials, (Rur = 0.0208 ±.019 /sec; Pur = 83.95 ± 88.21 cm
H2O).
34
DISCUSSION
Reduction of Urethral Spasm after Chronic SCI
This study provides the first evidence that patterned sensory stimulation of sacral
dermatomes can reduce spastic urethral activity after chronic spinal cord injury. These
data provide a key step in validating previous results after acute SCI (19) and justify
work to evaluate the voiding effectiveness of this novel approach to treat urethral
sphincter spasticity.
Location of the stimulation electrode significantly impacts the extent of reduction of
EUS spasticity, both at a large-scale level (lumbar vs. sacral dermatomes) and within the
sacral dermatomes (see Figures 2.1 and 2.3). Effective skin locations corresponded to the expected overlap of the L7 through S2 dermatomes, while higher lumbar levels or levels
below S2 were ineffective. Location specificity is consistent with our expectation that
sensory stimulation must be directed to specific spinal interneurons in the sacral spinal cord. The skin locations are consistent with neurophysiology studies placing spinal circuits controlling the external sphincter in cats at the S1-S2 level (58). Pharmacological work in spinal transected rats (61) demonstrating that local delivery of GABA agonists to the lumbosacral spinal cord could eliminate dyssynergic EUS contractions suggests that afferent stimulation could similarly act through a GABAergic inhibitory response.
EUS reflex reduction is dependent on the electrical stimulus pattern. Specific patterns reduced aberrant sphincter reflexes while others were ineffective. (.75 ON .25
OFF, 20 Hz) was effective in both responder animals, while 20 Hz continuous was only effective in one animal. This consistency suggests that pattern is an important factor.
Other stimulus features such as pulse width or wave shape may provide more optimal
35
sensory input. While stimulation location and pattern significantly affected the ability to
reduce urethral sphincter activity, these specific locations and patterns may not generalize to all animals.
Stimulus parameters found effective after chronic SCI were not tested in acute SCI, but the (.5 ON .5 OFF, 20 Hz) that was successful in acute animals (41) did not have a significant effect after chronic SCI. This suggests that the changes in the spinal reflex
pathways from after chronic SCI could impact the afferent input required for reflex
suppression. Demonstration that patterned afferent neuromodulation reduces urethral
spasms and improves functional outlet resistance after SCI is the first step towards
realizing a functional neuroprosthesis. Although complete bladder emptying was not
attempted, improvement in the first run (of a potential series) demonstrates that a
reduction in EUS spike pressure and frequency is functionally relevant to reducing the outlet pressures in animals with DSD. Evidence that a wire in the subcutaneous space can reduce sphincter pressures equivalent to surface stimulation, combined with existing
afferent neuromodulation technology for bladder inhibition, suggests a fully implantable
system is feasible.
Non-Responder Animals
Reflex suppression was only obtained in 2 of the 4 animals, which is consistent the
50% (3 of 6) success rate previously demonstrated in acute SCI animals (41). There is no clear pattern in the baseline activity of the four animals that would suggest why
suppression was achieved in some and not others. Other applications of afferent
stimulation have shown similar variability in individual responses (14,15,17). It is
currently unknown if we have not appropriately mapped the parameter space and missed
36
the correct input, if dermatome overlap results in too diffuse spinal input, or that the
neurophysiologic pathways were different in the non-responder animals after SCI (62).
CONCLUSIONS
This study demonstrates proof-of-concept that afferent stimulation of the sacral
dermatomes can significantly reduce urethral reflexes after chronic spinal cord injury.
Suppression is dependent on the location and pattern of the sensory input. Stimulation of
sacral dermatomes provides a minimally-invasive method for reducing urethral sphincter
spasms after chronic spinal cord injury.
37
Table 2-I: Control and Summarized dermatome data from individual animals
Table 2-I lists the individual and pooled baseline urethral reflex measures for all four
animals. Urethral spike rate (Rur) was significantly different for all four animals (P<.05).
Urethral reflex amplitude (Pur) was significantly different only for animal 2 (P<.05).
Pooled data are shown for all dermatome trials for non-responder animals 3 and 4. No difference was seen between baseline activity and any subset of dermatome stimulation.
**The pooled data for responder animals 1 and 2 is not shown, as pooling obscures the
effect of pattern and location on reflex activity. These data are more acurately shown in
Figure 2.3 and Figure 2.4.
38
Figure 2.1: Surface electrode placement and patterned stimulus waveform
A: Large 4 cm square electrode patches were used to stimulate sacral dermatomes [S]
(L7-S3) and lumbar dermatomes [L] (L4-L5). Smaller 2 cm round patches were used to improve spatial resolution within the sacral dermatomes [1-5]. Return patch [R] for anodal stim was used for all electrodes. Feline dermatome map adapted from Kuhn (59), showing the approximate boundaries and overlap of L4-S4 dermatomes. Electrode [S] overlaps the L7-S3 dermatomes; [1]: L5-L7; [2]: L7/S2; [3]: L7-S1; [4]: S3; [5]: L7-S2 dermatomes. B: The stimulus waveform was defined in terms of ON/OFF time, base frequency, and amplitude. On/ off times always summed to a total cycle time of 1 sec.
Base frequency was fixed at 20 Hz for duty cycles less than 100%; other base frequencies were tested for continuous stimulation.
39
Figure 2.2: Suppression of aberrant urethral reflexes after chronic SCI
Representative examples of aberrant urethral reflex activity with and without dermatome stimulation. All panels: Sacral root stimulation (20 Hz, solid bar) was applied to directly evoke bladder pressure and provoke DSD in a consistent, repeatable manner. Panel A:
Control trial showing baseline reflex activity. Panel B: The addition of patterned electrical stimulation to the sacral dermatomes (.75 ON .25 OFF, 20 Hz, dashed bar) suppressed the aberrant EUS reflexes. Panel C: Applying dermatome stimulation to the same electrode with different stimulus pattern (.5 ON .5 OFF, 20 Hz, dashed bar) did not significantly reduce urethral activity. This parameter space specificity is representative across both responder animals.
40
Figure 2.3: Spatial selectivity of stimulus location
Sacral dermatome stimulation reduced urethral reflexes. Data from large lumbar [L] and sacral [S] electrode locations (Black symbols, animal 1) show sacral locations [S] reduce
41 urethral reflexes. Finer resolution at location [S] using smaller electrodes [1-5] (gray symbols, animal 2) demonstrates further location specificity within the sacral dermatomes. Data are normalized to the baseline control evoked urethral activity Pur and
Rur. EUS reflex spike rate (Rur) is shown on the left axis (). EUS reflex spike amplitude is shown on the right axis (). Stimulus pattern shown for [L] and [S] is 20
Hz continuous; for [1-5] is [.75 ON .25 OFF, 20 Hz]. The number of trials (n) is shown for each location. *P<.05; ** P<.01.
42
Figure 2.4: Stimulus pattern affects reflex suppression
Urethral reflex suppression was dependent on stimulus pattern. Only (.75 ON .25 OFF,
20 Hz) patterned stimulation was effective in both responder animals. Increasing duty
43
cycle led to progressive reductions in spike rate Rur and amplitude Pur. 20 Hz continuous stimulation was effective only in animal 1. Data are normalized to the baseline control evoked urethral activity Pur and Rur. EUS reflex spike rate (Rur) is shown on the left axis (). EUS reflex spike amplitude is shown on the right axis ().
Data from animal 1 (black symbols, location [S]) and animal 2 (gray symbols, location
[5]) are shown separately for each stimulation pattern. The number of trials (n) is shown for each location. *P<.05; ** P<.01.
44
CHAPTER 3: Patterned Sacral Afferent Stimulation Produces Clinically Effective Voiding in Chronic SCI Felines
This chapter has been submitted and is under review:
McCoin, J. L., Bhadra, N. and Gustafson, K. J., Patterned Sacral Afferent Stimulation
Produces Effective Bladder Voiding in Chronic SCI Felines. Neurourol Urodyn, April
2013
45
ABSTRACT
Aims: Uncoordinated reflex contractions of the external urethral sphincter (EUS) prevent
efficient bladder emptying in individuals with spinal cord injury (SCI). Patterned
electrical stimulation of sacral afferent pathways can reduce EUS reflexes. This study
evaluated the ability of sacral afferent stimulation to achieve complete voiding and
replace standard bladder maintenance methods after chronic SCI.
Methods: Four male cats were implanted with bilateral extradural sacral root electrodes
for direct bladder activation and underwent subsequent spinal transection (T10-12). EUS
reflex suppression and bladder voiding was evaluated every 2 weeks under anesthesia.
Animals that displayed effective reflex suppression were voided with and without sacral dermatome stimulation. These animals were then electrically voided awake, replacing manual expression as the primary method of bladder maintenance.
Results: Sacral afferent stimulation suppressed urethral reflexes in 2 of 4 cats. In these
two animals EUS reflex rate decreased from 0.129 ± 0.014 to 0.016 ± 0.012 spikes/
second and amplitude decreased from 141.8 ±14.8 to 80.2 ± 15.4 cmH2O. Voiding
improved from 48.3 ± 11.7% to 83.9 ± 6.2% under anesthesia, and from 54.3±17.9% to
84.6±10.9 % while awake. Electric voiding was sufficient to replace manual expression
and both animals were maintained using only electric voiding (86.7 ±10.1%) during
weekdays for 1-3 weeks.
Conclusions: These are the first data to demonstrate that patterned sacral afferent
stimulation can produce effective voiding in animals with chronic SCI. Repeated daily
awake voiding demonstrates that it is possible to clinically maintain animals with chronic
SCI solely with this approach.
46
INTRODUCTION
A majority of individuals with thoracic or higher level spinal injuries will develop
bladder hyperreflexia and detrusor-sphincter-dyssynergia (DSD) following injury(12).
Uninhibited spinal reflexes produce frequent low-volume bladder contractions and external urethral sphincter (EUS) spasms that prevent efficient voiding. Urethral stents, sphincterotomies(18), and indwelling or intermittent catheterization fail to adequately address the neurologic component of sphincter spasm(63,64). Numerous neuroprosthetic approaches targeting detrusor hyperactivity have been developed, but a method for suppressing the somatic external sphincter is still lacking(65). A neuroprosthetic device that could selectively inhibit sphincter spasms and allow bladder voiding would improve current bladder management techniques. Neuromodulation of the sacral spinal centers provides such an approach.
Following the disruption of descending inputs from the pontine micturition center, bladder-sphincter reflex arcs become disordered and dependent entirely on local afferent inputs. Normally silent C-fiber afferents become sensitized to bladder distension and cause reflex contraction at low volumes(66). Subsequent fluid flow within the urethra excites a pudendal-pudendal reflex which, without supraspinal suppression, will cause continuous external sphincter spasm(64). It is possible to interrupt this pudendal reflex through dorsal rhizotomy(22), high frequency electrical nerve block(67), or afferent feedback at the dorsal sacral root level(39).
The anatomical connections between cutaneous afferents in the sacral dermatomes and the sphincter motoneurons have not yet been established; however the abundance of polysynaptic connections between sphincter motoneurons and perineal cutaneous
47
afferents at the S1-S2 spinal level(68,69) suggests that electrical stimulation of these
afferent pathways may provide an alternate approach to inhibit sphincter motoneurons in
a non-invasive way. A growing body of evidence in cats suggests that delivering
appropriately patterned input to the correct spinal level (via targeted skin locations) can
directly suppress external urethral spasms after both acute(41) and chronic SCI(71). This
reduction in outlet pressure should lead to significant improvement in urine output when
combined with reflexive or electrically-driven bladder pressure.
The objectives of this study were to determine if patterned sacral afferent
(dermatome) stimulation (PSAS) could be used to produce bladder voiding in both
anesthetized and awake-behaving cats following chronic SCI and to determine if PSAS
can temporarily replace clinically accepted methods of bladder maintenance in these
animals.
MATERIALS & METHODS
Chronic SCI animal model
Four sexually intact adult male cats had extradural sacral electrodes implanted to generate on-demand bladder pressure. Spinal transection was carried out after electrode performance was verified and animals were deemed behaviorally suitable for chronic SCI
(6 weeks post-implant). Animals underwent a terminal experiment under alpha- chloralose anesthesia 9 weeks post SCI; a suprapubic bladder catheter was placed to measure bladder pressures during the terminal experiment. All surgical procedures were carried out under general anesthesia (ketamine induction, isoflurane maintenance), with prior approval from the Case Western Reserve University IACUC. Details of the surgical
48
procedure have been described previously (71). Data from the animals in this study have
not been published elsewhere.
Urethral Reflex Suppression
Animals were tested every two weeks under light propofol IV anesthesia (PropoFlo
28, Abbot Animal Health; 1 mg/kg induction, 0.4-0.6 mL/min maintenance, 10 mg/mL)
beginning one week post spinal transection. Urethral reflexes were continuously
monitored using a high-fidelity 3.5 French microtransducer catheter (Gaeltec, Isle of
Skye, Scotland). Stimulation parameters and paradigms were consistent with methods reported previously(71). Sacral (L7-S3) dermatomes were shaved and prepped with a medical depilatory prior to application of surface electrode patches. Effective dermatome locations were found by probing a 2 cm round Ag/AgCl disk electrode (Cardinal Health,
US) across the sacral dermatomes thinly coated with electrolytic gel (Spectra 360, Parker
Laboratories). Stimulation patterns previously found to produce effective urethral reflex suppression(71) were delivered to the sacral dermatomes, typically (0.75 sec ON, 0.25 sec OFF, 20 Hz) and 20 Hz continuous stimulation. Stimulation amplitude was set to
90% of the visible muscle fasciculation threshold. Skin locations that visibly reduced
EUS spike activity were marked and surface patch electrodes (2 cm diameter round,
Cardinal Health) were placed over the effective location. The target electrode was bounded by additional electrodes in the rostral/caudal and medial/lateral directions to
verify spatial selectivity.
49
Voiding Protocol: Voiding Under Anesthesia
Bladder contractions were produced using sacral root stimulation (20 Hz, 2 s ON, 4 s
OFF) to obtain consistent, reproducible bladder pressures. Intermittent stimulation is
utilized in existing human neuroprostheses because sacral stimulation activates both the
bladder and sphincter; voiding occurs during the stimulation OFF periods(22,24).
However urethral reflexes after SCI prevent voiding and sacral dorsal rhizotomy is
required to achieve clinically acceptable voiding in humans(24).
Bladder voiding was evaluated during bladder activation with and without PSAS.
Voiding was initially tested under propofol anesthesia before being tested in awake-
behaving animals. During anesthetized tests bladder and sphincter pressures were
evaluated under isovolumetric conditions prior to voiding. Figure 1 illustrates the typical
bladder pressure and voiding pattern produced by the control and afferent voiding
paradigms.
Awake Voiding
Unanesthetized tolerance for sacral root stimulation was verified before awake voiding
was conducted. Intermittent root stimulation was delivered for 60 to 90 seconds in each
run. Runs were repeated 5 (range 3-7) times with 3 minutes of rest in between to allow recovery of any detrusor muscle fatigue. Voided volumes were measured in mL for each stimulation run. Residual volumes were calculated for anesthetized voids via intraurethral catheterization. Residual volumes for awake voiding trials were found via manual expression following electric voiding, or by bladder palpation during the bladder maintenance phase. Palpation estimates were calibrated during acute testing sessions with known bladder volumes.
50
Daily Maintenance Voiding
Electric voiding was considered successful if the total percentage emptied was greater
than or equal to the average manually expressed percentage for each animal. Manual
expression is the clinical standard of care for maintaining chronic SCI cats. After
demonstrating successful awake voiding with PSAS, an animal was transitioned to the
electric maintenance phase replacing twice daily manual expression with electric voiding as the primary method of bladder emptying. Electric maintenance was suspended temporarily when animals were transferred out of the facility on weekends, holidays, and
for bi-weekly anesthetized testing sessions. Animals were voided twice (range 1-4) daily with electrical stimulation separated by 8-12 hours. Voiding was repeated at shorter
intervals in cases of high urine output or large estimated residuals.
Subcutaneous Stimulation and Cutaneous Anesthesia
During the terminal test in animal 4, afferent stimulation was delivered subcutaneously
through a 22 gauge hypodermic needle inserted parallel to the skin surface immediately
beneath the effective surface location (left side). Voiding with needle stimulation was
conducted in 2 sets of control and afferent runs following successful reflex suppression.
In the same experiment lidocaine gel (Topicaine 4%, ESBA Laboratories) was applied
topically to the effective dermatome location (right side) to produce local anesthesia of cutaneous afferents. The gel was removed after 10 minutes, and surface stimulation reapplied to the effective area. Standard reflex measures were compared for (.75 ON, .25
OFF 20 Hz) stimulation afferent suppression before and after lidocaine anesthesia.
51
Data Analysis
Reflex activity was quantified by pressure spike rate and spike height. Each metric
was compared between control conditions (without PSAS) and with PSAS and analyzed
as described previously(71). Voiding efficiency was compared between conditions
(intermittent root voiding with and without PSAS, and manual expression) using bladder volumes and total percentage voided. For each animal one-way ANOVA was used to analyze the effect of dermatome stimulation on bladder voiding when compared to
control stimulation and hand expression. Tests were conducted separately for volumes
and percentages. The Tukey-Kramer method was used to identify post-hoc significance
(α = .05).
RESULTS
Suppression of EUS Reflex activity
All animals demonstrated active urethral reflexes within 1 week after SCI, and reflex
bladder contractions by 3 weeks post SCI. Urethral reflex suppression was achieved in
two of four animals (animals 1, 4) using PSAS patterns previously identified (0.75 ON,
0.25 OFF 20 Hz) and (20 Hz continuous)(71). Suppression was first achieved during
propofol-anesthetized tests 35 days (3rd test, animal 1) and 21 days (2nd test, animal 4) after SCI. Unilateral sacral dermatome stimulation reduced reflex spike rate from (0.129
± .014 to 0.016 ± .012 spikes/second, p<.001) and spike amplitude from (141.8 ±14.8 to
80.2 ± 15.4 cmH2O, p<.001). Surface suppression was achieved on only the left side
dermatomes in animal 1 and on both sides in animal 4. PSAS amplitudes for successful
suppression leading to voiding were 5.6 ±1.7 mA.
52
Two animals (animals 2, 3) did not demonstrate a significant reduction in urethral reflex activity in response to any PSAS parameters tested. These animals therefore did not advance to the voiding stages of the protocol. All animals had complete spinal transections, confirmed post mortem. Urethral reflex activity features were consistent for all types of anesthesia used (propofol IV, alpha-chloralose) across all animals. The ability to demonstrate suppression of reflexes and voiding improvement was independent of anesthetic, including unanesthetized conditions.
Voiding Under Anesthesia
Bladder voiding with PSAS was superior to sacral root stimulation alone under light propofol anesthesia in both animals (Figure 3.2). Data are pooled across animals. Sacral root stimulation alone (n=5) produced 48.3 ±11.75% bladder emptying, with an average voided volume of 22.0 ± 10.76 mL. Reducing urethral pressures with PSAS (n=7) improved voiding to 83.9 ± 6.2% (p<.01), with an average voided volume of 37.8 ± 7.0 mL (p<.05). The addition of deep isoflurane anesthesia completely eliminated urethral reflexes and provided a measure of the maximum voiding efficiency possible. Control sacral root stimulation under isoflurane (n= 3) produced 85.3 ± 3.0%, 38.2 ± 9.2 mL voiding, not significantly improving percentage (p=.629) or volume (p=.949) from afferent voiding.
Isovolumetric pressure recordings confirmed that surface stimulation did not impact bladder pressures during sacral root stimulation driven contractions. Surface stimulation also did not affect the magnitude or duration of distension-evoked reflex bladder pressures. Therefore increased bladder pressure was not the cause of improved voiding.
53
Awake and Daily Maintenance Voiding
Following voiding improvement with PSAS under anesthesia, animals 1 and 4 began awake testing 50 and 22 days post SCI, respectively. Forty-four total sessions of afferent voiding and eighteen control sessions were completed across both animals. Eight afferent sessions and two control sessions were excluded from awake statistical analysis due to electrode connectivity issues in one or more stimulation runs.
Bladder voiding with PSAS (84.6±10.9%) was superior to voiding without PSAS
(54.3±17.9%) under awake-behaving conditions in both animals (Figure 3.3). Manual expression voided 76% (n=112) of bladder volume in animal 1 and 80% (n=48) in animal
4. Twenty-nine of thirty-six voids with PSAS were equivalent to or greater than manual expression. Zero of sixteen control voids were greater than manual expression.
Additional awake stimulation runs produced diminishing returns toward total voided volume. The percentage of total starting volume contributed by individual stimulation runs were [13.4%; 12.3%; 9.1%; 7.9%; 7.3%] for control and [32.2%; 21.7%; 22.8%;
3.6%; 2.1%] when PSAS was added. All awake voids used at least 5 stimulation runs to maximize bladder emptying.
Both animals transitioned to the maintenance voiding phase (animal 1: 65 days post-
SCI; animal 4: 36 days post-SCI) where electric voiding was used twice daily to replace manual expression (Figure 3.4). Six maintenance voids shown in Figure 3.4 had stimulation problems in the first 5 runs and were therefore excluded from statistical analysis (not shown in Figure 3.3). Twenty-five of twenty-eight afferent voids were greater than manual expression percentage while zero of nine control maintenance voids were. Both animals maintained low residual volumes without habituation or appreciable
54
loss of reflex suppression during daily use. Stimulation (6.6±1.9 minutes/episode, range
5.0-10.5) was delivered 1.9 ± 0.7 (range 1-4) times/day. At most, stimulation of 30
minutes/ day was sufficient to replace other bladder maintenance methods.
Subcutaneous Stimulation and Cutaneous Anesthesia
Subcutaneous afferent stimulation (20.9 mA) reduced EUS pressures consistent with
surface stimulation (0.071 ± 0.129 spikes/second; 56.8 ± 51.6 cmH2O). Voiding using
subcutaneous PSAS (91.9±1.7%) was comparable to voiding using surface afferent
stimulation.
Lidocaine gel eliminated surface reflex suppression. Surface stimulation of (.75 ON
.25 OFF, 20 Hz) on the right S1-S2 dermatome suppressed reflex spikes prior to gel
application. Repeating afferent stimulation after cutaneous anesthesia failed to produce
any reduction in either spike rate or spike amplitude.
DISCUSSION
Previous work demonstrated that patterned afferent stimulation of sacral dermatomes
reduces aberrant urethral reflex activity in cats with chronic SCI(71). The present study
demonstrates for the first time that EUS reflex suppression with patterned afferent
stimulation can significantly improve bladder voiding after chronic SCI. Patterned sacral afferent stimulation for EUS reflex inhibition combined with sacral root stimulation for bladder activation was able to replace manual expression as the method of bladder maintenance in 2 awake-behaving animals with chronic SCI for 1-3 weeks.
55
Sacral afferent suppression of urethral reflexes
Urethral reflexes were completely suppressed via sacral surface stimulation in 2/4 animals. Urethral reflex suppression is strongly dependent on both stimulation location
and stimulation parameters(71). The level of suppression, stimulus patterns (0.75 ON,
0.25 OFF 20 Hz) and sacral locations found to be effective for urethral reflex suppression were equivalent to those previously found after chronic SCI(71). Complete urethral
reflex suppression was previously achieved in 2/4 chronic SCI animals(71), resulting in
suppression in 4/8 total chronic SCI animals. Suppression was achieved in 3/6 acute SCI
animals(41).
Previous demonstration of EUS suppression using patterned afferent stimulation was
limited to terminal experiments under alpha-chloralose anesthesia(41,71). Demonstration
of suppression under propofol anesthesia and in awake-behaving animals removes
concerns of anesthetic involvement.
No differences could be detected between the animals in which PSAS was effective
and those in which it was not. The time and testing required to identify effective
stimulation location and parameters for EUS suppression and voiding has decreased in
each animal during the previous study(71) and the current study. Bilateral surface
stimulation did not have an effect in the 2 non-responding animals. Only a limited
sample of the total stimulation parameter space was tested. It is possible that in the non-
responding animals either effective stimulus parameters were not attempted or that the
approach was not effective with the unique neural circuits after chronic SCI in those
animals.
56
Sacral afferent stimulation improves bladder voiding in awake animals
Sacral root stimulation does not produce voiding in humans without irreversible dorsal rhizotomy to eliminate sphincter reflexes(24). Patterned afferent stimulation of the
sacral dermatomes produced “clinically complete” bladder emptying (84.6±5.5%) in
awake-behaving chronic spinal injured cats in both acute sessions and with daily repeated
use over time (Figures 3.3 and 3.4). Maximum voiding in male cats after bilateral
pudendal transection or under deep anesthesia is limited to approximately 80% (82%
±19%)(67). Voiding in male cats is more difficult than females and is more
representative of the human SCI population. Clinically complete bladder emptying in the
chronic SCI cat model provides a greater degree of evidence than the acute SCI model that patterned afferent stimulation may translate to humans with chronic SCI.
Bladder maintenance of animals using only electric stimulation
The use of electric-only afferent voiding to chronically maintain bladder volumes
(Figures 3.3 and 3.4) provides proof-of-concept evidence that this novel approach can be clinically effective in cats with chronic SCI. Voiding was consistently obtained without
loss of effectiveness (Figure 3.4) or habituation.
A non-invasive treatment for urethral spasms or detrusor-sphincter-dyssynergia
would provide a valuable tool for individuals with SCI. Demonstration of EUS
suppression and bladder voiding through a subcutaneous needle electrode suggests that
suppression via an implanted subcutaneous electrode is feasible. Afferent stimulation for
EUS reflex suppression did not suppress distention evoked bladder reflex contractions;
therefore it may be possible to combine this approach with afferent stimulation for
bladder inhibition(72) in a neural prosthesis to restore bladder voiding and continence.
57
Comments on Neurophysiology
Effective dermatome locations are consistent with the lumbosacral levels known to control sphincter function(73,74). However, stimulation was not effective in the L7-S3 and perianal skin regions associated with the perianal-bladder micturition loop present in the early development stages of kittens and other animals(75,76).
The loss of EUS suppression after application of lidocaine gel suggests that cutaneous afferent fibers are involved in affecting the sacral reflex-inducing spinal circuits.
Selective suppression of urethral reflexes without suppression of distension-evoked bladder reflexes implies EUS-pathway specific suppression and not general sacral cord inhibition.
The 50% success rate of urethral suppression in chronic SCI animals is comparable to currently available human sacral neuromodulation devices(26,77). An improved understanding of the mechanisms involved in obtaining urethral reflex suppression should improve the number of potential candidates for patterned sacral afferent stimulation.
CONCLUSIONS
Patterned afferent stimulation of the sacral dermatomes reduces urethral reflexes after chronic SCI and can be used to produce effective bladder emptying in cats after chronic
SCI. Afferent voiding can be used over multiple days in awake-behaving cats to maintain clinically-acceptable bladder volumes. These proof-of-concept animal data suggest that patterned sacral afferent stimulation has potential as a treatment for voiding dysfunction caused by overactive sphincter activity.
58
ACKNOWLEDGEMENTS
Research reported in this publication was supported by the National Institute of Health
award numbers DK077089 and EB004314; Department of Veterans Affairs RR&D668;
and the Cleveland FES Center. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or VA. The authors thank
Tina Emancipator, Jennifer Mikulan, Julie Murphy, Daniel Young and Manfred Franke for technical assistance.
59
Figure 3.1: Sample voiding traces with and without afferent stimulation
Voided volume (black trace) was the primary outcome metric for comparing voiding with and without patterned sacral afferent stimulation (PSAS). Control voiding (Panel A) is inefficient due to urethral spasms impeding fluid flow during the post-stimulus period.
Applying effective afferent stimulation (solid black line, Panel B) suppressed these urethral reflexes and improved the total voided volume. Each voiding run utilized intermittent sacral root stimulation of (2 sec ON 4 sec OFF, 20 Hz), dashed gray bars, to generate peak bladder pressures. Only the first voiding run of a series is shown for each condition.
60
Figure 3.2: Voiding under light anesthesia
Intermittent sacral root stimulation (control, no afferent) produces inefficient bladder voiding in cats with chronic SCI due to aberrant urethral reflexes. Adding patterned sacral afferent stimulation to intermittent sacral root stimulation (patterned afferent) significantly improved voiding (percent at left; volume at right). Both conditions were conducted under light propofol anesthesia. Deep isoflurane anesthesia (control+
isoflurane) eliminates urethral reflexes and provides an anesthetic control that estimates
the maximum voiding efficiency possible; afferent stimulation was equivalent to isoflurane control. Individual data values are shown as empty circles; mean values are shown by filled black circles. Electric voiding was applied as shown in Figure 3.1.
Voided volume was calculated by summating the volume voided in each of 3 successive runs. (* p<.05, ** p<.01)
61
Figure 3.3: Awake voiding compared to manual expression
As in Figure 2, intermittent sacral root stimulation (control) produces only partial bladder
voiding in awake-behaving animals (percent at left; volume at right). The addition of
patterned afferent stimulation (afferent) significantly improved bladder voiding. Manual
expression represents the clinical standard for bladder maintenance in chronic SCI cats.
Intermittent stimulation alone was not sufficient for daily bladder maintenance without additional hand expression, but when combined with afferent stimulation was equivalent to the clinical standard. The subset of maintenance phase data is shown separately (at
right both panels). Afferent stimulation improved percentage voiding over manual
expression (p<.001), but not voided volume (p=.886). Data are pooled from both
animals, and all stimulation trials show the product of 5 voiding runs.
(*** p<.001, * p<.05)
62
Figure 3.4: Daily maintenance voiding over time
After awake testing sessions demonstrated voiding (gray symbols), afferent stimulation
replaced manual expression as the primary method of chronic bladder emptying
(maintenance voiding) in two animals. Additional control voids were conducted
63 intermittently. Voiding was stable over time for both conditions with stable stimulation parameters.
The longest consecutive period of electric maintenance was 5.5 days (animal 1); animals were not available for electric voiding during all weekends, holidays and anesthetized testing days. All data shown represent the output of 5 stimulus runs. Data shown are maintenance data subset from Figure 3.3. Six maintenance voids had stimulation problems and were excluded from statistical analysis, indicated by * symbols.
64
CHAPTER 4: Patterned Afferent Stimulation of Sacral Dermatomes for Suppression of Urethral Sphincter Reflexes in Human Subjects
This chapter is being prepared for submission:
McCoin, J. L., Bhadra, N. and Gustafson, K. J., Does patterned afferent stimulation of
sacral dermatomes suppress sphincter reflexes in individuals with spinal cord
injury?.Neurourol Urodyn June 2013
65
ABSTRACT
Aims: Dyssynergic contractions of the external urethral sphincter prevent efficient bladder voiding and lead to numerous health concerns. Patterned electrical stimulation of
the sacral dermatomes reduces urethral sphincter spasms and allows functional bladder emptying in cats after chronic SCI. Reflex suppression in animals is strongly dependent
on stimulus location and pattern. The purpose of this study was to determine whether the stimulation patterns and locations effective in animals suppress urethral sphincter spasms in humans with SCI.
Methods: Ten subjects with chronic SCI underwent bladder filling to elicit distention- evoked contractions. During reflex contractions patterned electrical stimulation was applied to the S2 or S3 dermatome in random 25 second intervals. Bladder and sphincter pressures were simultaneously recorded and compared between control and afferent stimulation periods.
Results: Six of the ten subjects demonstrated both reflex bladder and sphincter contractions with bladder filling. No significant reduction in urethral pressure was observed during stimulation for any stimulus locations and patterns tested.
Conclusions: Stimulation parameters and locations effective in SCI animals did not suppress reflex sphincter activity in these human subjects. It is likely that a broader set of stimulus patterns and dermatome locations will need to be tested to find the effective combination in humans.
66
INTRODUCTION
Sphincter dyssynergia after chronic SCI prevents efficient bladder emptying and can
produce high bladder pressures that lead to urinary reflux and kidney failure. Effective
bladder management techniques that address sphincter spasms remain a significant
expense and concern for individuals with SCI(12).
Sensory stimulation can modulate spinal reflex pathways and affect lower urinary
tract function in individuals with SCI. For example, surface stimulation of the dorsal genital nerves inhibits reflexive bladder contractions(50,78). However, no neuromodulation therapy currently exists for the targeted inhibition of dyssynergic urethral reflexes.
Patterned stimulation of the sensory afferents in the sacral dermatomes can disrupt urethral sphincter reflexes in cats after acute(41) and chronic SCI(71). Patterned sacral afferent stimulation can produce sufficient bladder emptying to replace veterinary- standard methods of bladder maintenance in chronic animals(79). The successful demonstration of patterned sacral afferent stimulation (PSAS) as a daily bladder treatment in a large animal chronic SCI model suggests that translation to human subjects may be feasible. Testing human subjects is the next step toward developing a beneficial neuromodulation therapy for detrusor-sphincter-dyssynergia. The non-invasive nature of
PSAS limits risk to subjects and facilitates early-stage testing. This general approach is similar to that used for urethral afferent stimulation for bladder excitation, which was demonstrated in acute SCI cats with a less invasive approach(54,80) before being attempted successfully in humans(81,82).
67
Sensory stimulation evokes different physiologic responses in the lower urinary tract
depending on the stimulus pattern (base frequency, cycle time, duty cycle, burst count, burst frequency, etc.)(43,44,78). Urethral reflex suppression with PSAS is dependent on
the combination of a specific sacral stimulus location and the correct stimulus
pattern(71,79). Similar parameter specificity is likely required to suppress urethral
reflexes in human subjects. Sacral dermatome stimulation has previously been attempted
in subjects with multiple sclerosis for the suppression of reflex bladder contractions,
though acute bladder reflex suppression was not observed(83).
The purpose of this study was to determine whether patterned sacral afferent
stimulation can suppress urethral sphincter spasms in humans with SCI. The stimulation
patterns and locations effective for producing suppression and voiding in chronic SCI
animals were chosen as the logical starting point.
MATERIALS AND METHODS
Subject selection
Subjects were recruited from the SCI patient population at the Cleveland Veteran’s
Administration Hospital. All procedures were approved by the Institutional Review
Board. Recruitment was limited to individuals with upper motor neuron SCI above T-12
and urodynamics test or physician assessment indicating bladder and sphincter spasticity.
The study included both genders with complete and incomplete lesions. Individuals with active sacral pressure wounds or who had sacral hyper-sensitivity and could not tolerate stimulation were excluded. Significant urethral stricture or prior sphincterotomy also
precluded participation.
68
Experiment Overview
Subjects underwent a series of modified urodynamics tests (Figure 1) during which bladder and sphincter pressures were simultaneously recorded. The bladder was filled with sterile saline until reflexive bladder contractions were observed. During reflex contractions, patterned electrical stimulation was randomly applied to the S2 or S3 dermatome in 25 second intervals. Differences in bladder and sphincter pressure were compared between control and afferent stimulation periods.
Equipment and Set-up
Vesicular and EUS pressures were continuously recorded through a 9 F triple-lumen fluid catheter. EUS position was determined by conducting a passive urethral pressure profile in accordance with International Continence Society guidelines(84). In men the
EUS transducer was set immediately rostral to the prostatic pressure plateau; in females it was set at the location of highest urethral pressure.
Subjects were monitored throughout the test session for spikes in blood pressure, sweating, headache, or other symptoms that would indicate the onset of autonomic dysreflexia. Blood pressure readings were taken by automatic sphygmomanometer every
5 minutes. Blood pressure increases over 40 mmHg or other symptoms of autonomic dysreflexia were immediately addressed by draining the bladder; testing did not resume until blood pressure returned to normal.
Surface electrodes (2”x4”, Re-Ply reusable snap) were placed on the S2 dermatome
(10 subjects; posterior mid-thigh) and the inferior buttock (S3 dermatome, 4 subjects).
The skin was prepared with alcohol but not shaved. The return electrode was always located rostral to the stimulating electrode.
69
The skin locations found effective for EUS suppression in cats match the S1/S2
overlap region in published dermatome maps(59,60). In cats EUS motoneurons are
concentrated in the S1 and S2 spinal levels. In humans these motoneurons are shifted to
the S2 and S3 levels(85). Therefore we expect that the corresponding dermatomes (S2-
S3) present the best surface target in humans.
Sacral Dermatome Stimulation
Patterned electrical stimulation was delivered to the S2 or S3 sacral dermatome using
monophasic constant-current pulses of 100-200 µsec (DS7A; Digitimer, Hertfordshire,
England). Stimulation amplitude was set to 90% of the current threshold for visible
muscle fasciculation immediately surrounding each electrode.
When reflex bladder contractions were observed, stimulation was randomly applied
in 25 second periods (Figure 2). Primary stimulation patterns were those found to
produce urethral reflex suppression and bladder voiding in chronic SCI animals: (0.75
ON 0.25 OFF, 20 Hz; 20 Hz continuous)(71,79). Additional patterns based on urethral reflex suppression (0.5 ON 0.5 OFF, 20 Hz)(41) and bladder excitation (8x20 Hz @ 1.2
Hz; 10x100 Hz @ 1.2 Hz)(43,44) in acute SCI animals were also tested as a preliminary expansion of the parameter space.
Stimulation patterns and data recording were controlled by a custom-built data-
acquisition program (Labview, National Instruments, Austin, TX).
70
Data Analysis
Bladder and sphincter pressures were filtered with a half-second moving average
filter to reduce infusion pump artifacts. Bladder contractions were registered as a 15 cmH2O or greater rise in vesicular pressure (PVES) over 3 seconds of baseline pressure
immediately prior. A trial was defined as the start of bladder contraction to the point
where vesicular pressure returned to within 10% of the pre-contraction baseline. Data within this trial period was divided into periods of stimulation or no stimulation, and the average vesicular and urethral sphincter pressures calculated for each period. Urethral sphincter pressure was the primary outcome measure. Vesicular pressure was evaluated as a secondary metric to rule out unintended bladder reflex suppression. Stimulation
periods were pooled for each trial; each trial tested only one stimulus pattern and
electrode location. Control periods were likewise pooled. Standard t-tests were
conducted on the pooled data sets for each trial (α=.05).
Following individual trial analysis, stimulation data and control data were
respectively pooled across subjects, across stimulus types, and across electrodes. 1-way
ANOVAs were conducted for each set.
71
RESULTS
Subject Response to Urodynamics
All recruited subjects had previous clinical urodynamics or clinical evaluations
indicating neurogenic bladder activity with impaired emptying. Eight of the ten subjects
demonstrated reflexive bladder activity during testing (Table I). Six subjects had suitable
urethral reflex activity. Two subjects (5 and 2) had reflex bladder activity without any
increase in urethral sphincter pressure, resulting in high levels (>80%) of bladder
leakage; both had reported problems with retention prior to recruitment. Subject 9 had
reflex bladder and sphincter activity but due to equipment failure was only able to
complete part of the testing set.
Sacral Dermatome Stimulation
Zero of the six subjects demonstrated urethral reflex suppression in response to any of
the stimulation patterns (Figure 3). Stimulation did not significantly impact bladder
pressure in any of the six subjects (39.73±21.15 without stimulation; 42 ±22.39 with stimulation).
There was no detectable difference between afferent stimulation of the S2 or S3
dermatomes (p= .493), therefore both stimulus locations were pooled. Inter-subject
variability had a significant effect on EUS pressure for both control and afferent
stimulation periods (p<.001); stimulation patterns were evaluated in each subject prior to
pooling to remove the inter-subject effect. Results were negative in all individual
subjects.
Subjects 2 and 9 required longer pulse width settings, 200 and 500 µs, respectively, to
observe muscle fasciculation within the 100mA amplitude limit of the stimulator.
72
Adverse Events
No subjects reported pain or adverse sensation during stimulation. Subject 4 did experience two episodes of elevated blood pressure during testing. This was alleviated when the bladder was emptied and was not correlated with dermatome stimulation.
Subjects 2, 3 and 7 experienced one or more lower limb spasms during testing. The spasms did not directly correspond to periods of stimulation but a secondary effect could not be ruled out.
DISCUSSION
Recent work in chronic SCI cats demonstrated that patterned afferent stimulation can suppress urethral reflexes(71) and allow repeated clinically effective bladder voiding in awake-behaving animals(79). These results suggest that testing human subjects is a logical step toward developing a beneficial neuromodulation therapy for detrusor- sphincter-dyssynergia. This study established a testing protocol for patterned sacral afferent stimulation in humans based on studies in chronic SCI cats. The stimulus patterns tested in this study did not affect reflexive sphincter or bladder activity. It is possible that the spinal neurophysiology in human subjects is too different from cats and that no set of input parameters exists to selectively suppress urethral spasms. Dermatome stimulation may not be sufficient to affect the spinal circuits involved in EUS control.
The majority of individuals with reflexive sphincter activity tested in this study had incomplete injuries, while the majority of animals tested to demonstrate PSAS suppression had complete injuries. It is unclear whether the remaining neural connections in these patients complicate the ability to achieve urethral reflex suppression.
73
Because patterned afferent stimulation is strongly dependent on stimulation pattern
and location(71,79) and there is a large parameter space for combinations of stimulation
frequency, cycle time, duty cycle and stimulus location, it is more likely that the effective
stimulation patterns and/or dermatomes in humans are different enough from cats that the
combinations attempted were not sufficiently effective to demonstrate urethral reflex
suppression. Animal studies that demonstrated urethral spasm suppression demonstrated
a progressive reduction in the amount of testing required to determine reflex suppression
parameters in each set of animals(71,79). More trials can be conducted per testing
session in animals than humans. Therefore it is expected that a greater number of testing
sessions or larger set of subjects will be necessary to conduct a comparable exploration of
the stimulus parameter space in humans.
A non-destructive, minimally-invasive treatment for urethral sphincter spasms would
have a significant clinical benefit. A larger stimulation parameter space may be required to identify effective PSAS parameters for urethral reflex suppression. An improved mechanistic understanding of patterned sacral afferent stimulation from additional neurophysiology tests in animals could lead to improved human subject testing.
CONCLUSIONS
Stimulation parameters and locations effective in SCI animals did not suppress reflex
sphincter activity in these human subjects. It is likely that a larger range of stimulus
patterns and/or locations are necessary to determine if urethral reflex suppression can be
achieved in humans.
74
ACKNOWLEDGEMENTS
Research reported in this publication was supported by the National Institute of
Health award numbers DK077089 and EB004314; Department of Veterans Affairs
RR&D668; and the Cleveland FES Center. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or VA. The authors thank Melissa Schmitt, Dennis Bourbeau, Dr. Steven Brose, and Matthew Fulgo for technical assistance.
75
Figure 4.1: Human System Diagram
Subjects underwent bladder filling with standard clinical urodynamics. Patterned electrical stimulation was delivered through 2”x4” surface electrodes placed on the posterior thigh (S2) or buttocks (S3) in order to target aberrant urethral reflex circuits in the sacral spinal cord. Stimulation was set at 90% of the visible muscle fasciculation amplitude and supplied by a Digitimer DS-7A stimulus generator. Bladder and sphincter pressures were measured by an intraurethral 9-french tri-lumen fluid catheter.
Abdominal pressure was measured through a rectal balloon catheter (not shown). All fluid pressure lines were digitized by Deltran pressure transducers and processed by a custom data acquisition system.
76
Figure 4.2: Sample pressure trace showing periods of afferent stimulation
Patterned sacral afferent stimulation was delivered in random 25 second intervals from
the start of a detected reflex bladder contraction. Detrusor (PVES) and sphincter (PEUS) pressures were simultaneously recorded; differences in pressure between stimulation and control periods were used to determine the effect of afferent stimulation on sphincter and bladder activity. This example shows (8x20 @ 1.2 Hz) stimulation on the S2 dermatome in subject 6. The subject voided 60 of 240 total ccs by the end of the bladder contraction
(after 120 seconds).
77
Subject ASIA Injury Stimulation Average Reflex ID Gender Score Level Threshold PVES Number 1 M A T-6 23.5 mA 2 F A C-6 70 mA* 57.6±7.6 3 M C T-10 54 mA; 49.7±9.5 65.7 mA 4 M B T-4 43.2 mA; 66.9±7.1 47.7 mA 5 M D C-5 61.5 mA 51.4±7.2 7 M A T-5 41 mA; 58 Areflexive mA 6 M B T-10 71.5 mA 34.2±12.1 8 M B C-5 73.62 mA 33.6±14.5 9 M B C-6 57.2 mA 31.2±2.2 10 M A C-6 84.0 mA** Areflexive
Table 4-I: Summarized clinical information for all subjects
Gender is indicated as male (M) or female (F). ASIA scores were derived from clinical records and were not directly evaluated for this study. Stimulation amplitude was 90% of the S2 stimulus threshold shown; where two numbers are shown the second represents S3 dermatome stimulation. Asterisks indicate individuals who required increased stimulation pulse width in order to observe muscle fasciculation within the available amplitude range (* = 200 µs, ** = 500 µs). Reflex bladder contractions (PVES) were not affected by PSAS and are pooled for all conditions.
78
Figure 4.3: Sphincter Reflex Pressures With and Without Afferent Stimulation
None of the patterned sacral afferent stimulation (PSAS) conditions tested on the S2 or
S3 dermatomes reduced urethral sphincter pressures compared to control periods without
79
PSAS stimulation. The panel at left shows the aggregate sphincter reflex activity with and without stimulation across all stimulation patterns and locations in all six subjects that demonstrated urethral reflex activity. The panel at right shows the sphincter response to individual stimulus patterns, combined across subjects and dermatomes.
Patterns 1 and 2 represent stimulation that was effective at producing voiding in chronic
SCI cats. Patterns 3-5 are an expansion of the parameter space based on other work done in animals. Each boxplot shows the median and interquartile ranges. Outliers are shown as + symbols. The mean of the aggregated data is represented by a black circle. Dark gray boxes represent periods of PSAS, light boxes represent control periods in both panels.
80
CHAPTER 5: Discussion of Results
81
SUMMARY OF KEY CONTRIBUTIONS
For approximately 60 years physicians have been searching for a way to eliminate
aberrant external urethral sphincter (EUS) reflexes to protect kidney function and
improve bladder voiding in individuals with SCI. Therapies have been developed that
circumvent sphincter overactivity (catheters, sphincterotomies, urethral stents, Botox injections), but none address the underlying issue of disordered spinal reflexes.
Afferent neuromodulation of sacral spinal circuits is a growing field that has to date
focused only on bladder dysfunction; this is the first example of LUT neuromodulation
that encompasses a somatic muscle component (EUS). The work of this dissertation
advances afferent stimulation of the sacral dermatomes toward a functioning
neuroprosthetic for bladder voiding in individuals with detrusor-sphincter dyssynergia.
Targeted urethral reflex suppression with sacral afferent stimulation was previously only demonstrated in cats with acute spinal injuries. The first logical milestone was to demonstrate that patterned sacral afferent stimulation (PSAS) can suppress urethral reflexes in animals after chronic spinal cord injury (Aim I). The sacral dermatome space was mapped and key skin locations identified as stimulation targets. The overlap of the
S1/S2 dermatomes was found to be the most effective. This matches the known spinal levels for urethral sphincter motoneuron pools in the cat, keeping with the hypothesis that afferent stimulation is modulating interneuron circuits in the sacral spinal cord. Key stimulus pattern features were identified out of the very large possible stimulus parameter space. Variations of duty cycle, cycle time and base frequency were found to adequately represent stimulus patterns that could selectively suppress urethral reflexes when combined with the correct dermatome location.
82
The most significant contribution of this work is the demonstration that patterned sacral afferent stimulation can serve as a therapy for daily bladder maintenance in chronic
SCI animals (Aim II). The ultimate goal of PSAS is to improve bladder emptying in humans with neurogenic LUT dysfunction, specifically detrusor-sphincter-dyssynergia.
Clinically-complete bladder emptying in animals under awake-behaving conditions effectively reduces PSAS to practice. The extension of awake voiding to repeated, daily bladder maintenance shows that PSAS can provide a clinical therapy in animals very similar to what would be used in human subjects.
The primary criticism of these data is the small number of chronic animals tested (8 animals). This concern was affirmed by reviewers for chapters 2 and 3, however the expense and difficult nature of conducting chronic SCI experiments in cats means that most studies utilizing this model have similarly small numbers of animals. The strength of the data within responder animals supports the “proof of concept” objective. The value of additional animal studies is addressed in the Continuing Research Proposal section of this chapter.
The third independent contribution made by this thesis is the investigation of patterned sacral afferent stimulation in human subjects (Aim III). The stimulus patterns tested in this early study did not affect reflex EUS or bladder activity. This first-pass through the parameter space suggests that the surface locations and stimulus parameters found successful for chronic SCI felines do not directly translate into urethral reflex suppression in people. These data show that while sphincter pressure recording during patterned sacral afferent stimulation is possible, the variability between subjects and logistical limits on urodynamics-based testing will require a significant increase in the
83
number of subject tests in order to adequately cover the large surface and stimulus pattern
parameter space.
The demonstration that patterned sacral afferent stimulation can selectively reduce urethral reflexes after chronic SCI in animals justifies additional studies to discover the
underlying mechanism and further pursue translation into human subjects. The
knowledge of how afferent stimulation modulates spinal reflex activity of the somatic
sphincter muscle may inspire studies utilizing patterned afferent stimulation for reduction
of other afferent -mediated spinal reflexes.
84
DISCUSSION OF SPECIFIC AIMS
Aim 1: Reduction of abnormal urethral reflex activity using patterned sacral afferent stimulation in chronic spinal-injured animals.
The experiments addressing Specific Aim 1 are summarized in Chapter 2. These are the first data successfully demonstrating that urethral reflex suppression can be achieved by patterned sacral afferent stimulation in animals with chronic spinal cord injuries. This provides the first evidence that chronic spinal reorganization does not guarantee changes that prevent activation of the suppression network/ pathway. The consistency of the location and stimulus parameter responses across animals that responded to stimulation suggests that there is a fundamental pathway that exists in (all*) animals prior to spinal cord injury. This spinal organization may be developmentally relevant in infancy but become dormant once supraspinal connections are solidified and conscious control of voiding is obtained in the mature animal. Similar responses to cutaneous stimulation are observed in both neonatal cats and human babies, leading us to surmise that a similar set of connections remains present in the human adult, and may likewise only become relevant after the loss of supraspinal inputs due to spinal injury.
Success in a chronic large animal SCI model is an important milestone in the translation of any therapy to human subjects. Many therapies that have been found promising in acute animal studies or smaller rodent studies, in either chronic or acute models, often fail to make the jump into large/chronic animal models. Large chronic animal studies are viewed as being the most representative of all animal studies before advancing to human subjects.
85
Because suppression was not demonstrated in all animals (5 responders of 8 animals)
and direct neurophysiology studies of spinal reorganization were not conducted, current
data cannot rule out the possibility that neurophysiologic changes prevent afferent reflex
suppression in some individuals. Baseline urethral reflex activity features were not obviously different across responder and non-responder animals; there were differences in reflex spike rate and spike height between animals regardless of their response to afferent stimulation. It is possible that anatomic variations in the nervous pathways between the sacral skin and the spinal cord prevent or contribute to the negative response
in some animals. This issue will likely not be resolved without a greater understanding of
the neurophysiologic mechanism.
Aim 2: Patterned Afferent Stimulation for Bladder Voiding in Chronic SCI animals
The experiments addressing Specific Aim 2 are summarized in Chapter 3. These data demonstrate significant improvement in bladder voiding with patterned sacral afferent stimulation over equivalent bladder drive without afferent stimulation. This is the first time that any form of afferent neuromodulation has been shown to produce clinically-
relevant bladder voiding after chronic SCI.
86
The most significant contribution of this work is the demonstration that patterned sacral afferent stimulation can serve as a therapy for daily bladder maintenance in chronic
SCI animals. The ultimate goal of PSAS is to improve bladder emptying in humans with neurogenic LUT dysfunction, specifically detrusor-sphincter-dyssynergia. Showing reductions in urethral pressure measurements during isovolumetric bladder contractions cannot fully represent the complex interaction between fluid flow through the urethra and other spinal reflexes; there was a possibility that PSAS would be able to eliminate urethral reflexes in the former condition but not the latter, rendering it clinically useless.
This was not the case.
Clinically-complete bladder emptying in animals under awake-behaving conditions effectively reduces PSAS to practice. The extension of awake voiding to repeated, daily bladder maintenance shows that PSAS can provide a clinical therapy in animals very similar to what would be used in human subjects. This is a key step towards demonstrating the long-term effectiveness of patterned sacral afferent stimulation in human subjects.
Aim 3: Application of Patterned Afferent Stimulation to Human Subjects with SCI
Chapter 4 details the initial investigation into PSAS suppression of urethral reflex spasms in humans with SCI. We have developed a system and methodology for testing the key parameters in human SCI subjects, but we have not been able to conclusively demonstrate that patterned surface stimulation of the sacral dermatomes will reduce sphincter spasticity in humans as it does in cats. Likewise we have not concluded that afferent stimulation cannot reduce human urethral sphincter spasms. Rather, we have
87
collected data from enough individuals to be able to develop an improved testing protocol
and to better estimate the number of subject tests that will be necessary in order to find
the effective parameters/ demonstrate suppression.
It is first necessary to demonstrate EUS reflex suppression using PSAS in at least one
individual as a “proof of concept” that the appropriate afferent mediated pathways exist
in humans as well as cats. Once feasibility is established, we will examine the role of the
parameter space on effective suppression.
Previous studies showing similar use of surface stimulation for reflex modulation
have been able to successfully translate pre-clinical animal studies into clinical studies.
The work that we have done here demonstrates that it is not as simple to translate sacral sensory stimulation from cats to humans, and that a greater expansion of the parameter
space will be necessary.
HYPOTHESIS OF MECHANISM
One major limitation of this work is the lack of a clear mechanistic pathway or
neurophysiologic understanding of how the reflex suppression is being achieved within
the spinal cord. As emphasized earlier, these data represent “proof of concept” for a
highly novel phenomenon and may be viewed by some as a “fishing expedition” in a
large sea of variables. This work effectively served to show that the “fish” of urethral
reflex suppression after chronic SCI is a robust-enough species that it warrants further
study, rather than an artifact of acute experimental conditions. Now that reflex
suppression has been shown to be repeatable and robust in chronic SCI animals,
investigation of the underlying mechanisms is justified. We put forward here a set of
88 conjectures of possible neural mechanisms, support for these conjectures and suggestions for how these hypotheses might be tested.
Revisiting Original Assumptions
Chapter 1, Aim 1 provides a list of all the possible variables that were considered when planning the original study design. Of these variables, only stimulus location, stimulus pattern and electrode size were studied in depth. Some data was collected regarding the impact of bladder volume, stimulus amplitude and unilateral vs. bilateral stimulation, but not enough to draw strong conclusions. Stimulus location and stimulus pattern were chosen as the primary target variables because of assumptions that delivering sensory input to the correct spinal level with the correct signal are important.
These variables were confirmed to be crucially important for achieving urethral reflex suppression. We will investigate why these variables play such an important role on influencing the spinal reflex micturition circuits.
Hypothesis 1
Cutaneous afferent stimulation selectively inhibits urethral reflexes (external urethral sphincter and internal urethral sphincter) by overriding visceral afferent inputs that are producing or exacerbating overactive reflexes.
Support
The urethral sphincter itself is somatic muscle under voluntary control (when the spinal column is intact) and there is little evidence to suggest that there are direct synaptic connections between cutaneous afferents and sphincter motoneurons, even within a given spinal level. However, in addition to supraspinal control sphincter motoneurons are also
89
intricately linked to pelvic reflex circuits that are under autonomic control(86). These
reflex circuits respond to visceral afferent inputs, particularly after spinal cord injury.
The sphincter is connected to bladder afferents as well, though through a more complicated set of interneuron connections(87).
Overactive sphincter reflexes are aggravated by untempered input from urethral smooth muscle and urothelium afferents, as can be seen in the chronic SCI cat model where movement of the catheter within the urethra would frequently set off long runs of sphincter spasms in an otherwise quiescent system. Likewise simultaneous pressure recordings at the sphincter and bladder neck in some animals showed that rhythmic pressure “humps” at the bladder neck were concurrently abolished with EUS spasms
when the correct PSAS stimulation was applied. Fluid flow within the urethra would normally produce visceral inputs that increase sphincter reflexes and prevent voiding; cutaneous PSAS overrode these visceral inputs to allow for complete bladder emptying.
Hypothesis 1.2
A set of sacral cutaneous afferents synapse directly on the same secondary neurons as visceral afferents which contribute sensory input from the bladder and or urethra that cause aberrant sphincter spasms. This set of cutaneous afferents is clustered within one subset of the total dermatome fibers, and gives rise to the location specificity observed in chronic SCI animal studies.
Support
It is known that visceral and cutaneous afferents share second order neuron connections because of the phenomena of referred pain. In cases of referred pain,
90
visceral irritation or damage is felt instead as pain on the skin or muscle at another part of
the body. The brain interprets the visceral afferent firing as having come from the
cutaneous afferent fiber instead, unable to distinguish which of the two inputs activated
the secondary neuron (cutaneous stimulation being more “common” than visceral
stimulation, it is the default offender)(88,89). In the case of PSAS, the inverse
relationship may be true— stimulation of cutaneous afferents sharing a secondary
synapse with pelvic visceral afferents may be interpreted by the spinal reflex circuits as
having come from the pelvic viscera instead.
Chapters 2 and 3 of this dissertation detail the narrow window of surface locations where PSAS could be applied and achieve complete reflex suppression. These locations were limited to the S1 and S2 dermatomes, but did not encompass all or even a significant percentage of those dermatomes. Therefore, targeting afferents in the correct spinal level is only part of the formula—not all dermatome afferents have the same level of input to the reflex circuit. It is likely that only a limited set of cutaneous fibers will share direct synaptic connections with key visceral afferents; secondary or tertiary connections may not be strong enough to override the visceral input, hence the very specific locations seen in animals.
It is valuable to note that manifestations of referred pain are not identical in all individuals. Left shoulder pain with myocardial infarction is the most common presentation of referred pain, but some individuals experience this pain in their right arm, neck, or stomach instead. The reason for this variability is unknown, however it fits with our data that there is not necessarily a one-to-one mapping relationship between skin location and key spinal targets in all individuals(88).
91
Testing
The use of anterograde nerve stain agents (axonally transmitted tracers such as herpes simplex virus and horseradish peroxidase) can be used for tracing the key afferent pathways. Using two different tracers, one on the skin and one in the urethra (with or without the bladder) can be used to gain more insight than a single marker traced from the effective PSAS skin location. The convergence of the two staining paths within the spinal cord indicates should clearly indicate which spinal level the target neural circuitry is located in. It can also be used to identify the earliest point of convergence, the first secondary neuron where the cutaneous input is overriding the visceral input.
Hypothesis 2
Stimulation of cutaneous afferents requires a specific firing pattern because the afferent input is interpreted by the spinal circuit as a more normal (i.e. pre-SCI) visceral state, rather than simply overwhelming the circuit with non-specific information.
Hypothesis 2.1
The one second time base of the patterned stimulation represents a key feature of the underlying neural circuitry.
Support
Modern dorsal column pain stimulators based on the Gate Control Theory of Pain
(90) operate under the assumption that simply overwhelming the ascending columns with non-noxious electrical stimulus of large A-beta fibers minimizes the impact/ transmission of small, unmyelinated C-fiber signals. The key to this approach is to simply keep the large fibers from adapting to the stimulus, however that can be achieved.
92
This is inconsistent with the dependence on stimulus patterning observed in both acute and chronic SCI animals. Chronic data suggest that there is something crucial about the 1 second time base, 75% duty cycle input signal that allows for selective
suppression of the urethral reflex response. Although experimental exploration of
stimulus patterns outside the one-second time base was too limited to be conclusive, a
limited number of trials in the terminal experiment of the first PSAS-responder cat
suggest that variance from the one-second cycle time did not produce reflex suppression.
(0.75 ON 0.75 OFF 20 Hz) and (1.5 ON, 0.5 OFF, 20 Hz) maintained the 0.75 second
input time and 75% duty cycle, but neither suppressed urethral spasms at the same
location that was effective with a (0.75 ON .25 OFF, 20 Hz). While 20 Hz continuous
stimulation did produce suppression, 40 Hz continuous stimulation did not, indicating
“more” input is not necessarily equal or better.
Testing
Testing for hypothesis 2 can flow directly from the testing study for hypothesis 1,
where key interneuron populations are identified by neurologic stain in chronic SCI
animals. Microelectrode recording can then be conducted in a terminal study. Recording
from a key secondary neuron with and without PSAS on the effective skin location would
allow for direct comparison of the circuit activity depending on input from visceral
versus cutaneous inputs (visceral afferents could be modulated by moving a catheter within the urethra or changing the resting bladder volume).
93
Hypothesis 3
Urethral reflex spasticity after chronic SCI is modulated in part by pelvic C-fiber
afferents, similar to bladder hyperreflexia. Progressive sensitization of C-fibers over time
may be responsible for the differences in response to stimulus patterns between acute and
chronic SCI feline preparations.
Support
There is evidence to suggest that sacral neuromodulation reduces detrusor
hyperactivity by acting on C-fiber afferents (91). Pain fibers in the bladder and pelvic
viscera are normally quiescent in healthy, intact individuals and play little role in the
micturition reflex. It has been shown that C-fiber afferents become active after spinal
cord injury and play a significant role in producing overactive bladder (and possibly
sphincter) contractions (62,92,93).
PSAS applied to the sacral dermatomes prior to SCI was not observed to affect the
lower urinary tract. It makes sense that the cutaneous afferents therefore would have little impact on the existing micturition reflex circuits because the visceral C-fibers have yet to play a significant part in controlling that reflex. However, this mechanistic postulate may be problematic for the case of reflex suppression in acute SCI cats. It is not clear how quickly C-fibers become sensitized after spinal injury in cats, or if C-fiber changes can occur within just a few hours after SCI. Therefore future investigation of acute SCI cat models should include immunochemistry experiments to tract the level of
c-fos and other C-fiber neurotransmitters over time(64).
94
Hypothesis 3.1
We can identify a specific class of sensory fibers in the skin that are uniquely responsible for transmitting patterned stimulation to target spinal circuits.
Support
It is likely that PSAS is similarly reducing urethral sphincter reflexes though a similar
pelvic C-fiber reflex circuit but we cannot immediately conclude that surface stimulation
is acting through cutaneous C-fibers. Referred pain is suggestive of cutaneous/ visceral
C-fiber co-synapses. However it is not clear whether cutaneous and visceral afferents
sharing a secondary neuron is limited to C-fibers (pain), or if it also occurs for A-delta
and other types of afferent fibers, or if there is a mix of C-fiber and other afferent fiber
types.
Testing
Preliminary investigation into cutaneous pathways with lidocaine gel could be
expanded upon by exchanging electrical activation (which activates all nerve fiber types)
for mechanical inputs. A piezoelectric vibrator capable of delivering vibrations to the
skin in equivalent patterns would determine whether or not mechanoreceptive fibers are
the primary target. Surface or subcutaneous application of chemotoxic capsacin would
selectively knock out cutaneous C-fibers and clarify whether PSAS is actually operating
through pain channels.
95
CONTINUING-RESEARCH PROPOSAL
This section is divided into Remaining Challenges and Future Research Aims.
Remaining Challenges addresses the three primary hurdles remaining to successfully transition from pre-clinical animal testing to clinical testing and neuroprosthetic development in humans. Future Research Aims proposes what next steps should be undertaken to tackle these challenges, and in what order.
Remaining Challenges
Three primary obstacles to clinically delivering a PSAS-based voiding device remain: successfully demonstrating reflex suppression is possible in human subjects; understanding of the neurophysiologic mechanism underlying afferent urethral reflex suppression; and development of a neuroprosthetic device for clinical trials. These obstacles are partially dependent on one another and to some degree will need to be tackled simultaneously.
Demonstrating reflex suppression in human subjects has the greatest potential impact, but will be challenging with first improving our understanding of how the suppression mechanism works in animals. Clinical testing in humans is also limited by our ability to provide a take-home device for early subjects to test PSAS for daily voiding. Long-term, it is likely that an implanted neuroprosthetic for PSAS would be preferable to surface stimulation but appropriate neurologic targets must be identified and tested in animals first. This likewise pairs well with neurophysiologic mechanism studies.
Successful demonstration of reflex suppression in human subjects
It will be essential to show that patterned sacral afferent stimulation can suppress urethral reflexes in human subjects as well as animals. Until suppression is achieved in at
96
least one person, we cannot be sure that the afferent suppression mechanism that exists in cats also exists in humans. Without human validation this technique has minimal clinical
value as a bladder dysfunction therapy. However, if suppression can be demonstrated in
a single individual, we can justify the pursuit of further testing in people and animal
models.
Understanding of the neurophysiologic mechanism underlying afferent urethral reflex
suppression
We do not currently have a clear understanding of how patterned sacral afferent
stimulation produces sphincter reflex suppression, nor do we know precisely which spinal
networks are involved. Pre-clinical realization to date has been significant even without a
well-defined mechanistic picture; however, this success validates the effort necessary to
conduct neurophysiology studies and highlights the value that mechanistic understanding
may produce. It could prove especially valuable in furthering human subject testing. As
identified in the Hypothesis section of this chapter, identifying which cutaneous fibers,
sensory nerves and interneuron pools are activated by patterned stimulation at the target
sacral skin location will allow us to better predict the “transfer function” between humans
and cats, and select human stimulus parameters accordingly.
Discovering the neural mechanism may also lead to additional areas of research, such
as new pharmaceutical targets or other nerve targets for delivering afferent stimulation.
Development and deployment of a neuroprosthetic device for clinical trials
Once a single human individual has been shown to have robust urethral reflex
suppression in acute tests, acute in-lab voiding tests are easy to implement. If it appears
that voiding is improved by surface stimulation, a “take-home” device will allow us to
97 investigate whether patterned sacral afferent stimulation can effectively improve voiding over time (through the use of voiding diaries). It will also allow us to gauge user satisfaction with the system and willingness to use this type of device over longer periods of time, including patients' feelings toward surface electrodes.
Non-invasive surface stimulation provides the easiest interface to identify the correct skin location and target cutaneous afferents in each individual tested. However, surface stimulation may not be the most ideal delivery system for long-term use due to the fragility of sacral skin in SCI patients, who are at increased risk of skin breakdown from moisture and the mechanical stress caused by surface electrodes. Therefore, it is worthwhile to investigate other suitable nervous system targets for a fully implanted afferent stimulation device. This could be done in conjunction with mechanistic exploration, delivering patterned afferent stimulus at points all along the pelvic afferent/ sacral spinal pathways. Ideally an implanted device would have the capability to selectively suppress both bladder and sphincter reflexes, providing both continence and voiding functions to the user without any additional hardware or destructive nerve transections.
Future Research Aims
This section details specific experimental steps that can be taken to address the remaining challenges.
Although effective EUS suppression in human subjects has the greatest potential impact, the first step should be to conduct neurophysiology/ mechanism studies in animals. Mechanism studies are the most “independent” of the three remaining
98
challenges and will contribute most to the other two aims. Neuroprosthetic development
can proceed simultaneously, with additional human subject tests following.
Aim 1: Improve understanding of the neurophysiologic mechanism
Aim 1.1: Identify nerve pathways connecting cutaneous afferents to urethral sphincter
motoneurons
In order to identify the peripheral and central pathways that can serve as
neuroprosthetic targets, we must first improve our understanding of the neurophysiology and mechanisms behind urethral reflex suppression. Additionally, increasing insight into the spinal circuits should allow us to set better stimulus parameters and gain an understanding of inter-subject variability. It will be necessary to first conduct these mechanistic studies in animal models due to their invasive and potentially destructive nature.
Nerve tracing studies will allow us to identify key nerve pathways and cell types from the cutaneous afferents to the spinal circuit interneurons. After target neuron locations
have been identified, microelectrode stimulation and recording can be used to explore the
neurophysiologic mechanism.
Once the key peripheral pathways have been established, the specific spinal levels
fibers enter/ synapse in the spinal cord can be established. This improves the potential for accurate spinal micro-recording placement and will provide greater insight for possible interneuron/ circuit connections to other known micturition pathways.
Nerve tracing studies
Nerve tracing studies can be done in animals with acute and chronic SCI with identical methodology. The direct comparison of acute and chronic results will provide
99 information about the specific reflex circuit changes that take place Nerve tracing studies have been widely used to track anatomical pathways throughout the body. The easiest and most widely used tracer is horseradish peroxidase (HRP). HRP is taken up by axons and transported in both anterograde and retrograde directions. The storage mechanisms are different for anterograde and retrograde, however, meaning that different techniques are required to visualize the anterograde pathway that would be used in this application. Injecting horseradish peroxidase into the skin at the effective surface stimulation location, followed by patterned stimulation, should induce afferent fiber uptake back through the PFCN, through the sacral plexus and into the sacral spinal cord
(2). This would allow us to view the target neural pathways both for a mechanistic understanding and potential identification of new interface targets.
Micro-recording studies
Once we know which axons to target, we can use intracellular microrecordings to ascertain firing rates/ firing properties of these neurons with and without effective surface stimulation. This will allow us to answer valuable questions such as: do afferent fibers train to the surface stimulation? Are there synaptic delays, or is there a level of summation/ additional processing that changes the signal altogether? These are all important pieces for understanding how sensory stimulation may be altering the function of the underlying spinal network.
100
Additional concerns and considerations
Identifying stimulation targets for implanted device
The primary concern for development of an implanted neuroprosthetic is addressed in
future aim 3, but it is worth noting that such work will benefit from co-investigation of
the anatomical reflex pathway. Delivering patterned stimulation to central nerve targets
may provide a more stable long-term interface than surface electrodes.
Variability between individuals
It is possible that the inability to achieve reflex suppression in ~50% of the animals
we tested is related to the inter-individual variability of cutaneous innervation and
distribution through the sacral plexus. It is known that anatomical variability exists between individuals at the level of the sacral plexus(6,7), the pudendal nerve(7), and the femoral cutaneous nerve(8) , but it is unclear how this variability might impact the response to sensory stimulation. Delivering patterned stimulation to the sacral roots or other central nerve targets may reduce the impact of peripheral variability and increase the number of individuals that respond to patterned stimulation.
Aim 3: Demonstration of reflex suppression in humans with patterned sacral afferent stimulation
We expect to gain a great deal of insight from the mechanistic studies in animals, therefore it may be more valuable to wait for further knowledge about potential stimulus patterns than to attempt to “brute force” our way through the parameter space. However, there are still a number of lessons that we can already apply given the experiences with the initial 10 subjects.
101
Key changes to existing protocol
Determining the effective skin location in cats required multiple test sessions, even with the development of a “probing” protocol that was able to move the electrical stimulus across different skin areas and allowed real-time feedback on the suppression effect. It is possible that the first human tests were limited by the used of by fixed electrode patches. Therefore, it will be necessary to bring individual subjects back for multiple test sessions or attempt to extend the amount of time in the single test session
(which is unlikely, given the limits of subject tolerance and number of bladder contractions in the existing 4 hour test block). Barring these we will need to test a much greater number of subjects.
We will also revise the existing testing setup to be able to “probe” different locations while simultaneous recording tonic EUS spasms, as was done in cats. Since the existing urodynamics chair is only designed to allow patients to lay on their back, it will be necessary to have subjects tested on a hospital gurney or bed. Subjects would need to lay on their side, preferably with a foam support between their knees to prevent pressure concerns. This would allow for easier probing of the thigh and buttocks (S2-S4 dermatomes) on the higher side of the body using an approach similar to the one in cats
(moving a non-sticky electrode patch over the skin coated by a thin layer of conductive gel)
If patients prove to be intolerant to the side-laying position for the large amount of time likely needed to conduct a thorough probe, it may be beneficial to develop a wearable “sleeve” that has multiple electrode patches built into it.
102
Additional concerns and considerations
Unlike in our animal subjects, human patients often have had numerous health complications and medical interventions before we see them. Pressure wounds are common in our target population, particularly of the sacral and ischial tissues. In the best cases these wounds are small and have healed through the natural scarring process, however larger wounds may fail to ever completely heal or may have required a surgical skin flap procedure to cover the wound. In such cases the target cutaneous afferents may have been heavily eroded or moved. Individuals with skin flaps may require “transform maps” to estimate the new location of the target cutaneous afferents. Subjects who have had significant scarring in the target location will likely not qualify for patterned afferent stimulation, at least not at the dermatome level.
It has been noted that there are different classifications of DSD among human subjects, each characterized by a different spastic EUS reflex pattern. It is likely that the underlying neurophysiology generating each of these patterns of activity is slightly different and will likely cause subjects to respond differently to patterned afferent stimulation. Classifying subjects by which type of DSD they exhibited and grouping them into appropriate test categories may provide additional insight into “classes” of patterned sacral afferent stimulation most likely to work with each group.
Type 1—dyssynergic up to the point of maximum bladder pressure, then EUS
relaxes and becomes synergistic
Type 2—phasic-ly dyssynergic, producing spurts of urine (similar to the PFPAPG
seen in numerous other animal species (94) )
103
Type 3—Tonically dyssynergic, with continuous active sphincter reflex activity even
with reflex bladder contractions
An evaluation of which type of dyssynergia each subject presents with could quickly be
conducted at the beginning of the test session, and intra-experimental decisions made
about which set of stimulus patters to test with that subject.
Aim 3: Develop a neuroprosthetic device for future clinical trials
This aim will encompass both short-term and long-term neuroprosthetic development.
Once a suitable human subject has been identified for a take-home study, the short-term
goal will be to utilize existing and accessible technology for the rapid deployment of
surface stimulation for bladder voiding with PSAS. The long-term goal will be to take
steps toward a fully implanted neuroprosthetic device.
As noted earlier, however, even if surface stimulation is effective at improving
voiding with long-term use there are concerns about the risks for skin irritation and breakdown caused by regular use of surface electrodes in the sacral region. Some individuals may be opposed to a neuroprosthetic requiring more invasive surgical procedures(95) and therefore opt to continue using the surface therapy or resume more traditional voiding methods if the side effects are too great. Other individuals may be open to receiving an implanted device similar to ones currently available for bladder overactivity(6).
It will be necessary to first develop the implantable neuroprosthetic in animals. We have identified three potential interface locations, listed in order of increasing invasiveness. The first has already been attempted in a limited number of trials in
104
chronic animals; the others will likely depend on the outcome of the mechanism/ pathway
investigation experiments.
Short-term development of a surface stimulation device
Once a human subject has shown evidence of EUS suppression in an acute test
session, we should be prepared to invite that individual to participate in a longer term
“take-home” study to investigate the effect of PSAS on their bladder maintenance
routine. This will allow us to investigate the potential for patterned sacral surface
stimulation to function as a voiding neuroprosthesis in a chronic setting. The bladder
voiding data collected from subject voiding diaries will allow us to analyze the
robustness of stimulation over time. We also expect to receive feedback from the user about their satisfaction with our device, whether they find it sufficiently useful and if they desire to continue using it as a replacement for frequent catheterization.
This device will utilize surface stimulation of the sacral dermatomes identical to those used in the acute test setting. This will be most easily achieved with the Universal
External Control Unit (UECU) developed by the Case Western Technology Development
Laboratory (TDL) and commercially available surface electrodes. The internal components of the UECU are already capable of delivering patterned surface stimulation.
The system will need to be programmed through a custom MATLAB Simulink interface to deliver the specific stimulus patterns found effective in acute testing sessions.
Long-term development of a fully implanted neuroprosthetic device
While a fully implanted device may eventually be warranted in humans, all invasive implant procedures must be fully developed in animals first. The mechanistic work previously proposed will help identify target nerves and locations that may (tolerate)
105 implanted electrodes for a neuroprosthetic device. Each proposed location below would be tested in a chronic SCI animal and afferent stimulation at that location evaluated for use with chronic maintenance voiding)
Subcutaneous wire placed directly under the effective surface location
The work contained in this thesis presents preliminary evidence that subcutaneous stimulus delivery is as effective for reflex suppression as surface stimulation. However, these experiments were limited in number and scope, and did not examine the effect of encapsulation or scarring. This may be a greater concern when the implant is placed in the subcutaneous space immediately below the target fiber endings.
Our early investigation into subcutaneous stimulation (Chapter 2, 3) indicates that a single metal wire placed immediately subcutaneous to the effective skin location identified by surface stimulation can indeed suppress urethral reflexes and produce voiding equivalent to that achieved with surface stimulation. These trials were conducted with 22 gauge stainless steel needles that entered and exited the skin immediately adjacent to the target location, and were only conducted in two animals under anesthetized conditions. Additionally, the current levels needed to achieve suppression with the needle electrode were on the order of 20 mA, a significant current draw for an implanted device.
Utilizing a spiral design would allow us to increase the electrical surface area and decrease the necessary current density, while maximizing current flow in the target region. A self-coiling memory wire would allow for easy implantation without the need for a large incision if a disk electrode was inserted subcutaneously (with risk for potential disruption of the target cutaneous fibers).
106
Implanted electrode on the posterior femoral cutaneous nerve
Reviewing the anatomical literature, the most promising nerve branch to investigate is
the posterior femoral cutaneous nerve (PFCN). The descending cutaneous branch of the
posterior cutaneous nerve innervates the back of the thigh—including the sacral
dermatomes— and joins with the perineal branch which innervates the perineum to
produce the posterior femoral cutaneous nerve.
Work done by Tai and colleagues(45) showed that PFCN stimulation was able to
improve bladder capacity in spinal intact cats anesthetized with alpha-chloralose. The
stimulus patterns used ranged from 3-10 Hz continuous stimulation, which is known to be ineffective at reducing urethral sphincter reflexes at the sacral dermatome level. It may be possible that a single interface on the PFCN could be used to control both continence and voiding.
Though this implant would require an invasive surgery, the single interface would reduce the total amount of implanted hardware necessary to restore natural bladder control. It is unclear whether simply delivering different stimulus patterns to the whole nerve would be sufficient to produce different LUT responses, or if patters would need to be selectively delivered to specific fascicles within the nerve. This could be combined with anatomical/ mechanistic studies; acute stimulation with a FINE cuff electrode would determine if a selective PCFN implanted nerve interface is feasible.
107
BROADER IMPLICATIONS OF THIS WORK
Treating detrusor-sphincter-dyssynergia in other neurologic disorders
The data presented here were limited to animals and individuals with spinal cord
injuries, but detrusor-sphincter-dyssynergia can impact individuals with other neurologic
disorders as well, including multiple sclerosis(96) and brain stem stroke(97). Patterned
sacral afferent stimulation may provide an applicable treatment for DSD in these
populations.
Individuals with brain stem stroke have intact sacral spinal circuits and are comparable to individuals with upper motor neuron spinal cord injuries, so will likely
respond to the same stimulus parameters.
Multiple sclerosis (MS) is a progressive degenerative disease that causes
demyelination of central axons. MS sufferers may also have gray matter lesions that disrupt local circuits and spinal reflex processing. The disruption of central spinal circuits presents an additional challenge for patterned sacral afferent stimulation. The disease progression is highly variable among and even within individuals, which may change the necessary sensory stimulus over time. However, sacral root neuromodulation has been used in MS patients to successfully decrease symptoms of bladder overactivity
(98). This suggests that there may be a window of opportunity in which MS patients can benefit from PSAS before the disruption of the sacral spinal circuits becomes too extreme.
Application of patterned afferent stimulation to other aberrant reflexes
Although this thesis was limited to investigation of the sacral spinal- lower urinary
tract interaction, the techniques developed here have the potential to improve quality of
108
life for individuals with spinal cord injuries beyond bladder maintenance. Patterned afferent stimulation may be applied to different dermatomes or spinal levels to target other aberrant spinal circuits.
Patterned afferent stimulation may be especially valuable for suppressing lower limb spasticity in individuals with SCI or other neurologic lesions. Spasticity negatively
impacts sleep and activities of daily living in the majority of chronic spinal cord
patients(99). Like EUS spasms, lower limb spasticity is aggravated by sensory input and
can be triggered by irritated skin, tight or bunched clothing, bladder or bowel distension,
among other factors. Rapid extension of the limbs may also trigger spasms, interfering
with rehabilitation and transfers(100).
CONCLUSIONS
The suppression of urethral reflexes by patterned sacral afferent stimulation in chronic SCI cats conquers one of the major hurdles before translation into human subjects. Further demonstration that afferent stimulation can reduce sphincter spasms to a level that allows for the restoration of clinically-complete bladder emptying in animals
provides great impetus to pursue the development of a bladder voiding neuroprosthetic
based on this technique. Key steps toward successful translation to human subjects have
been identified.
109
APENDIX I: Additional Animal Study Details Not Previously Included
This appendix contains additional details that are valuable for future researchers attempting to continue or reproduce the chronic SCI animal data, but were not otherwise included in publications (Chapters 2 and 3). The recommendations are broken up into three sections, one for each stage of the chronic animal study and one specifically relating to the spinal transection surgery.
CHRONIC SCI ANIMAL MODEL
Electrode Implant Phase
It is especially important to make sure that the electrodes are placed on the correct
(S2) roots in order to maximize bladder pressure for potential voiding tests.
Ideally cuff electrodes will be properly handed, meaning that cuff wraps are
designed to go on the left or right side. This significantly eases the implant
process and minimizes chances for nerve damage.
Lead wires should be silicon tubed up to approximately half their length, then
untubed to the end to encourage encapsulation by the skin. This makes electrode
removal and reuse significantly more difficult post-mortem, but cuts down on the
danger of leads being pulled out of the skin and damaging the nerve roots.
Suturing lead connectors to the skin and protecting them with jackets further
prevents lead damage.
Animals’ temperament was monitored closely for significant changes after initial
surgery; poor response to the electrode implant surgery indicated that an animal
may not be suitable for daily handling and maintenance after SCI
110
Deliver patterned afferent stimulation to the sacral and lumbar dermatomes
exactly as planned for after SCI and verify that dermatome stimulation does not
produce any changes in bladder or sphincter activity pre-SCI.
o Should test a full range of locations, stimulus patterns, and amplitudes.
In 8 cats tested pre-SCI, 0 animals showed a measurable impact on lower
urinary tract function up to very high amplitudes. High amplitudes (>2x
fasciculation threshold) would produce strong lower limb contractions that
referred movement to the urethral catheter and any pressure changes could not be
distinguished from motion artifacts.
Spinal Transection
A dorsal laminectomy was performed at lower thoracic vertebral levels and the
dura was opened by a longitudinal incision held open by stay sutures. The
exposed spinal cord was transected with micro-scissors using a surgical
microscope (Storz Urban US-1, Urban Engineering Inc., CA) after instillation of
0.1 mL Bupivacaine locally. Care was taken not to injure the blood vessels,
particularly in the ventral part of the cord. Small triangular sponges were used to
staunch bleeding within the surgical field. A square of gelatin foam (Gelfoam,
Pfizer Inc., NY) approximately 1x2x3 mm3 was placed between the cut ends of
the cord to prevent fiber regrowth across the transection and absorb any additional
hemorrhaging.
111
When possible, the dura was closed with 6-O proline suture. If dural tears
prevented adequate suturing, tissue flaps of overlying muscle were used to cover
the cord and minimize CSF leakage. The wound was closed in layers.
Chronic SCI Maintenance
Sterile catheterization during testing sessions is always important, but it is
especially crucial in chronic SCI cats. The need for repeat applications of
antibiotics combined with reduced gastric function can produce chronic diarrhea
in these animals. The need to bathe animals on a daily basis is stressful for both
animals and care givers, and this can contribute to skin problems or increased risk
of secondary infection. The presence of UTI also irritates the urinary system and
may interfere with LUT reflex suppression.
Zero of eight chronic SCI cats reported here demonstrated pathologic or
aggressive behavior towards their denervated limbs. However aggressive licking
or chewing was seen in two other chronic animals in a separate study; care should
be taken to check daily for signs of chewing or other skin damage and the areas
protectively bandaged accordingly.
Reflexive sphincter contractions were typically present within 5 hours of spinal
transection; reflex bladder contractions took from 4 days to 2 weeks to appear
after SCI. Manual expression is significantly easier once bladder reflexes appear,
as the bladder is hard to empty when it is large and floppy. Catheterization was
frequently needed for appropriate bladder emptying in the first week after SCI.
112
APPENDIX II: Animal and Human Testing Setup Documents
113
IMPORTANT POINTS TO AVOID CURRENT GROUND LOOPS
(unwanted stimulation which interferes with the nervous system and may potentially damage nerves):
Refer to cat testing diagram, previous page
• Polarity of connectors to DLS100 (output stages of DS8000) matters. Red to red,
black to black.
• BNC cables from different channels must not touch each other on the metallic
connection points (outside to outside touch) or unwanted / unknown “ground
loops” can occur.
• To avoid DC currents, blocking capacitors are needed to insure that no DC current
drives through the electrodes sitting on nerves effectively damaging the nerve.
• Real world current controlled systems are more likely to have a DC offset. This
DC bias current results from an unbalanced stimulator output. Any real stimulator
will be somewhat unbalanced, which means that the real current signal will also
not be 100% balanced even though the electronic theoretical signal says it is. The
result is a pulsating or continuous bias DC current. This bias current will load
blocking capacitors if in place and cause the stimulator’s output stage to run into
the rails. If DC blocking capacitors are used to overcome this problem, shunting
inductors (or switched shunting resistors) are necessary to level out charges
collecting on either of the capacitors.
• If using an oscilloscope to check signal output, the scope should be running on
battery (aka car battery powered 115V inverter) or it will cause ground loops via
the scope’s own ground line as part of the 115V plug that is coupled to the
114
outsides of the BNC-in of the scope. Furthermore, the scope should not be
connected to anything else than power and the device it’s supposed to measure.
• When the scope any other device is powered by car battery powered 115V
inverter, the voltage of the car battery before the experiment without load should
be at least 12.4Vdc. This ensures power for 10+ hrs of operation.
• When the scope is used, it is advised to measure the current passing through the
electrode instead of the voltage across the electrode. The former adds a small
extra impedance (0.5kR) that is smaller than 10% of the average implanted
electrode (6..8kR). The latter can lead to additional (capacitive) ground loops.
115
116
APPENDIX III: Comparison of Acute and Chronic Feline SCI Data
PREVIOUS ACUTE STUDY
An acute spinal animal preparation that would exhibit urethral reflexes was developed
by Timothy Mariano, MD Ph.D. in conjunction with Kenneth Gustafson, Ph.D. and
Narendra Bhadra, MD Ph.D. This acute SCI cat model was intended to evaluate
neuroprosthetic approaches for suppression of urethral reflexes without the additional
expense and challenges of maintaining chronic SCI animals, and served as the precursor
to the chronic patterned afferent sacral stimulation studies presented in Chapters 2 and 3
of this thesis.
Patterned sacral afferent stimulation (PSAS) was able to achieve urethral reflex
suppression in both acute and chronic SCI felines, but there are considerable differences
between the two experimental setups that makes straightforward comparison of the
outcomes challenging. This appendix details the most important of those details, and
attempts to compare acute and chronic SCI reflex metrics in spite of experimental
differences.
MATERIALS AND METHODS: COMPARISON OF ACUTE AND
CHRONIC PREPARATIONS
Acute SCI preparations are possible in a feline model because of the short period of
spinal shock, especially when compared to other species (97, 98). Acute SCI tests were
experimentally similar to the terminal tests in chronic SCI animals described in detail in
Chapter 2. The use of alpha chloralose anesthesia and insertion of a suprapubic catheter
117
line were present in all acute tests but only during the terminal tests for chronic animals.
The electrode interfaces (sacral root electrodes) were thoroughly encapsulated in the chronic animals and were not accessible for adjustment as in the acute animals. Most significantly, the quality of the spinal transection surgery was of utmost importance in the acute preparation—an overdose of Marcaine or rough handling of the cord could significantly extend the time necessary to observe urethral reflexes or ruin the prep
entirely. The chronic animals all had 9+ weeks of recovery and spinal reorganization,
therefore reflexes were consistently more robust in the chronic SCI animals than the
acute SCI animals.
Urethral reflex pressures were detected and recorded differently in acute and chronic
SCI animals. In the acute animals, urethral pressure (PUR; PEUS) was measured with a 3.5
F perfusion catheter placed at the location of the EUS (generally 3.0 cm to 4.5 cm
proximal to the meatus) as determined by an active urethral pressure profile. In chronic
animals a 3.5F microtransducer catheter was used, as detailed in Chapters 2 and 3. The
temporal resolution and quality of these different measuring systems has a considerable
effect on the final data recordings and therefore the ability to directly compare data
across preparations. Both studies utilized spike rate (spikes/ second) and spike height
(average spike pressure) as their primary variables. However, attempts to record EUS
pressure using a fluid transduction catheter in chronic SCI animals indicated that the
microtransducer produced significantly cleaner recordings of spikes with larger
amplitudes and less baseline shifting.
118
Acute experiments were performed on nine sexually mature, adult male cats. Chronic
SCI experiments were done with 8 adult, sexually intact male cats from the same vendor.
Data Processing
A custom software interface (LabVIEW 7.1, National Instruments Corporation) was
used to record data for both acute and chronic studies. Custom MATLAB programs (The
MathWorks, MA) were used to process the data; processing code was different between acute and chronic studies, but both algorithms extracted average spike amplitude (Pspk,ave)
and spike rate (events per second, Rspk). Statistical analysis for acute data was conducted
in (SPSS 16.0.2 GP, SPSS, IL), and for chronic data in (JMP). Both analyses included
computation of Pearson product-moment correlation coefficients with two-tailed
significances, analyses of variance (ANOVA), and independent-samples t-tests.
COMPARISON OF RESULTS
Urethral Reflexes
Post-spinalization urethral reflex pressure spikes occurred in all cats following sacral
root stimulation (Figure 3A.1). In the chronic animal they also occurred in response to
catheter motion within the urethra, separate from any electrical stimulation. In 8 acute
cats, the amplitude (Pspk,ave) was 44.41 ± 3.03 cm H2O (Mean ± standard error of the
mean), approximately one-third of the average urethral pressure (Pur,ave) during
stimulation. The reflex spike rate (Rspk) was 0.222 ± 0.023 spikes/sec. The reflexes
typically ceased within approximately 30-75 seconds, although reflex durations as short
as 10 s and as long as 2-3 minutes were observed. In the chronic animal, reflex activity
119
following sacral stimulation lasted well over 100 seconds (upwards of 5 minutes of continuous spike activity). For consistency, only spikes occurring within 100 seconds
after the end of sacral stimulation were counted. In twelve control trials, Rspk was 0.398
± 0.057 and Pspk,ave was 69.43 ± 2.76.
Bladder Contractions Impact EUS Reflexes Only in Acute SCI Animals
Generated bladder contractions evoked an average pressure (Pves,ave) of 22.45 ± 1.53
cm H2O with an average maximum pressure (Pves,max) of 57.49 ± 3.92 cm H2O (n = 215
trials). The mean basal bladder pressure (Pves,base) was consistently low at 5.32 ± 0.24 cm
H2O. Generated urethral pressures had a mean Pur,ave of 148.17 ± 10.11 cm H2O and a
mean Pur,max of 214.55 ± 14.63 cm H2O. The average basal urethral tone (Pur,base) was
17.27 ± 1.18 cm H2O.
In acute SCI, animals with higher bladder test volumes had higher urethral pressures
and larger reflex spike amplitudes. Significant positive correlations were found between
the bladder test volume and reflex metrics Pspk,ave and Rspk (all p < 0.002). This was in
strong contrast to chronic SCI animals, where urethral reflexes could be seen at all levels
of bladder filling from 0 to over 50ccs, and where no dependence on bladder volume was
observed. Similarly, spontaneous or distension-evoked bladder contractions did not
occur after acute spinalization whereas they could be evoked with large bladder volumes
in all chronic animals after 3 weeks post-SCI.
120
DISCUSSION
These data define an animal preparation that displays urethral responses after acute spinalization that are functionally similar, although not neurophysiologically equivalent to those seen after chronic SCI. All acute animals exhibited spike-like urethral contractions shortly after spinalization. These responses continued for a variable period following sacral root stimulation and were absent in the spinal intact animal. In our previous work with a different cohort of animals receiving the same preparation (Mariano et al., 2010), seven of nine cats exhibited such responses—totaling 16 of 18 cats across both studies. The dependence of these urethral pressure spikes on bladder volume, their reversible abolition by isoflurane, and their independence from bladder drive strongly
suggest that they are spinal reflexes. This preparation offers a potential means of testing novel neuroprosthetic or pharmacological treatments of aberrant urethral reflexes that cause post-SCI voiding dysfunctions.
Post-Spinalization Reflexes
Anal reflexes returned to 7 of 9 acute animals within 0.5-1.5 hours post-spinalization.
All 9 animals exhibited urethral reflexes when first checked within 7.5 hours after spinalization. This matches results seen in chronic SCI cats.
Urethral reflex activity spikes did not depend on bladder stimulation modality for
either acute or chronic preparations. In the acute prep they occurred after both direct
sacral drive and reflex bladder activation via pudendal nerve stimulation. In chronic
animals spastic EUS reflexes were persistent with and without sacral drive, and could be
elicited by movement of the catheter in the urethra; pudendal reflex activation was not
121 attempted. The pressure responses may contain non-EUS components, as the perineal musculature was observed to contract in synchrony with the spike-like urethral pressure responses, but similar behavior is expected in humans with SCI.
In both acute and chronic SCI preparations it was not possible to observe urethral reflexes during the duration of sacral stimulation (i.e. while bladder pressure was rising) because the extradural sacral root stimulation simultaneously activated both bladder and
EUS, producing a fused EUS contraction that prevented observation of any reflex activity. However, the urethral pressure spikes occurring after the end of stimulation were present while bladder pressure was still elevated (Fig. 1), suggesting a lack of coordination between bladder and EUS.
ACKNOWLEDGEMENTS
The acute work reported here was conducted by Timothy Y. Mariano, MD PhD. The chronic work was conducted by Jaime L. McCoin, Ph.D. The authors thank Adam
Boger, Ph.D., Tim Bruns, Ph.D., Tina Emancipator, Dan Young, Obinna Nwanna, and
Petar Bajic for assistance. This work was supported by NIH DK077089, Department of
Veterans Affairs RR&D B3675R and B6685R, and Department of Education GAANN
P200A040207.
122
Figure AIII.1: Representative trials from two cats illustrating shorter and longer
durations of aberrant urethral reflexes
In both panels, the black trace is Pur, the gray trace is Pves, and the horizontal black bar
indicates application of sacral root stimulation. Note the slightly different horizontal axis
scaling.
123
Figure AIII.2: Direct comparison of acute to chronic SCI key metric data
The panel on the left (A) shows the average urethral spike pressure for two sets of acute
SCI cats as well as all chronic SCI cats. The right panel (B) shows the average spike rate for the same subsets of animals. The high level of inter-animal variability is noteworthy, because it may mask the effects of acute to chronic comparison . However, we can see that the urethral spike pressures are significantly higher in the chronic study cats, while spike rates are equivalent. The differences in equipment used to measure the urethral reflexes may have contributed significantly to the difference in spike amplitude, therefore we cannot conclusively say whether this difference is due to the neurophysiologic reorganization of the spinal circuits in the chronic prep.
124
APPENDIX IV: Alternate Analyses of Urethral Reflex Data
**The work presented in this chapter was contributed by Daniel R. Young, BS with oversight and editing by Jaime McCoin
METHODOLOGY
Frequency Content Exploration – Spectrograms
The frequency content of the EUS pressure signal was investigated using multiple methods. First, the power spectral density (PSD) was calculated using Short-Time
Fourier Transforms on the EUS pressure data using MATLAB (R2011a, MathWorks,
Natick, MA). A window size of .64 seconds with 80% overlap was utilized using the provided SPECTROGRAM function. These parameters were chosen by visual comparison of window sizes ranging from .04-5.12 seconds and overlap ranging from 0-
100%. The output of this analysis was a spectrogram of the pressure data, showing the magnitude of power in each frequency band (0-50Hz, 1.56Hz resolution) for the temporal length of the trial.
Figure A3.3 shows that the instantaneous signal power lumped waveforms are mostly consistent between frequency bins, indicating that no specific frequency range will be a better indicator of EUS reflexive activity than any other. However, the magnitude of the instantaneous power is much larger in the lower frequency bins than in the higher, indicating that the EUS pressure signal is mainly comprised of lower frequency components. Because each of the frequency bins showed similar content, but with different magnitudes, it was determined that the total instantaneous power would be a
125
better measure of EUS activity than arbitrarily picking a narrow frequency band to
examine.
Signal Power Exploration – Total (Mean) Power
Once it was determined that the total power should be used, the Short-Time Fourier
Transform method of obtaining the instantaneous power was replaced by a numerical
integration technique implementing the definition of signal power. This method change
both improved temporal resolution and removed variability due to the input parameters to the Fourier transform, such as window size and overlap. Prior to the calculation of power, the signal was filtered using a 10th order Butterworth filter with a cutoff frequency of
25Hz. The mean power was calculated using the definition of signal power, where x(t) is
the filtered EUS pressure trace and P is the mean power.
1 t2 (Eq 1) Pxtdt |()|2 tt 21t1
MATLAB was used to integrate the square of the signal over 60 seconds and divide
the result by the length of time. A trapezoidal numerical approximation of the integral
was utilized using the provided TRAPZ function. The output of this analysis was the
mean signal power over the 60 second post stimulation period.
Frequency Content Exploration – Lumped Frequency Bins
To better analyze the frequency content of the signal, the instantaneous power in each
frequency band was lumped into bins (0-50Hz, 10Hz resolution). The instantaneous
power for each bin was obtained by summing the magnitude of the scaled complex
Fourier coefficients for each frequency contained in the bin. A moving average was
126
performed on the power signals of each frequency bin. The average was performed in
both the forward and reverse directions with a 6.5 second window. The output from this
analysis was a temporal trace of instantaneous signal power for each lumped 10Hz bin.
These moving averages were inspected to see if any frequency bin trace contained unique
information during dermatome stimulation
RESULTS
New metrics showed suppression in 5 of 8 cats. All results were in agreement with
previous metrics (spike rate, spike amplitude).
DISCUSSION
To date, the non-linear and spectral power analyses have provided an interesting
alternative way of looking at the urethral reflex pressure traces, particularly being able to
consider the total power contained within a trace. While it is good that the results from
these new methods match the outcomes (as defined by a significant reduction in pressure
for any combination of stimulus pattern and location compared to control periods) of the old metrics, they also do not currently supply us with any “new” information. However these preliminary investigations suggest that signal power may provide a better way to categorize “partial suppression” cases where the effect on reflexes is not so obviously clear when looking solely at spikes. The development of a “good/ bad/ worst” categorizing system may provide additional insight into which patterns and locations have a smaller impact on the reflex system than is currently detectable.
127
Figure AIV.1: Spectrogram of a typical trial showing periods of suppression and control
The urethral sphincter activity from a typical trial during the terminal test of cat CA001.
Color map shows red as the highest frequency components and blue as the lowest
frequency components. The spectrogram closely resembles the standard pressure graph.
128
Figure AIV.2: Sample Trace of Frequency Power Bins
A log plot of the moving average of power during successful stimulation for EUS suppression. The large spike in power around 20 seconds occurs during root stimulation, and the subsequent drops in power around 200 and 300 seconds occur during a second
and third application of dermatome stimulation. The different frequencies (sorted into
bins of 10Hz) all share similar traces, justifying the use of total power and not a specific
frequency.
129
Figure AIV.3: Sample Trace of Frequency Power Bins
The power metrics calculated using Short-Time Fourier Transforms were used mostly for qualitative analysis of the data and investigation into the frequency content of the EUS pressure signal. The final power metric data, calculated using the squared signal, were analyzed using one-way ANOVAs with JMP (SAS ©, 2012) statistical software. The mean power was compared for all stimulus locations and parameters against control trials. Only (P<.05) significance was used to label trials as suppression.
130
BIBLIOGRAPHY
1. Jung J, Ahn HK, Huh Y. Clinical and functional anatomy of the urethral sphincter.
International neurourology journal. 2012 Sep;16(3):102–6. Available from:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3469827&tool=pmcen
trez&rendertype=abstract
2. Hole JW. Human Anatomy and Physiology. 2nd Editio. Dubuque, IA: McGraw-
Hill College; 1996.
3. National Spinal Cord Injury Statistical Center T. Spinal cord injury facts and
figures at a glance. The journal of spinal cord medicine. 2012 Jan;35(1):68–9.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/22330194
4. Kirchberger I, Biering-Sørensen F, Charlifue S, Baumberger M, Campbell R,
Kovindha a, et al. Identification of the most common problems in functioning of
individuals with spinal cord injury using the International Classification of
Functioning, Disability and Health. Spinal cord : the official journal of the
International Medical Society of Paraplegia. 2010;48(3):221–9. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/19752871
5. Groutz A, Blaivas JG. Non-neurogenic female voiding dysfunction. Current
Opinion in Urology. 2002;12(4):311–6.
131
6. Gaunt RA, Prochazka A. Control of urinary bladder function with devices:
successes and failures. Progress in brain research. 2006 Jan;152:163–94.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/16198700
7. Anderson KD. Targeting recovery: priorities of the spinal cord-injured population.
Journal of neurotrauma. 2004;21(10):1371–83. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/15672628
8. Snoek GJ, IJzerman MJ, Hermens HJ, Maxwell D, Biering-Sorensen F. Survey of
the needs of patients with spinal cord injury: impact and priority for improvement
in hand function in tetraplegics. Spinal cord. 2004;42(9):526–32. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/15224087
9. French DD, Campbell RR, Sabharwal S, Nelson AL, Palacios P a, Gavin-
Dreschnack D. Health care costs for patients with chronic spinal cord injury in the
Veterans Health Administration. The journal of spinal cord medicine.
2007;30(5):477–81. Available from:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2141733&tool=pmcen
trez&rendertype=abstract
10. Bootsma MJ, Buurman BM, Geerlings SE, De Rooij SE. Urinary incontinence and
indwelling urinary catheters in acutely admitted elderly patients: relationship with
mortality, institutionalization, and functional decline. Journal of the American
Medical Directors Association. Elsevier Ltd; 2013;14(2):147.e7–12. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/23206725
132
11. Pellatt G. Neurogenic continence. Part 1: pathophysiology and quality of life.
British journal of Nursing. 2008;17(13):836–42. Available from:
http://www.internurse.com/cgi-bin/go.pl/library/abstract.html?uid=30534
12. Weld KJ, Graney MJ, Dmochowski RR. Clinical significance of detrusor sphincter
dyssynergia type in patients with post-traumatic spinal cord injury. Urology. 2000
Oct 1;56(4):565–8. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/11018603
13. Blaivas JG, Sinha HP, Zayed AA, Labib KB. Detrusor-external sphincter
dyssynergia: a detailed electromyographic study. The Journal of Urology.
1981;125(4):545–8. Available from:
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt
=Citation&list_uids=7218457
14. Burns S, Rivas D, Ditunno J. The management of neurogenic bladder and sexual
dysfunction after spinal cord injury. Spine. 2001;26(24 Suppl):S129–36. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/11805620
15. Yoshimura N, Chancellor MB. Neurophysiology of Lower Urinary. 2003;5:3–10.
16. Yoshimura N, Smith CP, Chancellor MB, de Groat WC. Pharmacologic and
potential biologic interventions to restore bladder function after spinal cord injury.
Current opinion in neurology. 2000 Dec;13(6):677–81. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/11148669
133
17. Madhuvrata P, Singh M, Hasafa Z, Abdel-Fattah M. Anticholinergic drugs for
adult neurogenic detrusor overactivity: a systematic review and meta-analysis.
European urology. 2012 Nov;62(5):816–30. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/22397851
18. Pan D, Troy A, Rogerson J, Bolton D, Brown D, Lawrentschuk N. Long-term
outcomes of external sphincterotomy in a spinal injured population. The Journal of
urology. American Urological Association; 2009;181(2):705–9. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/19091341
19. Ahmed HU, Shergill IS, Arya M, Shah PJR. Management of detrusor-external
sphincter dyssynergia. Nature clinical practice. Urology. 2006;3(7):368–80.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/16835625
20. Boone TB. Urological Neurology and Urodynamics BLADDER LEAK POINT
PRESSURE : THE MEASURE FOR SPHINCTEROTOMY SUCCESS IN
SPINAL CORD INJURED PATIENTS WITH EXTERNAL DETRUSOR-
SPHINCTER DYSSYNERGIA. 1998;6–10.
21. Jonas U, Tanagho EA. Studies on the feasibility of urinary bladder evacuation by
direct spinal cord stimulation. II. Poststimulus voiding: a way to overcome outflow
resistance. Investigative Urology. 1975;13(2):151–3.
22. Brindley GS. An implant to empty the bladder or close the urethra. Journal of
neurology, neurosurgery, and psychiatry. 1977 Apr;40(4):358–69. Available from:
134
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=492703&tool=pmcentr
ez&rendertype=abstract
23. Martens FMJ, Heesakkers JPF a. Clinical results of a brindley procedure: sacral
anterior root stimulation in combination with a rhizotomy of the dorsal roots.
Advances in urology. 2011;2011:709708. Available from:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3124141&tool=pmcen
trez&rendertype=abstract
24. Kirkham PS, Knight SL, Craggs MD, M CT, Shah PJR. Neuromodulation through
sacral nerve roots 2 to 4 with a Finetech-Brindley sacral posterior and anterior root
stimulator. Spinal cord. 2002;40(6):272–81. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/12037708
25. Krames ES, Peckham PH, Rezai AR, editors. Neuromodulation: Volume 2. 1st ed.
Neuromodulation. Burlington MA: Elselvier Academic Press; 2009. p. 1–1089.
26. Van Kerrebroeck PE , Van Voskuilen AC, Heesakkers JPF, Lycklama á Nijholt,
Siegel S, Jonas U, et al. Results of sacral neuromodulation therapy for urinary
voiding dysfunction: outcomes of a prospective, worldwide clinical study. The
Journal of urology. 2007;178(5):2029–34. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/17869298
27. Lombardi G, Del Popolo G. Clinical outcome of sacral neuromodulation in
incomplete spinal cord injured patients suffering from neurogenic lower urinary
135
tract symptoms. Spinal cord. Nature Publishing Group; 2009;47(6):486–91.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/19238164
28. Hubsher CP, Jansen R, Riggs DR, Jackson BJ, Zaslau S. Sacral nerve stimulation
for neuromodulation of the lower urinary tract. The Canadian Journal of Urology.
2012; 19(5):6480–4. Available from:
http://link.springer.com/chapter/10.1007/978-1-84628-510-3_17
29. Kessler TM, Fowler CJ. Sacral neuromodulation for urinary retention. Nature
clinical practice. Urology. 2008;5(12):657–66. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/19002127
30. Medtronic Inc. InterStim TM Patient Therapy Guide, For Bladder Control
Problems. 2013 p. 1–12.
31. Pham K. Unilateral versus bilateral stage I neuromodulator lead placement for the
treatment of refractory voiding dysfunction. Neurourology and Urodynamics.
2008;27(8):779–81. Available from:
http://onlinelibrary.wiley.com/doi/10.1002/nau.20577/abstract
32. Kondo A, Otani T, Takita T. Suppression of bladder instability by penile squeeze.
British journal of urology. 1982;54(4):360–2.
33. Lee Y, Creasey GH, Y-h AL, Self-controlled CGH. Self-Controlled Dorsal Penile
Nerve Stimulation to Inhibit Bladder Hyperreflexia in Incomplete Spinal Cord
136
Injury : A Case Report. Archives of Physical Medicine and Rehabilitation.
2002;83(Feb):273–7.
34. Lee Y, Creasey GH, Lim H, Song J. Detrusor and Blood Pressure Responses to
Dorsal Penile Nerve Stimulation During Hyperreflexic Contraction of the Bladder
in Patients With Cervical Cord Injury. Blood Pressure. 2003;84(January):136–40.
35. Amarenco G, Ismael SS, Even-Schneider A, Raibaut P, Demaille-Wlodyka S,
Parratte B, et al. Urodynamic effect of acute transcutaneous posterior tibial nerve
stimulation in overactive bladder. The Journal of urology. 2003;169(6):2210–5.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/12771752
36. Al-Shaiji TF, Banakhar M, Hassouna MM. Pelvic electrical neuromodulation for
the treatment of overactive bladder symptoms. Advances in urology.
2011;2011:757454. Available from:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3113365&tool=pmcen
trez&rendertype=abstract
37. Burks FN, Bui DT, Peters KM. Neuromodulation and the neurogenic bladder. The
Urologic clinics of North America. Elsevier Ltd; 2010;37(4):559–65. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/20955907
38. Bhadra N, Grunewald V, Creasey G, Mortimer JT. Selective Suppression of
Sphincter Activation During Sacral Anterior Nerve Root Stimulation.
Neurourology and Urodynamics. 2002;21(1):55–64.
137
39. Bhadra N, Mortimer TJ, Grunewald H. Combined Stimulation of Ventral and
Dorsal Sacral Roots for Control of Bladder Function. USPO; 2006.
40. DeGroat W, Douglas J, Glass J. Changes in somato-vesical reflexes during
postnatal development in the kitten. Brain research. 1975;94:150–4. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/1148861
41. Mariano TY, Bhadra N, Gustafson KJ. Suppression of reflex urethral responses by
sacral dermatome stimulation in an acute spinalized feline model. Neurourology
and urodynamics. 2010;29(3):494–500. Available from:
http://onlinelibrary.wiley.com/doi/10.1002/nau.20717/abstract
42. Bristol AS, Carew TJ. Differential role of inhibition in habituation of two
independent afferent pathways to a common motor output. Learning & memory
(Cold Spring Harbor, N.Y.). 2005;12(1):52–60. Available from:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=548496&tool=pmcentr
ez&rendertype=abstract
43. Bruns TM, Bhadra N, Gustafson KJ. Intraurethral stimulation for reflex bladder
activation depends on stimulation pattern and location. Neurourology and
Urodynamics. 2009;28(6):561–6. Available from:
http://onlinelibrary.wiley.com/doi/10.1002/nau.20703/abstract
44. Bruns TM, Bhadra N, Gustafson KJ. Variable patterned pudendal nerve stimuli
improves reflex bladder activation. IEEE transactions on neural systems and
rehabilitation engineering : a publication of the IEEE Engineering in Medicine and
138
Biology Society. 2008;16(2):140–8. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/18403282
45. Tai C, Shen B, Mally A, Zhang F. Inhibition of micturition reflex by activation of
somatic afferents in posterior femoral cutaneous nerve. The Journal of … .
2012;590(Pt 19):4945–55. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/22869011
46. Park J, Bloom D. The guarding reflex revisited. British journal of urology.
1997;80(6):940–5. Available from:
http://onlinelibrary.wiley.com/doi/10.1046/j.1464-410X.1997.00488.x/abstract
47. Newman DK, Willson MM. Review of intermittent catheterization and current best
practices. Urologic nursing. 2011;31(1):12–28, 48; quiz 29. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/21542441
48. Ahmed HU, Shergill IS, Arya M, Shah PJR. Management of detrusor-external
sphincter dyssynergia. Nature clinical practice. Urology. 2006;3(7):368–80.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/16835625
49. Sanders PMH, Ijzerman MJ, Roach MJ, Gustafson KJ. Patient preferences for next
generation neural prostheses to restore bladder function. Spinal cord. Nature
Publishing Group; 2011;49(1):113–9. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/20531360
139
50. Goldman HB, Amundsen CL, Mangel J, Grill J, Bennett M, Gustafson KJ, et al.
Dorsal Genital Nerve Stimulation for the Treatment of Overactive Bladder
Symptoms. Neurourology and Urodynamics. 2007;(October):1–5.
51. MacDiarmid SA, Peters KM, Shobeiri SA, Wooldridge LS, Rovner ES, Leong FC,
et al. Long-term durability of percutaneous tibial nerve stimulation for the
treatment of overactive bladder. The Journal of urology. 2010;183(1):234–40.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/19913821
52. Bosch JL, Groen J. Neuromodulation: urodynamic effects of sacral (S3) spinal
nerve stimulation in patients with detrusor instability or detrusor hyperflexia.
Behavioural brain research. 1998;92(2):141–50. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/9638956
53. Boggs JW, Wenzel BJ, Gustafson KJ, Grill WM. Frequency-dependent selection
of reflexes by pudendal afferents in the cat. The Journal of physiology.
2006;577(Pt 1):115–26. Available from:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2000666&tool=pmcen
trez&rendertype=abstract
54. Woock JP, Yoo PB, Grill WM. Intraurethral stimulation evokes bladder responses
via 2 distinct reflex pathways. The Journal of urology. 2009;182(1):366–73.
Available from:
http://www.sciencedirect.com/science/article/pii/S002253470900531X
140
55. Wang J, Liu H, Shen B, Roppolo JR, Groat WC De. Bladder Inhibition or
Excitation by Electrical Perianal Stimulation in the Chronic SCI Cat. BJU Int.
2009;103(4):530.
56. Yoo PB, Horvath EE, Amundsen CL, Webster GD, Grill WM. Multiple pudendal
sensory pathways reflexly modulate bladder and urethral activity in patients with
spinal cord injury. The Journal of urology. 2011;185(2):737–43. Available from:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3056321&tool=pmcen
trez&rendertype=abstract
57. de Groat WC, Yoshimura N. Plasticity in reflex pathways to the lower urinary tract
following spinal cord injury. Experimental neurology. 2011; Available from:
http://www.ncbi.nlm.nih.gov/pubmed/21596038
58. Blok BF. Central pathways controlling micturition and urinary continence.
Urology. 2002;59(5 Suppl 1):13–7. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/12007517
59. Kuhn RA. ORGANIZATION OF TACTILE DERMATOMES IN CAT AND
MONKEY. J Neurophysiol. 1953;16(2):169–82.
60. Brown PB, Koerber HR. Cat hindlimb tactile dermatomes determined with single-
unit recordings. Journal of neurophysiology. 1978; 41(2):260–7. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/650266
141
61. Miyazato M, Sasatomi K, Hiragata S, Sugaya K, Chancellor MB, de Groat WC, et
al. Suppression of detrusor-sphincter dysynergia by GABA-receptor activation in
the lumbosacral spinal cord in spinal cord-injured rats. American journal of
physiology. Regulatory, integrative and comparative physiology.
2008;295(1):R336–42. Available from: http://www.pubmedcentral.nih.gov/
articlerender.fcgi?artid=2494799&tool=pmcentrez&rendertype=abstract
62. De Groat WC, Yoshimura N. Changes in Afferent Activity After Spinal Cord
Injury. Neurourology and Urodynamics. 2010;76(December 2009):63–76.
63. Mahfouz W, Corcos J. Management of detrusor external sphincter dyssynergia in
neurogenic bladder. Eur J Phys Rehabil Med. 2011;47(4):639–50. Available from:
http://ukpmc.ac.uk/abstract/MED/22081065
64. Tai C, Roppolo J, De Groat WC. Spinal reflex control of micturition after spinal
cord injury. Restorative neurology and Neuroscience. 2006;24(2):69–78. Available
from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3119351/?tool=pubmed
65. Gaunt R a, Prochazka A. Control of urinary bladder function with devices:
successes and failures. Progress in brain research. 2006;152:163–94. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/16198700
66. Fowler CJ, Griffiths D, De Groat WC. The neural control of micturition. Nature
reviews. Neuroscience. 2008;9(6):453–66. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/18490916
142
67. Boger A, Bhadra N, Gustafson KJ. Bladder Voiding by Combined High Frequency
Electrical Pudendal Nerve Block and Sacral Root Stimulation. Neurourology and
urodynamics. 2008;439(October 2007):435–9.
68. Buss RR, Shefchyk SJ. Sacral dorsal horn neurone activity during micturition in
the cat. The Journal of physiology. 2003;551(Pt 1):387–96. Available from:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2343146&tool=pmcen
trez&rendertype=abstract
69. Shefchyk SJ. Sacral spinal interneurones and the control of urinary bladder and
urethral striated sphincter muscle function. The Journal of physiology.
2001;533(Pt 1):57–63. Available from:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2278618&tool=pmcen
trez&rendertype=abstract
70. McCoin JL, Bhadra N, Gustafson KJ. Electrical stimulation of sacral dermatomes
can suppress aberrant urethral reflexes in felines with chronic spinal cord injury.
Neurourology and Urodynamics. 2013;32:92–7. Available from:
http://onlinelibrary.wiley.com/doi/10.1002/nau.22276/full
71. McCoin JL, Bhadra N, Gustafson KJ. Electrical stimulation of sacral dermatomes
can suppress aberrant urethral reflexes in felines with chronic spinal cord injury.
Neurourology and Urodynamics. 2013;32:92–7. Available from:
http://onlinelibrary.wiley.com/doi/10.1002/nau.22276/full
143
72. Opisso E, Borau a, Rodríguez a, Hansen J, Rijkhoff NJM. Patient controlled versus
automatic stimulation of pudendal nerve afferents to treat neurogenic detrusor
overactivity. The Journal of urology. 2008;180(4):1403–8. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/18710774
73. De Groat WC, Wickens C. Organization of the neural switching circuitry
underlying reflex micturition. Acta physiologica (Oxford, England). 2012 Oct 3;1–
19. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23033877
74. Chang H, Cheng C, Chen JJ, De Groat WC. Serotonergic drugs and spinal cord
transections indicate that different spinal circuits are involved in external urethral
sphincter activity in rats. American journal of physical medicine & rehabilitation.
2007;292(3):F1044–F1053. Available from:
http://ajprenal.physiology.org/content/292/3/F1044.short
75. de Groat WC, Douglas JW, Glass J, Simonds W, Weimer B, Werner P. Changes in
somato-vesical reflexes during postnatal development in the kitten. Brain
Research. 1975;94:150–4.
76. De Groat WC. Plasticity of bladder reflex pathways during postnatal development.
Physiology & behavior. 2002;77(4-5):689–92. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/12527020
77. Al-Shaiji T, Banakhar M, Hassouna M. Pelvic electrical neuromodulation for the
treatment of overactive bladder symptoms. Advances in urology. 2011;2011:1–7.
Available from: http://www.hindawi.com/journals/au/2011/757454/abs/
144
78. Woock JP, Yoo PB, Grill WM. Activation and inhibition of the micturition reflex
by penile afferents in the cat. American journal of physiology. Regulatory,
integrative and comparative physiology [Internet]. 2008;294(6):R1880–9.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/18434446
79. McCoin JL, Bhadra N, Gustafson KJ. Patterned Sacral Afferent Stimulation
Produces Effective Bladder Voiding in Chronic SCI Felines. Neurourology and
Urodynamics. 2013;(In Submission).
80. Gustafson KJ, Creasey GH, Grill WM. A catheter based method to activate
urethral sensory nerve fibers. The Journal of urology. 2003;170(1):126–9.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/12796663
81. Gustafson KJ, Creasey GH, Grill WM. A urethral afferent mediated excitatory
bladder reflex exists in humans. Neuroscience letters. 2004;360(1-2):9–12.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/15082166
82. Yoo P, Horvath E. Intraurethral activation of excitatory bladder reflexes in persons
with spinal cord injury. Conf Proc IEEE Eng Med Biol Soc. 2009. p. 6781–4.
Available from: http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=5333982
83. Fjorback M V, Rey FS Van, Petersen T, Heesakkers JP. Electrical Stimulation of
Sacral Dermatomes in Multiple Sclerosis Patients With Neurogenic Detrusor
Overactivity. Event (London). 2007;530(December 2005):525–30.
145
84. Lose G, Gri D, Hosker G, Kulseng-hanssen S, Perucchini D, Schìfer W, et al.
Standardisation of Urethral Pressure Measurement : Report from the
Standardisation Sub-Committee of the International Continence Society. Society.
2002;260:258–60.
85. Pullen a. H, Tucker D, Martin JE. Morphological and morphometric
characterisation of Onuf’s nucleus in the spinal cord in man. Journal of Anatomy.
1997;191(2):201–13. Available from: http://doi.wiley.com/10.1046/j.1469-
7580.1997.19120201.x
86. Shefchyk S. Spinal mechanisms contributing to striated sphincter control during
continence and micturition: “How good things might go bad”. In: Weaver LC,
Polosa C, editors. Progress in Brain Research Vol 152. 2006. p. 85–95.
87. Shefchyk SJ. Sacral spinal interneurones and the control of urinary bladder and
urethral striated sphincter muscle function. The Journal of physiology.
2001;533(Pt 1):57–63. Available from:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2278618&tool=pmcen
trez&rendertype=abstract
88. Fields HL. Pain. 1st ed. Hamburg: McGraw-Hill; 1987. p. 354.
89. Procacci P, Maresca M. Referred pain from somatic and visceral structures.
Current Review of Pain. 1999;3(2):96–9.
146
90. Melzack R, Wall PD. Pain mechanisms: a new theory. Science (New York, N.Y.).
1965;150(3699):971–9. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/5320816
91. Shaker H, Wang Y, Loung D. Role of C-afferent fibres in the mechanism of action
of sacral nerve root neuromodulation in chronic spinal cord injury. BJU
International. 2000;85(7):905–10. Available from:
http://onlinelibrary.wiley.com/doi/10.1046/j.1464-410x.2000.00559.x/full
92. Yoshimura N, Seki S, Erickson K a, Erickson VL, Hancellor MB, De Groat WC.
Histological and electrical properties of rat dorsal root ganglion neurons
innervating the lower urinary tract. The Journal of neuroscience : the official
journal of the Society for Neuroscience. 2003;23(10):4355–61. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/12764124
93. De Groat WC, Kawatani M, Hisamitsu T, Cheng CL, Ma CP, Thor K, et al.
Mechanisms underlying the recovery of urinary bladder function following spinal
cord injury. Journal of the autonomic nervous system. 1990 Jul;30 Suppl:S71–7.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/2212495
94. Fraser MO. New Insights into the Pathophysiology of Detrusor-Sphincter
Dyssynergia. Current Bladder Dysfunction Reports. 2011;6(2):93–9. Available
from: http://www.springerlink.com/index/10.1007/s11884-011-0083-x
147
95. Anderson KD. Consideration of user priorities when developing neural prosthetics.
Journal of neural engineering. 2009;6(5):055003. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/19721182
96. Nakipoglu GF, Kaya a Z, Orhan G, Tezen O, Tunc H, Ozgirgin N, et al. Urinary
dysfunction in multiple sclerosis. Journal of clinical neuroscience : official journal
of the Neurosurgical Society of Australasia. 2009;16(10):1321–4. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/19560927
97. Sakakibara R, Hattori T, Yasuda K, Yamanishi T. Micturitional disturbance and
the pontine tegmental lesion: urodynamic and MRI analyses of vascular cases.
Journal of the neurological sciences. 1996;141(1-2):105–10. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/8880701
98. Minardi D, Muzzonigro G. Sacral neuromodulation in patients with multiple
sclerosis. World journal of urology. 2012;30(1):123–8. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/21400258
99. Little J, Micklesen P. Lower extremity manifestations of spasticity in chronic
spinal cord injury. American journal of physical medicine & rehabilitation.
1989;68(1):32–6. Available from:
http://journals.lww.com/ajpmr/Abstract/1989/02000/Lower_Extremity_Manifestati
ons_of_Spasticity_in.9.aspx
148
100. Stevenson VL. Rehabilitation in practice: Spasticity management. Clinical
rehabilitation. 2010;24(4):293–304. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/20360150
101. McCouch GP. 1947. Reflex development in the chronically spinal cat and dog.
Journal of neurophysiology 10(6):425-428.
102. McCouch GP, Snape WJ, Stewart WB. 1935. Note on reflex thresholds in the cat
during spinal shock. Am J Physiol 111(2):263-271.
149