- Email: [email protected]
Bladder and urethral sphincter responses evoked by microstimulation of S2 sacral spinal cord in spinal cord intact and chronic spinal cord injured cats
Experimental Neurology 190 (2004) 171 – 183 www.elsevier.com/locate/yexnr
Bladder and urethral sphincter responses evoked by microstimulation of S2 sacral spinal cord in spinal cord intact and chronic spinal cord injured cats Changfeng Tai, August M. Booth, William C. de Groat, James R. Roppolo* Department of Pharmacology, University of Pittsburgh, Pittsburgh, PA 15261, USA Received 11 February 2004; revised 29 April 2004; accepted 1 July 2004
Abstract Urinary bladder and urethral sphincter responses evoked by bladder distention, ventral root stimulation, or microstimulation of S2 segment of the sacral spinal cord were investigated under a-chloralose anesthesia in cats with an intact spinal cord and in chronic spinal cord injured (SCI) cats 6–8 weeks after spinal cord transection at T9–T10 spinal segment. Both SCI and normal cats exhibited large amplitude reflex bladder contractions when the bladder was fully distended. SCI cats also exhibited hyperreflexic bladder contractions during filling and detrusor–sphincter dyssynergia during voiding, neither was observed in normal cats. Electrical stimulation of the ventral roots revealed that the S2 sacral spinal cord was the most effective segment for evoking large amplitude bladder contractions or voiding in both types of cats. Microstimulation with a stimulus intensity of 100 AA and duration of 30–60 s via a single microelectrode in the S2 lateral ventral horn or ventral funiculus evoked large amplitude bladder contractions with small urethral contractions in both normal and SCI cats. However, this stimulation evoked incomplete voiding due to either co-activation of the urethral sphincter or detrusor–sphincter dyssynergia. Stimulation in the S2 dorsal horn evoked large amplitude sphincter responses. The effectiveness of spinal cord microstimulation with a single electrode to induce prominent bladder and urethral sphincter responses in SCI animals demonstrates the potential for using microstimulation techniques to modulate lower urinary tract function in patients with neurogenic voiding dysfunctions. D 2004 Elsevier Inc. All rights reserved. Keywords: Spinal cord injury; Bladder; Urethral sphincter; Cat, Microstimulation; Sacral spinal cord; Micturition; Electrical stimulation
Introduction Spinal cord injury disrupts the normal storage and elimination functions of the lower urinary tract (de Groat, 1975; de Groat et al., 1993, 2001). The bladder does not store well due to detrusor hyperreflexia (DH) (Burns et al., 2001; Jamil, 2001) and does not empty well due to simultaneous contractions of the bladder and the urethral sphincter (detrusor–sphincter dyssynergia, DSD) (Burns et al., 2001; Iwatsubo et al., 1999). DSD prevents complete * Corresponding author. Department of Pharmacology, University of Pittsburgh, E1356 Biomedical Science Tower, Pittsburgh, PA, 15261. Fax: +1 412 648 1945. E-mail address: [email protected] (J.R. Roppolo). 0014-4886/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.expneurol.2004.07.001
elimination of urine, generates high bladder pressures and necessitates the use of intermittent urethral catheterization to empty the bladder. High bladder pressures also contribute to vesicoureteral reflux and renal failure in long-term SCI patients; whereas residual urine in the bladder and urethral catheterization cause cystitis and infection. In addition, DH reduces bladder storage capacity and can induce incontinence, increase risk for kidney damage, and bladder hypertrophy (Burns et al., 2001; Jamil, 2001). The problems of the lower urinary tract contribute to the large costs for medical care of SCI patients and have a tremendous social and psychological impact on the patients and their families (North, 1999; Tyroch et al., 1997). Currently no oral medication can effectively treat DSD, and the pharmacological treatments
172
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
Fig. 1. The experimental setup to measure the bladder and urethral pressures simultaneously. A ligature was placed near the bladder neck to separate the bladder and urethral pressures. Note: the urethra was left intact during voiding experiments (no catheter in urethra and no ligature around the bladder neck).
for DH (e.g., antimuscarinic agents) are limited by their side effects (Yoshimura et al., 2000). In the 1970s, Brindley et al. developed implantable sacral anterior root stimulators to restore micturition function after SCI. This system is now commercially available and has been implanted in over 2000 patients around the world (Brindley, 1994; Jamil, 2001). To achieve optimal results, sacral posterior root rhizotomy is required to prevent DSD or DH (Van Kerrebroeck et al., 1996). The sacral posterior root rhizotomy, which is irreversible, also results in the loss of reflex sexual and defecation functions (Brindley, 1994; Brindley and Rushton, 1990; Creasey, 1993). Voiding induced by the sacral anterior root stimulation generates non-physiologically high pressures in the bladder (Creasey, 1993), which can induce vesicoureteral reflux and renal failure in the long term (Blaivas, 1982; Chancellor et al., 1999). Many studies (Bhadra et al., 2002; Rihkhoff et al., 1997; Shaker et al., 1998; Sievert et al., 2002) have been conducted in an attempt to selectively activate the bladder without activation of the urethral sphincter, but with only limited success. Functional microstimulation of the sacral spinal cord to restore lower urinary tract function after SCI is a very promising approach (Carter et al., 1995; Grill et al., 1999; Roppolo et al., 1993, 1994). Neurons that control the bladder are in a different region of the sacral spinal cord than neurons innervating the urethral sphincter (de Groat et al., 1996; Nadelhaft et al., 1980; Thor et al., 1989; Ueyama et al., 1984). Thus, microelectrodes implanted in the sacral spinal cord could in theory activate bladder without
activating urethral sphincter if the microstimulation was sufficiently localized. Previous studies in our laboratory (Roppolo et al., 1993, 1994) and other laboratories (Carter et al., 1995; Grill et al., 1999) have obtained encouraging results demonstrating micturition in cats following microstimulation of the spinal cord. Bladder contractions could be evoked without significant urethral sphincter contractions (Grill et al., 1999; Roppolo et al., 1993), or even with relaxation of urethral sphincter (Carter et al., 1995). However, most of these studies were conducted in cats with an intact spinal cord (Carter et al., 1995; Grill et al., 1999; Roppolo et al., 1993). As noted above, the neural control of the lower urinary tract undergoes significant changes after SCI (Cheng et al., 1999; de Groat, 1995; de Groat et al., 1993) including the emergence of DSD and DH, which could alter the effectiveness of spinal microstimulation to induce micturition. More experimental data will be needed to test the feasibility of using microstimulation to restore micturition after SCI. In this study, we compared the bladder and urethral sphincter responses induced by microstimulation of sacral spinal cord in spinal intact and chronic SCI cats. Preliminary results of this study were reported previously in an abstract (Roppolo et al., 1994).
Materials and methods All protocols involving the use of animals in this study were approved by the Animal Care and Use Committee of the University of Pittsburgh School of Medicine. Spinalization Five normal (2.5–3.5 kg, three females and two males) and five chronic SCI cats (2.5 –3.5 kg, three females and two males) were used. Cats were spinalized under halothane anesthesia using aseptic surgical techniques. After performing a dorsal laminectomy at T9–T10 vertebral level, a local anesthetic (lidocaine 1%) was applied to the surface of the spinal cord and then injected into the cord through the dura. The spinal cord was then cut completely between the T9 and T10 spinal segments, leaving the parasympathetic and sympathetic innervations to the lower urinary tract intact. A piece of gel foam was placed between the cut ends (usually a separation of 2–3 mm). The muscle and skin were then sutured and after full recovery from anesthesia, the animal was returned to its cage. Following spinal section, bladders were emptied by manual expression at least twice daily. If manual expression was not successful, a sterile catheter was inserted into the bladder through the urethra to empty the bladder. Antibiotics (Polyflex) were given for 10 days following surgery. Six to 8 weeks following spinal cord transection animals were used for spinal cord microstimulation experiments.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
Animal preparation for microstimulation The animals were anesthetized with a-chloralose (60 mg/ kg, i.v.) following induction with halothane. Systemic blood pressure was monitored via a cannula placed in the right carotid artery. A tracheotomy was performed and a tube was inserted to secure the airway. A catheter for i.v. infusion was introduced into right ulnar vein. Ureters were cut and drained externally. Bladder pressure was recorded via a catheter placed through the bladder dome and connected to a pressure transducer while urethral pressure was recorded via a urethral catheter (see Fig. 1). The urethra was slowly infused continuously with saline solution at the rate of 0.1 ml/min. When the urethral sphincter contracts the resistance to flow of the saline infusion is increased, which will cause a pressure elevation within the urethra. The bladder was infused at the rate of 0.5–2 ml/min to determine the volume to induce a reflex bladder contraction, then bladder volume was set to about 80% of the volume necessary to produce rhythmic bladder activity. The proximal urethra was ligated near the bladder neck to maintain isovolumetric conditions in the bladder and to allow separate recordings of bladder and urethral pressures. For the experiments where the voiding induced by microstimulation was investigated (N = 2 for each type of cats), the urethra was left intact and the urine was collected in a beaker attached to a strain gauge to
173
measure the voided volume. In some of the voiding experiments, stainless steel wire electrodes (0.5 mm in diameter) were inserted into the external urethral sphincter to monitor the EMG activity. A dorsal laminectomy from L4 to the first caudal vertebra exposed the lumbosacral spinal cord and roots. The dura mater was opened and each of the sacral segments was identified. The animal was mounted in a modified Narishige spinal cord investigation apparatus. The hip was supported by metal pins and the dorsal spinal process at the rostral end of the laminectomy was secured with a clamp. Skin flaps were retracted to form a pool, which was filled with Krebs solution or mineral oil. The temperature of the animal was monitored and maintained between 358C and 378C using a heating pad and heating tubes in the Krebs or oil pool. Stimulation Before intraspinal microstimulation, each sacral ventral root (S1–S3) was stimulated with a bipolar (cathode distal to anode) platinum-iridium hook electrode to determine the spinal segment that was most effective in inducing a large bladder response or voiding. Stimulation parameters used for ventral root stimulation were 15 Hz frequency, 0.05 ms pulse width, at intensities between 10 and 500 AA for 10 s. Then the spinal segment that was most effective (always S2)
Fig. 2. Bladder and urethral activity after SCI in a female cat. (A) Urethral pressure increased with a slow saline infusion into the bladder. (B) Urethral pressure decreased with large rhythmic bladder contractions. (C) Urethral pressure decreased with large rhythmic bladder contractions induced by perigenital stimulation. Recordings in A, B and C are from the same animal. On/Off marks the start and stop times for bladder infusion (2 ml/min). PG—perigenital stimulation.
174
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
was probed with fine tipped (200 Am2 surface area) activated iridium microelectrodes that were advanced from the dorsal surface of the spinal cord in 200 Am increments. The microelectrode was the cathode, and the anode was placed in the shoulder muscle of the animals. The stimulation parameters used for spinal cord microstimulation were 15 Hz frequency, 0.2 ms pulse width, at an intensity of 50 AA for 10 s. These stimulation parameters were determined based on preliminary experiments. Longer stimulation durations (0.5–4 min) with a higher intensity
(100 AA) were also tested in some locations in an attempt to evoke more efficient voiding. Stimulation for both ventral roots and spinal cord was monophasic and capacitively coupled. The microstimulation of the sacral spinal cord was always started in the middle of S2 segment adjacent to the dorsal root entry zone. Then the microelectrode was withdrawn and moved 200–400 Am rostral/caudal and/or medial/lateral to an adjacent location where the testing was repeated. In each experiment, only one side of the spinal cord was stimulated (usually the left side) and successive
Fig. 3. Bladder hyperactivity after SCI. (A) Hyperactive bladder contractions during a slow infusion of saline (1 ml/min) in a chronic SCI male cat. Very small volume was voided with each hyperactive contraction at a large bladder volume. Total 60 ml was in the bladder when the infusion was stopped. (B) Bladder hyperactivity in another chronic SCI female cat with less frequent, more irregular hyperactive contractions (total 75 ml infused at rate of 2 ml/min). (C) Bladder and urethral activity in a normal male cat. Hyperactive bladder contractions did not occur during slow infusion of saline (0.5 ml/min), and 90% of the infused volume was evacuated during the reflex voiding (13.5 ml out of the total infused volume of 15 ml). The second panels in A and B are bladder pressure recordings continuous with those in the first panels. The third panel in A is corresponding in time to the second panel in A. Insert in C shows at a faster timebase suppression of sphincter EMG activity during voiding and increased activity following voiding.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
penetrations were made as long as the animal’s physical condition permitted (usually 12–24 h). Histology At the end of each experiment, the region of the sacral cord probed with the microelectrode was removed and fixed in formalin (10%). After the cord was fixed, 70-Am-thick sections were cut serially, stained with cresyl violet and examined using light microscopy. Most of the microelectrode tracks could be identified in these histological sections by lesions made by the microelectrode insertion. The microelectrode locations were then correlated with the pressure recordings.
Results Different types of bladder and urethral sphincter activity after SCI Both SCI and normal cats exhibited reflex activity of the bladder and urethral sphincter during bladder filling (Figs. 2
175
and 3). In Fig. 3A, the hyperreflexic bladder contractions in a SCI cat started at a relatively low volume with small amplitude and short duration, and increased in amplitude with the continuation of infusion. However, bladder hyperreflexia was not observed in all chronic SCI cat (see Fig. 2A), and contraction frequency and amplitude varied between animals (see Fig. 3B). In spinal intact cats, the bladder was not hyperreflexic during bladder filling as shown in Fig. 3C. Bladder pressure during the infusion slowly increased without detectable contractile activity until the initiation of reflex voiding. As shown in Fig. 2A, when the bladder of a SCI cat was filled at a rate of 2 ml/min to a maximum of 40 ml, rhythmic activity in the intra-urethral pressure recording was very obvious and especially large during the large bladder contractions at the end of the infusion. This activity reflected hyperreflexia of the external urethral sphincter. Although no voiding could occur due to the ligature at the bladder neck (see Fig. 1), the co-contraction of the urethral sphincter with the bladder during the large bladder contractions presumably represented detrusor–sphincter dyssynergia (DSD). DSD was also evident in experiments where the urethra was open and the animal could void (N =
Fig. 4. Bladder, urethral pressure and voiding responses evoked by stimulation of either S1, S2, or S3 ventral root. (A) Normal female cat: bladder (upper traces) and urethral pressure (lower traces) responses. (B) SCI male cat: bladder (upper traces) and voiding (lower traces) responses. The voiding volume traces are integrated (cumulative) volume levels. The numbers (10–500) at the bottom of figures A and B represent the stimulus intensities in AA, and the stimulation durations are marked by the short black bars (10 s).
176
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
2). As shown in Fig. 3A, only a very small amount of fluid was voided in a SCI cat with every large bladder contraction (over 50 cm H2O) even at a very large bladder volume (about 60 ml). In contrast, reflex voiding in spinal cord intact cats was very efficient (Fig. 3C) because urethral activity which increased when bladder volume was large decreased during reflex voiding, followed by a short burst of urethral EMG activity which occurred after the large reflex bladder contraction and voiding (Fig. 3C). Although DSD was observed frequently in all of the SCI cats, inhibition of the urethral sphincter during a bladder contraction was also detected. In one cat (Fig. 2), the hyperactive contraction of the urethral sphincter continued during a voiding contraction induced by small bladder volume (Fig. 2A), but was inhibited (Fig. 2B) by very large amplitude bladder contractions induced by an additional 10 ml infusion into the bladder. The pressure traces in Fig. 2B were continuous recordings after those shown in Fig. 2A. The bladder contractions induced in Fig. 2A stopped after a period of time presumably as the bladder accommodated to the initial 40 ml infusion. Additional infusion again induced large bladder contractions and inhibition of urethral sphincter activity. As shown in Fig. 2C, inhibition of the urethral sphincter was also observed during the large bladder contractions induced by perigenital stimulation which was applied by repeatedly stroking (approximate twice per second) the
perigenital skin area with a cotton swab. Fig. 2C is a continuation of the recording in Fig. 2B but with only 80% of the bladder volume (10 ml was withdrawn from the bladder). After removal of 10 ml from the bladder, the hyperactivity of the urethral sphincter appeared again although the baseline pressure of the urethra was low (see the urethral pressure at the beginning in Fig. 2C). The first perigenital stimulation was applied with very light stroking and only induced a small bladder contraction. The second and third periods of stimulation with strong stroking induced bladder contractions larger than the rhythmic contractions and inhibited urethral sphincter. Ventral root stimulation Based on ventral root stimulation the S2 sacral spinal cord was the most effective segment for producing bladder contractions or voiding in both normal and chronic SCI cats. Simultaneous measurements of isovolumetric bladder and isotonic urethral pressures (Fig. 4A) or measurements of isotonic bladder pressures and voiding with urethra open (Fig. 4B) revealed that stimulation of S2 ventral root produced large amplitude bladder contractions (56.5 cm H2O mean F 4.5 SEM) with smaller urethral pressures (27.6 cm H2O mean F 7.7 SEM) and voiding (Fig. 4B). Stimulation of S1 ventral root evoked large amplitude urethral pressures (49.7 cm H2O mean F 11.6 SEM) with
Fig. 5. Bladder and urethral activity evoked by microstimulation of S2 spinal cord in a SCI female cat. Five microelectrode penetrations (marked as A, B, C, D, and E on the spinal cord drawing) were made. Electrode E is parallel with electrode D, but at a location 800 Am rostral to electrode D. The corresponding responses evoked by these five penetrations are also marked by A, B, C, D, and E, respectively. The black bars under each response mark the stimulation durations (10 s). The black bars in sequence also represent a 200-Am ventral advancement of the microelectrode in the spinal cord. The first black bar always represents the electrode at the surface of the spinal cord. The most effective locations to evoke bladder contraction without a significant urethral contraction are marked by symbol b*Q along the histologically identified microelectrode tracks on the spinal cord drawing. The corresponding bladder contractions are also marked by the same symbol. Similarly, the most effective locations for urethral sphincter responses are marked by symbol boQ. Microstimulation: 50 AA intensity, 15 Hz frequency and 0.2 ms pulse width.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
relatively small (13.9 cm H2O mean F 4.1 SEM) or no bladder contractions and no voiding, whereas stimulation of S3 ventral root evoked small amplitude contractions of both bladder (39.2 cm H2O mean F 5.7 SEM) and urethra (18.6 cm H2O mean F 12.5 SEM) without voiding. Microstimulation S2 sacral spinal cord was probed with a single microelectrode to determine the most effective location for inducing large amplitude bladder contractions without a significant contraction of the urethral sphincter. Fig. 5 shows a typical experiment in a chronic SCI cat. The microelectrode was positioned at an angle so that the electrode tracks would pass through the sacral parasympathetic nucleus. The effective locations for eliciting bladder contractions were in the lateral ventral horn and deep in the ventral funiculus, where the sacral preganglionic efferent axons to the bladder are located (de Groat et al., 1996; Morgan et al., 1993; Nadelhaft et al., 1980). The maximum evoked bladder pressure was 72.6 cm H2O mean F 5.2 SEM in SCI cats. The bladder pressure induced by 50 AA microstimulation with a single microelectrode implanted in either lateral ventral horn or ventral funiculus of the S2 spinal cord was in the range of 50–100 cm H2O (see Figs. 5 and 6), which was comparable to the bladder response evoked by stimulation of the S2 ventral root (see Fig. 4). This indicates that a considerable proportion of the bladder efferent nerve fibers in the S2 segment were activated by a single microelectrode. However, the effective locations for eliciting urethral sphincter contractions were in the lateral dorsal horn where
177
the visceral and somatic (urethral) afferent axons are located (Morgan et al., 1981; Thor et al., 1989; Ueyama et al., 1984). The maximal evoked urethral pressure was 17.7 cm H2O mean F 4.0 SEM in SCI cats. The small urethral contractions evoked in the locations marked by the symbol b*Q had a relatively slow time course compared to the large urethral contractions evoked in the dorsal horn (marked by symbol boQ). This indicated that the smooth muscle rather than the striated muscle of the urethra might be activated by microstimulation in the lateral ventral horn or ventral funiculus of the S2 spinal cord, or activated by the evoked bladder contractions via a bladder-urethral reflex. Only small bladder contractions (which were not closely linked temporally with the stimulation) could be induced by stimulation in the dorsal horn in chronic SCI animals (Figs. 5A and B). The most effective locations in S2 spinal cord for inducing large bladder contractions by microstimulation in normal animals (Fig. 6) were similar to those sites in chronic SCI animals (Fig. 5). Voiding could be evoked in either lateral ventral horn or ventral funiculus of the S2 spinal cord in normal cats (Fig. 6). The maximal evoked bladder pressure was 46.1 cm H2O mean F 4.6 SEM. It was also possible to induce a large amplitude reflex bladder contraction and voiding (which outlasted the time of stimulation) in normal animals with the microelectrode in the dorsal horn (Fig. 6A). The effective locations for eliciting urethral sphincter contractions were also in the lateral dorsal horn (Fig. 7B second column). The maximal evoked urethral pressure was 25 cm H2O mean F 5 SEM. Fig. 7 shows the mapping of the bladder pressure (Fig. 7 first column), and urethral pressure or voiding response
Fig. 6. Bladder pressure and voiding evoked by microstimulation of S2 spinal cord in a normal male cat. Five microelectrode penetrations (marked as A, B, C, D, and E on the spinal cord drawing) were made. The corresponding responses evoked by these five penetrations are also marked by A, B, C, D, and E, respectively. The black bars under each response mark the stimulation durations (10 s). The black bars in sequence also represent a 200 Am ventral advancement of the microelectrode in the spinal cord. The first black bar always represents the electrode at the surface of the spinal cord. The most effective locations to evoke a bladder contraction are marked by symbol b*Q along the histologically identified microelectrode tracks on the spinal cord drawing. The corresponding bladder contractions are also marked by the same symbol. The voiding volume traces are integrated (cumulative) volume levels. Microstimulation: 50 AA intensity, 15 Hz frequency and 0.2 ms pulse width.
178
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
groups of animals (Fig. 7 second column). There are more locations in S2 spinal cord of the normal cats for inducing small bladder contractions (pressure less than 30 cm H2O) than in the SCI cats (Fig. 7 first column), but there are more locations in the SCI cats for inducing small urethral contractions (pressure less than 10 cm H2O) than in the normal cats (Fig. 7 second column). However, the interpretation of the mapping results should always keep in mind that the electrode penetrations were targeting the sacral parasympathetic nucleus, which resulted in the central area of the S2 spinal cord being unprobed (as shown in Fig. 7 third column where all tested microstimulation sites are presented). Influence of microstimulation parameters
Fig. 7. Mapping of the bladder, urethral pressure and voiding response in S2 spinal cord. (A) From five SCI animals. (B) From five normal animals. Microstimulation: 50–100 AA intensity, 15 Hz frequency, 0.2 ms pulse width and 10–30 s duration.
(Fig. 7 second column) in S2 spinal cord for SCI (Fig. 7 first row) and normal (Fig. 7 second row) animals. Large bladder contractions (pressure greater than 30 cm H2O) were induced by microstimulation in either lateral ventral horn or ventral funiculus of the S2 spinal cord (Fig. 7 first column) with small urethral contractions (pressure less than 10 cm H2O) or no/incomplete (b90%) voidings (Fig. 7 second column) in both SCI and normal animals. Large urethral contractions (pressure greater than 10 cm H2O) were mostly induced in the lateral dorsal horn in both
The influence of the microstimulation parameters on the evoked responses was tested with an electrode implanted in the S2 lateral ventral horn or ventral funiculus at locations similar to those marked by symbol b*Q in Figs. 5 and 6. Fig. 8 shows the influence of the stimulation parameters on bladder and urethral sphincter responses in a chronic SCI cat. Repeated stimulation evoked large bladder contractions with little attenuation of amplitude (Fig. 8A). Long duration stimulation (4 min) evoked an initial sustained large amplitude bladder contraction lasting over 1 min, which then converted into a series of rhythmic contractions (Fig. 8B). A higher stimulation frequency (15 Hz) evoked larger sustained bladder contractions than a lower frequency (3.5 Hz) (Figs. 8B and C). There was no obvious influence of changing the stimulation parameters on intra-urethral pressure responses (Figs. 8A, B and C) that were very small. The influence of the stimulation parameters and stimulation location on bladder contractions is shown in Fig. 9. At a stimulus intensity of 50 AA and duration of 10 s, only small bladder contractions were evoked at sites in the ventral horn (2.2–2.8 mm depth). However, at a higher intensity (100 AA) and longer stimulation duration (30 s),
Fig. 8. Influence of the stimulation parameters on bladder and urethral sphincter responses in a SCI female cat. (A) Influence of repeated stimulation (frequency 15 Hz, duration 10 s). (B) Low frequency (3.5 Hz) for a longer duration (4 min). (C) High frequency (15 Hz) for a longer duration (1 min). Stimulation durations are marked by the black bars at the bottom of each trace. Other stimulation parameters: intensity 50 AA, pulse width 0.2 ms.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
179
Fig. 9. Influence of stimulation parameters and stimulation location on bladder contractions and voiding in a SCI male cat. The black bars under the bladder pressure trace mark the stimulation durations. The numbers on top of the black bars indicate the stimulus intensities. The numbers below the bars indicate the electrode depths (mm) from the surface of the spinal cord. The microelectrode was advanced ventrally into the spinal cord in 0.2 mm increments to a depth of 2.8 mm, then the stimulus intensity was changed from 50 to 100 AA and the electrode was withdrawn back to 2.4 and 2.2 mm depths, respectively. The voiding volume trace is an integrated (cumulative) volume level. Other stimulation parameters: 15 Hz frequency, 0.2 ms pulse width, 10- or 30-s duration.
Fig. 10. Influence of stimulation parameters on bladder, urethra and voiding responses in normal cats. (A) Different stimulation intensities (female). (B) Different stimulation durations (male). The voiding volume trace is an integrated (cumulative) volume level. The black bars at the bottom of each trace mark the stimulation durations. Other stimulation parameters: frequency 15 Hz, pulse width 0.2 ms. Electrodes in S2 lateral ventral horn.
180
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
larger bladder contractions were evoked at the same locations (2.2–2.8 mm depth). The results shown in Fig. 9 indicated that lower stimulation intensity (50 AA) might be better to determine more precise locations for effective microstimulation, but a higher intensity (100 AA) with longer duration (30 s or 1 min as shown in Fig. 8C) can be used to compensate for a non-optimal placement of the microelectrode. The small bladder contractions evoked by stimulation in the dorsal horn (0.2–1.6 mm depth in Fig. 9) were not all synchronized with the stimulation, indicating that they were not due to direct activation of efferent pathways but were more likely due to indirect activation of preganglionic neurons via stimulation of other elements of a reflex circuit (e.g., afferent axons or interneurons). Similar to chronic SCI cats, higher stimulus intensities and longer stimulation durations also induced larger amplitude, longer lasting bladder contractions in normal cats (Figs. 10A and B). Voiding evoked by microstimulation Voiding was evoked in both spinal intact and chronic SCI cats by a single microelectrode implanted in the S2 lateral ventral horn or ventral funiculus at locations similar to those shown in Figs. 5 and 6. Stimulation with a higher intensity and longer duration was also better in inducing voiding (see Fig. 9). Fig. 11 shows the voided volumes and the corresponding bladder pressures generated by stimulation at an intensity of 100 AA, and duration of 1 min in a chronic SCI cat. The stimulation was repeated three times at approximately 4-min intervals to allow the bladder to accommodate to the smaller volume after each void. Although sustained large amplitude bladder contractions (over 60 cm H2O) were evoked, only small amounts of fluid (about 5 ml) were voided during each stimulation. With a total of 3 min of stimulation, only 15 ml of the 48-ml bladder volume was voided (31%) in this experiment.
Similarly, complete voiding was never observed in other SCI animals with the stimulation strategy shown in Fig. 11. This indicated that the urethral sphincter must have been contracting reflexively in concert with every bladder contraction (Fig. 2A) or activated by every stimulation (Figs. 5C, D and E). Voiding induced by microstimulation in spinal intact cats was also incomplete (b90%) even with a 30 second stimulation at intensity of 100 AA (see Fig. 10B) due to the co-activation of the bladder and urethral sphincter (see Fig. 10A), although a complete (N90%) voiding did occur during reflex micturition induced by bladder distention (see Fig. 3C). The average voiding efficiency during a single void evoked by microstimulation was 13.0 F 0.5% (mean F SEM) in SCI cats and 14.1 F 5% in normal cats.
Discussion This study revealed that microstimulation of the sacral spinal cord with a single electrode in S2 lateral ventral horn or ventral funiculus could induce large amplitude bladder contractions with minimal activation of the urethral sphincter in a-chloralose anesthetized cats with the spinal cord intact or with the spinal cord chronically transected. However, microstimulation elicited relatively inefficient voiding in both groups of animals. SCI animals also exhibited DH, DSD and inefficient voiding in response to bladder distention. Normal animals did not exhibit DH and DSD, and also had efficient reflex voiding in response to bladder distention. This study shows that spinal neural circuitry after SCI is sufficient to allow a single microelectrode to induce large bladder contractions for voiding, but additional procedures are needed to promote urethral relaxation during voiding. The bladder and urethral sphincter responses evoked by microstimulation of S2 sacral spinal cord are similar to those evoked by S2 ventral root stimulation. Both stimulations
Fig. 11. Voiding evoked by microstimulation of S2 sacral spinal cord in a SCI male cat. The black bars mark the stimulation duration (1 min). Total 48 ml saline in the bladder (about 80% of the volume to induce rhythmic bladder contractions). Electrode at 2.2 mm depth in the spinal cord. The voiding volume trace is an integrated (cumulative) volume level. Stimulation: 100 AA intensity, 15 Hz frequency, 0.2 ms pulse width, and 4-min interval between each stimulation.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
evoked large bladder contractions (50–100 cm H2O) with small urethra contractions (Figs. 4–6), and also evoked incomplete voiding (Figs. 4B, 10B and 11). With the stimulus intensity less than 100 AA delivered through a single microelectrode, the probable area of direct activation in the spinal cord is only about 1.0 mm in diameter (Bagshaw and Evans, 1976; Yeomans et al., 1986). Thus, the number of directly activated neurons/axons may still be very small. However, the stimulation via a single microelectrode induced responses similar to those evoked by stimulation of the entire S2 ventral root. This suggests that microstimulation via a single electrode may indirectly activate a complex spinal neural circuit which may spread rostrally and caudally within the same spinal segment and probably into adjacent spinal cord segments through axon collaterals, interneurons and inter-segmental circuits along the spinal cord. Spread to the opposite side of the spinal cord is also possible via axon collaterals (Morgan et al., 1991). In both chronic SCI (Figs. 5 and 7A) and normal (Figs. 6 and 7B) cats, the most effective locations to induce large bladder contractions without a significant contraction of the urethral sphincter were in the S2 lateral ventral horn or ventral funiculus where the efferent axons from the sacral parasympathetic nucleus passed through to exit from the cord (de Groat et al., 1996; Morgan et al., 1993; Nadelhaft et al., 1980). Therefore, it is likely that the axon, not the neuron, was activated by microstimulation (McIntyre and Grill, 1999; Rattay, 1998), although the microelectrodes passed through the sacral parasympathetic nucleus in this study. This result also agrees well with other studies on cats with an intact spinal cord (Carter et al., 1995; Grill et al., 1999). Large contractions of the urethral sphincter were evoked in the S2 lateral dorsal horn in both SCI cats (Figs. 5 and 7A) and normal cats (Fig. 7B), where the afferent axons from the pelvic and pudendal nerves project (Morgan et al., 1981; Thor et al., 1989; Ueyama et al., 1984). The activation of visceral and somatic (urethral) primary afferent axons could induce large amplitude reflex bladder contractions and voiding in normal cats (Fig. 6A), but only induced small bladder contractions in SCI cats that were not well synchronized with the stimulation (Figs. 5A and B, and 9). This indicates that microstimulation in S2 dorsal horn area might involve different reflex mechanisms in normal and chronic SCI cats. Only a spinal reflex mediated by C fiber afferents can be evoked in chronic SCI cats, whereas a spino-bulbospinal reflex mediated by lower threshold Ay fiber afferents might be elicited in normal cats (de Groat, 1995; de Groat et al., 1993). This study revealed different types of bladder and urethral sphincter activity existing among different chronic SCI cats and even at different times in the same animal. Bladder hyperactivity was seen in the animals shown in Figs. 3A and B, but not seen in another animal shown in Fig. 2A. For the same animal shown in Fig. 2, the activity of
181
urethral sphincter could be enhanced by increasing bladder volume or by bladder contractions (Fig. 2A), but it could also be inhibited by very large bladder contractions when bladder volume was increased (Figs. 2B and C). The different types of bladder and urethral sphincter activity in the chronic SCI cats indicates that the neural re-organization or plasticity after SCI (Cheng et al., 1999; de Groat, 1995; de Groat et al., 1993) might occur differently for different animals, since their neuro-urological systems might experience different external stimuli (i.e., How much urine was accumulated in the bladder between the twice daily manual expressions? How much urine was expressed each time? How many times was urethral catheterization required to empty the bladder since this could damage the urethra and increase the urethral resistance?). Thus, there is a need for a more detailed understanding of the mechanisms underlying the synaptic re-organization and recovery of lower urinary function after SCI. Inhibition of the urethral sphincter activity with very large reflex bladder contractions (Figs. 2B and C) indicated that the neural circuitry to inhibit urethral sphincter activity might still be present in some of the spinal animals as noted by Rampal and Mignard (1975). Since the spinal cord was transected at T9–T10 spinal segment, this inhibitory neural circuit must exist in the lumbosacral spinal cord. One inhibitory circuitry could consist of bladder afferent projections to g-aminobutyric acid (GABA) immunoreactive inhibitory neurons in the dorsal commissure area (Blok and Holstege, 1997; Sie et al., 2001) which in turn make synaptic connections with pudendal motor neurons in the Onuf’s nucleus (Blok and Holstege, 1997; Blok et al., 1998; Nadelhaft et al., 1980; Sie et al., 2001; Thor et al., 1989). Although this putative inhibitory pathway was not able to effectively coordinate bladder and sphincter activity after SCI in anesthetized or awake cats (Rampal and Mignard, 1975), it is possible that microstimulation in the dorsal commissure area might be more effective in inhibiting urethral sphincter activity (Blok et al., 1998; Carter et al., 1995). However, in this study, we only targeted lateral sites containing sacral parasympathetic neurons or their axons to activate bladder. Due to the significant re-organization of the neural circuitry after SCI (Cheng et al., 1999; de Groat, 1995; de Groat et al., 1993), the feasibility of facilitating voiding in chronic SCI cats by inhibiting the urethral sphincter with microstimulation of the GABA-ergic inhibitory neurons needs to be tested in the future. Even though small urethral pressures were recorded during large bladder contractions evoked by microstimulation (Figs. 5C, D and E, Fig. 8 and Fig. 10A), an efficient voiding was never observed (Figs. 9, 10B and 11). This indicated that the pressure recording from the urethra using a catheter with a very slow infusion (0.1 ml/min) might not be a good measurement to represent the entire activity along the urethra. A recent study (Wang et al., 1999) in cats showed that the urethral pressure could change as much as 5-fold at different locations along the length of urethra.
182
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
a-Chloralose anesthesia can significantly reduce the autonomic reflex of the lower urinary tract (Rudy et al., 1991), which might have influenced the evoked bladder or urethral reflex responses in this study. However, this study also showed that electrical stimulation (both ventral root and spinal cord) could evoke large bladder or urethral responses, which indicates that a-chloralose anesthesia may have only modest effects on the electrically evoked efferent responses (i.e., stimulation of efferent axons) where reflexes are not dominant. In addition to bladder and urethral contractions, a variety of somatic responses were also seen with ventral root stimulation. These include tail and hindlimb movements and contraction of perineal muscles. These somatic responses were especially large with S1 ventral root stimulation and smaller with S2 or S3 ventral root stimulation. The hindlimb movement with microstimulation of the S2 spinal cord was often absent or very small at the stimulation intensities used in this study. Some tail and perineal activity was seen with microstimulation deep in the S2 ventral horn. A stimulation strategy to control micturition by microstimulation of sacral spinal cord needs to be developed in a chronic SCI animal model before any long-term study with chronic implantation in humans should be carried out. This study suggests that one electrode implanted in the S2 lateral ventral horn or ventral funiculus might be sufficient to generate the necessary bladder pressure for voiding (Figs. 9 and 11). However, an efficient/complete voiding will also require relaxation of the urethral sphincter during voiding. If we assume that microstimulation of the GABA-ergic inhibitory neurons around the dorsal commissure area could still inhibit the urethral sphincter activity in chronic SCI animals, one or several microelectrodes will also need to be implanted in this area. Even with the microelectrodes implanted both in the S2 lateral ventral horn or ventral funiculus and in the dorsal commissure area, the stimulation strategy to control micturition after SCI might still be incomplete. Brindley employed sacral posterior root rhizotomy to prevent both DSD and DH (Brindley, 1994; Brindley and Rushton, 1990; Creasey, 1993; Van Kerrebroeck et al., 1996). To improve on Brindley’s method, microstimulation has to be able to inhibit hyperactive bladder activity during urine storage. Although the dorsal horn where the pudendal afferents terminate (Thor et al., 1989; Ueyama et al., 1984) is a promising location (Chartier-Kastler et al., 2001; Kirkham et al., 2002) for microstimulation to inhibit bladder hyperactivity, it needs to be tested in chronic SCI animals since visceral afferents also terminate in this area (Morgan et al., 1981). To identify effective stimulation locations for bladder activation/inhibition and/or urethral sphincter relaxation in chronic SCI animals, experiments using a single microelectrode to acutely probe the entire sacral spinal cord will be more efficient than studies using chronically implanted electrodes, since fewer locations could be tested in one animal with the chronically implanted electrodes. Therefore,
more acute experiments on chronic SCI animals should be performed to identify those effective stimulation locations. Although many questions need to be answered before microstimulation of sacral spinal cord can be successfully applied to control micturition after SCI, the results from this study are encouraging. Bladder pressure suitable for voiding was generated by a single microelectrode. Voiding was also evoked even with a single microelectrode although it was incomplete. Microstimulation of sacral spinal cord is a promising technique to restore the function of the lower urinary tract after SCI.
Acknowledgments This work was supported by the NIH/NINDS under contract N01-NS-2301 and NIH grants 1 R01 DK 068566-1 and 1 P01 HD 39768-02.
References Bagshaw, E.V., Evans, M.H., 1976. Measurement of current spread from microelectrodes when stimulating within the nervous system. Exp. Brain Res. 25, 391 – 400. Bhadra, N., Grunewald, V., Creasey, G., Mortimer, J.T., 2002. Selective suppression of sphincter activation during sacral anterior nerve root stimulation. Neurourol. Urodyn. 21, 55 – 64. Blaivas, J.G., 1982. The neurophysiology of micturition: a clinical study of 550 patients. J. Urol. 127, 958 – 963. Blok, B.F.M., Holstege, G., 1997. The pontine micturition center projects to sacral cord GABA immunoreactive neurons in the cat. Neurosci. Lett. 233, 109 – 112. Blok, B.F.M., van Maarseveen, J.T.P.W., Holstege, G., 1998. Electrical stimulation of the sacral dorsal gray commissure evokes relaxation of the external urethral sphincter in the cat. Neurosci. Lett. 249, 68 – 70. Brindley, G.S., 1994. The first 500 sacral anterior root stimulator implants: general description. Paraplegia 32, 795 – 805. Brindley, G.S., Rushton, D.N., 1990. Long-term follow-up of patients with sacral anterior root stimulator implants. Paraplegia 28, 469 – 475. Burns, A.S., Rivas, D.A., Ditunno, J.F., 2001. The management of neurogenic bladder and sexual dysfunction after spinal cord injury. Spine Suppl. 26, s129 – s136. Carter, R.R., McCreey, D.B., Woodford, B.J., Bullara, L.A., Agnew, W.F., 1995. Micturition control by microstimulation of the sacral spinal cord of the cat: acute studies. IEEE Trans. Rehabil. Eng. 3, 206 – 214. Chancellor, M.B., Gajewski, J., Ackman, C.F.D., Appell, R.A., Bennett, J., Binard, J., Boone, T.B., Chetner, M.P., Crewalk, J.A., Defalco, A., Foote, J., Green, B., Juma, S., Jung, S.Y., Linsenmeyer, T.A., MacMillan, R., Mayo, M., Ozawa, H., Roehrborn, C.G., Shenot, P.J., Stone, A., Vazquez, A., Killorin, W., Rivas, D.A., 1999. Long-term follow-up of the North American multicenter UroLume trial for the treatment of external detrusor–sphincter dyssynergia. J. Urol. 161, 1545 – 1550. Chartier-Kastler, E.J., Denys, P., Chancellor, M.B., Haertig, A., Bussel, B., Richard, F., 2001. Urodynamic monitoring during percutaneous sacral nerve neurostimulation in patients with neurogenic detrusor hyperreflexia. Neurourol. Urodyn. 20, 61 – 71. Cheng, C.L., Liu, J.C., Chang, S.Y., Ma, C.P., de Groat, W.C., 1999. Effect of capsaicin on the micturition reflex in normal and chronic spinal cats. Am. J. Physiol. 277, R786 – R794.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183 Creasey, G.H., 1993. Electrical stimulation of sacral roots for micturition after spinal cord injury. Spinal Cord Inj. 20, 505 – 515. de Groat, W.C., 1975. Nervous control of the urinary bladder of the cat. Brain Res. 87, 201 – 211. de Groat, W.C., 1995. Mechanisms underlying the recovery of lower urinary tract function following spinal cord injury. Paraplegia 33, 493 – 505. de Groat, W.C., Booth, A.M., Yoshimura, N., 1993. Neurophysiology of micturition and its modification in animal models of human disease. In: Maggi, C.A. (Ed.), The Autonomic Nervous System, Nervous Control of the Urogenital System, Auton. Nerv. Syst. vol. 3. Harwood Academic Publishers, London, U.K, pp. 227 – 289. Chap. 8. de Groat, W.C., Vizzard, M.A., Araki, I., Roppolo, J.R., 1996. Spinal interneurons and preganglionic neurons in sacral autonomic reflex pathways. Prog. Brain Res. 107, 97 – 111. de Groat, W.C., Fraser, M.O., Yoshiyama, M., Smerin, S., Tai, C., Chancellor, M.B., Yoshimura, N., Roppolo, J.R., 2001. Neural control of the urethra. Scand. J. Urol. Nephrol. (207), 35 – 43. Grill, W.M., Bhadra, N., Wang, B., 1999. Bladder and urethral sphincter pressures evoked by microstimulation of the sacral spinal cord in cats. Brain Res. 836, 19 – 30. Iwatsubo, E., Koyanagi, T., Wyndaele, J.J., El Masri, W.S., Diokno, A.C., 1999. Urological management of tetraplegic patients. Spinal Cord 37, 1 – 5. Jamil, F., 2001. Toward a catheter free status in neurogenic bladder dysfunction: a review of bladder management options in spinal cord injury (SCI). Spinal Cord 39, 355 – 361. Kirkham, A.P.S., Knight, S.L., Craggs, M.D., Casey, A.T.M., Shah, P.J.R., 2002. Neuromodulation through sacral nerve roots 2 to 4 with a Finetech–Brindley sacral posterior and anterior root stimulator. Spinal Cord 40, 272 – 281. McIntyre, C.C., Grill, W.M., 1999. Excitation of central nervous system neurons by nonuniform electric fields. Biophys. J. 76, 878 – 888. Morgan, C.W., Nadelhaft, I., de Groat, W.C., 1981. The distribution of visceral primary afferents from the pelvic nerve to Lissauer’s tract and the spinal gray matter and its relationship to the sacral parasympathetic nucleus. J. Comp. Neurol. 201, 415 – 440. Morgan, C.W., de Groat, W.C., Felkins, L.A., Zhang, S.J., 1991. Axon collaterals indicate broad interspinal role for sacral preganglionic neurons. Proc. Natl. Acad. Sci. 88, 6888 – 6892. Morgan, C.W., de Groat, W.C., Felkins, L.A., Zhang, S.J., 1993. Intracellular injection of neurobiotin or horseradish peroxidase reveals separate types of preganglionic neurons in the sacral parasympathetic nucleus of the cat. J. Comp. Neurol. 331, 161 – 182. Nadelhaft, I., de Groat, W.C., Morgan, C., 1980. Location and morphology of parasympathetic preganglionic neurons in the sacral spinal cord of the cat revealed by retrograde axonal transport of horseradish peroxidase. J. Comp. Neurol. 193, 265 – 281. North, N.T., 1999. The psychological effects of spinal cord injury: a review. Spinal Cord 37, 671 – 679. Rampal, G., Mignard, P., 1975. Behaviour of the urethral striated sphincter and of the bladder in the chronic spinal cat. Pflugers Arch. 353, 33 – 42.
183
Rattay, F., 1998. Analysis of the electrical excitation of CNS neurons. IEEE Trans. Biomed. Eng. 45, 766 – 772. Rihkhoff, N.J.M., Hendridx, L.B.P.M., van Kerrebroeck, P.E.V., Debruyne, F.M.J., Wijkstra, H., 1997. Selective detrusor activation by electrical stimulation of the human sacral nerve roots. Artif. Organs 21, 223 – 226. Roppolo, J.R., Booth, A.M., de Groat, W.C., 1993. Microstimulation of the sacral cord with two microelectrodes produces additive and synergistic bladder contraction (BC). Abstr. - Soc. Neurosci. 19, 510. Roppolo, J.R., Booth, A.M., de Groat, W.C., 1994. Bladder and external urethral sphincter responses elicited by microstimulation of the sacral spinal cord in chronic spinal cord transected cats. Abstr. - Soc. Neurosci. 20, 24. Rudy, D.C., Downie, J.W., McAndrew, J.D., 1991. alpha-Chloralose alters autonomic reflex function of the lower urinary tract. Am. J. Physiol. 261 (6 pt. 2), R1560 – R1567. Shaker, H.S., Tu, L.M., Robin, S., Arabi, K., Hassouna, M., Sawan, M., Elhilali, M.M., 1998. Reduction of bladder outlet resistance by selective sacral root stimulation using high-frequency blockade in dogs: an acute study. J. Urol. 160, 901 – 907. Sie, J.A.M.L., Blok, B.F.M., de Weerd, H., Holstege, G., 2001. Ultrastructural evidence for direct projections from the pontine micturition center to glycine-immunoreactive neurons in the sacral dorsal gray commissure in the cat. J. Comp. Neurol. 429, 631 – 637. Sievert, K.D., Gleason, C.A., Junemann, K.P., Alken, P., Tanagho, E.A., 2002. Physiologic bladder evacuation with selective sacral root stimulation: sinusoidal signal and organ-specific frequency. Neurourol. Urodyn. 21, 80 – 91. Thor, K.B., Morgan, C., Nadelhaft, I., Houston, M., de Groat, W.C., 1989. Organization of afferent and efferent pathways in the pudendal nerve of the female cat. J. Comp. Neurol. 288, 263 – 279. Tyroch, A.H., Davis, J.W., Kaups, K.L., Lorenzo, M., 1997. Spinal cord injury: a preventable public burden. Arch. Surg. 132, 778 – 781. Ueyama, T., Mizuno, N., Nomura, S., Konishi, A., Itoh, K., Arakawa, H., 1984. Central distribution of afferent and efferent components of the pudendal nerve in cat. J. Comp. Neurol. 222, 38 – 46. Van Kerrebroeck, P.E.V., Koldewijn, E.L., Rosier, P.F.W.M., Wijkstra, H., Debruyne, F.M.J., 1996. Results of the treatment of neurogenic bladder dysfunction in spinal cord injury by sacral posterior root rhizotomy and anterior sacral root stimulation. J. Urol. 155, 1378 – 1381. Wang, B., Bhadra, N., Grill, W.M., 1999. Functional anatomy of the male feline urethra: morphological and physiological correlations. J. Urol. 161, 654 – 659. Yeomans, J., Prior, P., Bateman, F., 1986. Current-distance relation of axons mediating circling elicited by midbrain stimulation. Brain Res. 327, 95 – 106. Yoshimura, N., Smith, C.P., Chancellor, M.B., de Groat, W.C., 2000. Pharmacologic and potential biologic interventions to restore bladder function after spinal cord injury. Curr. Opin. Neurol. 13, 677 – 681.
Bladder and urethral sphincter responses evoked by microstimulation of S2 sacral spinal cord in spinal cord intact and chronic spinal cord injured cats Changfeng Tai, August M. Booth, William C. de Groat, James R. Roppolo* Department of Pharmacology, University of Pittsburgh, Pittsburgh, PA 15261, USA Received 11 February 2004; revised 29 April 2004; accepted 1 July 2004
Abstract Urinary bladder and urethral sphincter responses evoked by bladder distention, ventral root stimulation, or microstimulation of S2 segment of the sacral spinal cord were investigated under a-chloralose anesthesia in cats with an intact spinal cord and in chronic spinal cord injured (SCI) cats 6–8 weeks after spinal cord transection at T9–T10 spinal segment. Both SCI and normal cats exhibited large amplitude reflex bladder contractions when the bladder was fully distended. SCI cats also exhibited hyperreflexic bladder contractions during filling and detrusor–sphincter dyssynergia during voiding, neither was observed in normal cats. Electrical stimulation of the ventral roots revealed that the S2 sacral spinal cord was the most effective segment for evoking large amplitude bladder contractions or voiding in both types of cats. Microstimulation with a stimulus intensity of 100 AA and duration of 30–60 s via a single microelectrode in the S2 lateral ventral horn or ventral funiculus evoked large amplitude bladder contractions with small urethral contractions in both normal and SCI cats. However, this stimulation evoked incomplete voiding due to either co-activation of the urethral sphincter or detrusor–sphincter dyssynergia. Stimulation in the S2 dorsal horn evoked large amplitude sphincter responses. The effectiveness of spinal cord microstimulation with a single electrode to induce prominent bladder and urethral sphincter responses in SCI animals demonstrates the potential for using microstimulation techniques to modulate lower urinary tract function in patients with neurogenic voiding dysfunctions. D 2004 Elsevier Inc. All rights reserved. Keywords: Spinal cord injury; Bladder; Urethral sphincter; Cat, Microstimulation; Sacral spinal cord; Micturition; Electrical stimulation
Introduction Spinal cord injury disrupts the normal storage and elimination functions of the lower urinary tract (de Groat, 1975; de Groat et al., 1993, 2001). The bladder does not store well due to detrusor hyperreflexia (DH) (Burns et al., 2001; Jamil, 2001) and does not empty well due to simultaneous contractions of the bladder and the urethral sphincter (detrusor–sphincter dyssynergia, DSD) (Burns et al., 2001; Iwatsubo et al., 1999). DSD prevents complete * Corresponding author. Department of Pharmacology, University of Pittsburgh, E1356 Biomedical Science Tower, Pittsburgh, PA, 15261. Fax: +1 412 648 1945. E-mail address: [email protected] (J.R. Roppolo). 0014-4886/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.expneurol.2004.07.001
elimination of urine, generates high bladder pressures and necessitates the use of intermittent urethral catheterization to empty the bladder. High bladder pressures also contribute to vesicoureteral reflux and renal failure in long-term SCI patients; whereas residual urine in the bladder and urethral catheterization cause cystitis and infection. In addition, DH reduces bladder storage capacity and can induce incontinence, increase risk for kidney damage, and bladder hypertrophy (Burns et al., 2001; Jamil, 2001). The problems of the lower urinary tract contribute to the large costs for medical care of SCI patients and have a tremendous social and psychological impact on the patients and their families (North, 1999; Tyroch et al., 1997). Currently no oral medication can effectively treat DSD, and the pharmacological treatments
172
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
Fig. 1. The experimental setup to measure the bladder and urethral pressures simultaneously. A ligature was placed near the bladder neck to separate the bladder and urethral pressures. Note: the urethra was left intact during voiding experiments (no catheter in urethra and no ligature around the bladder neck).
for DH (e.g., antimuscarinic agents) are limited by their side effects (Yoshimura et al., 2000). In the 1970s, Brindley et al. developed implantable sacral anterior root stimulators to restore micturition function after SCI. This system is now commercially available and has been implanted in over 2000 patients around the world (Brindley, 1994; Jamil, 2001). To achieve optimal results, sacral posterior root rhizotomy is required to prevent DSD or DH (Van Kerrebroeck et al., 1996). The sacral posterior root rhizotomy, which is irreversible, also results in the loss of reflex sexual and defecation functions (Brindley, 1994; Brindley and Rushton, 1990; Creasey, 1993). Voiding induced by the sacral anterior root stimulation generates non-physiologically high pressures in the bladder (Creasey, 1993), which can induce vesicoureteral reflux and renal failure in the long term (Blaivas, 1982; Chancellor et al., 1999). Many studies (Bhadra et al., 2002; Rihkhoff et al., 1997; Shaker et al., 1998; Sievert et al., 2002) have been conducted in an attempt to selectively activate the bladder without activation of the urethral sphincter, but with only limited success. Functional microstimulation of the sacral spinal cord to restore lower urinary tract function after SCI is a very promising approach (Carter et al., 1995; Grill et al., 1999; Roppolo et al., 1993, 1994). Neurons that control the bladder are in a different region of the sacral spinal cord than neurons innervating the urethral sphincter (de Groat et al., 1996; Nadelhaft et al., 1980; Thor et al., 1989; Ueyama et al., 1984). Thus, microelectrodes implanted in the sacral spinal cord could in theory activate bladder without
activating urethral sphincter if the microstimulation was sufficiently localized. Previous studies in our laboratory (Roppolo et al., 1993, 1994) and other laboratories (Carter et al., 1995; Grill et al., 1999) have obtained encouraging results demonstrating micturition in cats following microstimulation of the spinal cord. Bladder contractions could be evoked without significant urethral sphincter contractions (Grill et al., 1999; Roppolo et al., 1993), or even with relaxation of urethral sphincter (Carter et al., 1995). However, most of these studies were conducted in cats with an intact spinal cord (Carter et al., 1995; Grill et al., 1999; Roppolo et al., 1993). As noted above, the neural control of the lower urinary tract undergoes significant changes after SCI (Cheng et al., 1999; de Groat, 1995; de Groat et al., 1993) including the emergence of DSD and DH, which could alter the effectiveness of spinal microstimulation to induce micturition. More experimental data will be needed to test the feasibility of using microstimulation to restore micturition after SCI. In this study, we compared the bladder and urethral sphincter responses induced by microstimulation of sacral spinal cord in spinal intact and chronic SCI cats. Preliminary results of this study were reported previously in an abstract (Roppolo et al., 1994).
Materials and methods All protocols involving the use of animals in this study were approved by the Animal Care and Use Committee of the University of Pittsburgh School of Medicine. Spinalization Five normal (2.5–3.5 kg, three females and two males) and five chronic SCI cats (2.5 –3.5 kg, three females and two males) were used. Cats were spinalized under halothane anesthesia using aseptic surgical techniques. After performing a dorsal laminectomy at T9–T10 vertebral level, a local anesthetic (lidocaine 1%) was applied to the surface of the spinal cord and then injected into the cord through the dura. The spinal cord was then cut completely between the T9 and T10 spinal segments, leaving the parasympathetic and sympathetic innervations to the lower urinary tract intact. A piece of gel foam was placed between the cut ends (usually a separation of 2–3 mm). The muscle and skin were then sutured and after full recovery from anesthesia, the animal was returned to its cage. Following spinal section, bladders were emptied by manual expression at least twice daily. If manual expression was not successful, a sterile catheter was inserted into the bladder through the urethra to empty the bladder. Antibiotics (Polyflex) were given for 10 days following surgery. Six to 8 weeks following spinal cord transection animals were used for spinal cord microstimulation experiments.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
Animal preparation for microstimulation The animals were anesthetized with a-chloralose (60 mg/ kg, i.v.) following induction with halothane. Systemic blood pressure was monitored via a cannula placed in the right carotid artery. A tracheotomy was performed and a tube was inserted to secure the airway. A catheter for i.v. infusion was introduced into right ulnar vein. Ureters were cut and drained externally. Bladder pressure was recorded via a catheter placed through the bladder dome and connected to a pressure transducer while urethral pressure was recorded via a urethral catheter (see Fig. 1). The urethra was slowly infused continuously with saline solution at the rate of 0.1 ml/min. When the urethral sphincter contracts the resistance to flow of the saline infusion is increased, which will cause a pressure elevation within the urethra. The bladder was infused at the rate of 0.5–2 ml/min to determine the volume to induce a reflex bladder contraction, then bladder volume was set to about 80% of the volume necessary to produce rhythmic bladder activity. The proximal urethra was ligated near the bladder neck to maintain isovolumetric conditions in the bladder and to allow separate recordings of bladder and urethral pressures. For the experiments where the voiding induced by microstimulation was investigated (N = 2 for each type of cats), the urethra was left intact and the urine was collected in a beaker attached to a strain gauge to
173
measure the voided volume. In some of the voiding experiments, stainless steel wire electrodes (0.5 mm in diameter) were inserted into the external urethral sphincter to monitor the EMG activity. A dorsal laminectomy from L4 to the first caudal vertebra exposed the lumbosacral spinal cord and roots. The dura mater was opened and each of the sacral segments was identified. The animal was mounted in a modified Narishige spinal cord investigation apparatus. The hip was supported by metal pins and the dorsal spinal process at the rostral end of the laminectomy was secured with a clamp. Skin flaps were retracted to form a pool, which was filled with Krebs solution or mineral oil. The temperature of the animal was monitored and maintained between 358C and 378C using a heating pad and heating tubes in the Krebs or oil pool. Stimulation Before intraspinal microstimulation, each sacral ventral root (S1–S3) was stimulated with a bipolar (cathode distal to anode) platinum-iridium hook electrode to determine the spinal segment that was most effective in inducing a large bladder response or voiding. Stimulation parameters used for ventral root stimulation were 15 Hz frequency, 0.05 ms pulse width, at intensities between 10 and 500 AA for 10 s. Then the spinal segment that was most effective (always S2)
Fig. 2. Bladder and urethral activity after SCI in a female cat. (A) Urethral pressure increased with a slow saline infusion into the bladder. (B) Urethral pressure decreased with large rhythmic bladder contractions. (C) Urethral pressure decreased with large rhythmic bladder contractions induced by perigenital stimulation. Recordings in A, B and C are from the same animal. On/Off marks the start and stop times for bladder infusion (2 ml/min). PG—perigenital stimulation.
174
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
was probed with fine tipped (200 Am2 surface area) activated iridium microelectrodes that were advanced from the dorsal surface of the spinal cord in 200 Am increments. The microelectrode was the cathode, and the anode was placed in the shoulder muscle of the animals. The stimulation parameters used for spinal cord microstimulation were 15 Hz frequency, 0.2 ms pulse width, at an intensity of 50 AA for 10 s. These stimulation parameters were determined based on preliminary experiments. Longer stimulation durations (0.5–4 min) with a higher intensity
(100 AA) were also tested in some locations in an attempt to evoke more efficient voiding. Stimulation for both ventral roots and spinal cord was monophasic and capacitively coupled. The microstimulation of the sacral spinal cord was always started in the middle of S2 segment adjacent to the dorsal root entry zone. Then the microelectrode was withdrawn and moved 200–400 Am rostral/caudal and/or medial/lateral to an adjacent location where the testing was repeated. In each experiment, only one side of the spinal cord was stimulated (usually the left side) and successive
Fig. 3. Bladder hyperactivity after SCI. (A) Hyperactive bladder contractions during a slow infusion of saline (1 ml/min) in a chronic SCI male cat. Very small volume was voided with each hyperactive contraction at a large bladder volume. Total 60 ml was in the bladder when the infusion was stopped. (B) Bladder hyperactivity in another chronic SCI female cat with less frequent, more irregular hyperactive contractions (total 75 ml infused at rate of 2 ml/min). (C) Bladder and urethral activity in a normal male cat. Hyperactive bladder contractions did not occur during slow infusion of saline (0.5 ml/min), and 90% of the infused volume was evacuated during the reflex voiding (13.5 ml out of the total infused volume of 15 ml). The second panels in A and B are bladder pressure recordings continuous with those in the first panels. The third panel in A is corresponding in time to the second panel in A. Insert in C shows at a faster timebase suppression of sphincter EMG activity during voiding and increased activity following voiding.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
penetrations were made as long as the animal’s physical condition permitted (usually 12–24 h). Histology At the end of each experiment, the region of the sacral cord probed with the microelectrode was removed and fixed in formalin (10%). After the cord was fixed, 70-Am-thick sections were cut serially, stained with cresyl violet and examined using light microscopy. Most of the microelectrode tracks could be identified in these histological sections by lesions made by the microelectrode insertion. The microelectrode locations were then correlated with the pressure recordings.
Results Different types of bladder and urethral sphincter activity after SCI Both SCI and normal cats exhibited reflex activity of the bladder and urethral sphincter during bladder filling (Figs. 2
175
and 3). In Fig. 3A, the hyperreflexic bladder contractions in a SCI cat started at a relatively low volume with small amplitude and short duration, and increased in amplitude with the continuation of infusion. However, bladder hyperreflexia was not observed in all chronic SCI cat (see Fig. 2A), and contraction frequency and amplitude varied between animals (see Fig. 3B). In spinal intact cats, the bladder was not hyperreflexic during bladder filling as shown in Fig. 3C. Bladder pressure during the infusion slowly increased without detectable contractile activity until the initiation of reflex voiding. As shown in Fig. 2A, when the bladder of a SCI cat was filled at a rate of 2 ml/min to a maximum of 40 ml, rhythmic activity in the intra-urethral pressure recording was very obvious and especially large during the large bladder contractions at the end of the infusion. This activity reflected hyperreflexia of the external urethral sphincter. Although no voiding could occur due to the ligature at the bladder neck (see Fig. 1), the co-contraction of the urethral sphincter with the bladder during the large bladder contractions presumably represented detrusor–sphincter dyssynergia (DSD). DSD was also evident in experiments where the urethra was open and the animal could void (N =
Fig. 4. Bladder, urethral pressure and voiding responses evoked by stimulation of either S1, S2, or S3 ventral root. (A) Normal female cat: bladder (upper traces) and urethral pressure (lower traces) responses. (B) SCI male cat: bladder (upper traces) and voiding (lower traces) responses. The voiding volume traces are integrated (cumulative) volume levels. The numbers (10–500) at the bottom of figures A and B represent the stimulus intensities in AA, and the stimulation durations are marked by the short black bars (10 s).
176
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
2). As shown in Fig. 3A, only a very small amount of fluid was voided in a SCI cat with every large bladder contraction (over 50 cm H2O) even at a very large bladder volume (about 60 ml). In contrast, reflex voiding in spinal cord intact cats was very efficient (Fig. 3C) because urethral activity which increased when bladder volume was large decreased during reflex voiding, followed by a short burst of urethral EMG activity which occurred after the large reflex bladder contraction and voiding (Fig. 3C). Although DSD was observed frequently in all of the SCI cats, inhibition of the urethral sphincter during a bladder contraction was also detected. In one cat (Fig. 2), the hyperactive contraction of the urethral sphincter continued during a voiding contraction induced by small bladder volume (Fig. 2A), but was inhibited (Fig. 2B) by very large amplitude bladder contractions induced by an additional 10 ml infusion into the bladder. The pressure traces in Fig. 2B were continuous recordings after those shown in Fig. 2A. The bladder contractions induced in Fig. 2A stopped after a period of time presumably as the bladder accommodated to the initial 40 ml infusion. Additional infusion again induced large bladder contractions and inhibition of urethral sphincter activity. As shown in Fig. 2C, inhibition of the urethral sphincter was also observed during the large bladder contractions induced by perigenital stimulation which was applied by repeatedly stroking (approximate twice per second) the
perigenital skin area with a cotton swab. Fig. 2C is a continuation of the recording in Fig. 2B but with only 80% of the bladder volume (10 ml was withdrawn from the bladder). After removal of 10 ml from the bladder, the hyperactivity of the urethral sphincter appeared again although the baseline pressure of the urethra was low (see the urethral pressure at the beginning in Fig. 2C). The first perigenital stimulation was applied with very light stroking and only induced a small bladder contraction. The second and third periods of stimulation with strong stroking induced bladder contractions larger than the rhythmic contractions and inhibited urethral sphincter. Ventral root stimulation Based on ventral root stimulation the S2 sacral spinal cord was the most effective segment for producing bladder contractions or voiding in both normal and chronic SCI cats. Simultaneous measurements of isovolumetric bladder and isotonic urethral pressures (Fig. 4A) or measurements of isotonic bladder pressures and voiding with urethra open (Fig. 4B) revealed that stimulation of S2 ventral root produced large amplitude bladder contractions (56.5 cm H2O mean F 4.5 SEM) with smaller urethral pressures (27.6 cm H2O mean F 7.7 SEM) and voiding (Fig. 4B). Stimulation of S1 ventral root evoked large amplitude urethral pressures (49.7 cm H2O mean F 11.6 SEM) with
Fig. 5. Bladder and urethral activity evoked by microstimulation of S2 spinal cord in a SCI female cat. Five microelectrode penetrations (marked as A, B, C, D, and E on the spinal cord drawing) were made. Electrode E is parallel with electrode D, but at a location 800 Am rostral to electrode D. The corresponding responses evoked by these five penetrations are also marked by A, B, C, D, and E, respectively. The black bars under each response mark the stimulation durations (10 s). The black bars in sequence also represent a 200-Am ventral advancement of the microelectrode in the spinal cord. The first black bar always represents the electrode at the surface of the spinal cord. The most effective locations to evoke bladder contraction without a significant urethral contraction are marked by symbol b*Q along the histologically identified microelectrode tracks on the spinal cord drawing. The corresponding bladder contractions are also marked by the same symbol. Similarly, the most effective locations for urethral sphincter responses are marked by symbol boQ. Microstimulation: 50 AA intensity, 15 Hz frequency and 0.2 ms pulse width.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
relatively small (13.9 cm H2O mean F 4.1 SEM) or no bladder contractions and no voiding, whereas stimulation of S3 ventral root evoked small amplitude contractions of both bladder (39.2 cm H2O mean F 5.7 SEM) and urethra (18.6 cm H2O mean F 12.5 SEM) without voiding. Microstimulation S2 sacral spinal cord was probed with a single microelectrode to determine the most effective location for inducing large amplitude bladder contractions without a significant contraction of the urethral sphincter. Fig. 5 shows a typical experiment in a chronic SCI cat. The microelectrode was positioned at an angle so that the electrode tracks would pass through the sacral parasympathetic nucleus. The effective locations for eliciting bladder contractions were in the lateral ventral horn and deep in the ventral funiculus, where the sacral preganglionic efferent axons to the bladder are located (de Groat et al., 1996; Morgan et al., 1993; Nadelhaft et al., 1980). The maximum evoked bladder pressure was 72.6 cm H2O mean F 5.2 SEM in SCI cats. The bladder pressure induced by 50 AA microstimulation with a single microelectrode implanted in either lateral ventral horn or ventral funiculus of the S2 spinal cord was in the range of 50–100 cm H2O (see Figs. 5 and 6), which was comparable to the bladder response evoked by stimulation of the S2 ventral root (see Fig. 4). This indicates that a considerable proportion of the bladder efferent nerve fibers in the S2 segment were activated by a single microelectrode. However, the effective locations for eliciting urethral sphincter contractions were in the lateral dorsal horn where
177
the visceral and somatic (urethral) afferent axons are located (Morgan et al., 1981; Thor et al., 1989; Ueyama et al., 1984). The maximal evoked urethral pressure was 17.7 cm H2O mean F 4.0 SEM in SCI cats. The small urethral contractions evoked in the locations marked by the symbol b*Q had a relatively slow time course compared to the large urethral contractions evoked in the dorsal horn (marked by symbol boQ). This indicated that the smooth muscle rather than the striated muscle of the urethra might be activated by microstimulation in the lateral ventral horn or ventral funiculus of the S2 spinal cord, or activated by the evoked bladder contractions via a bladder-urethral reflex. Only small bladder contractions (which were not closely linked temporally with the stimulation) could be induced by stimulation in the dorsal horn in chronic SCI animals (Figs. 5A and B). The most effective locations in S2 spinal cord for inducing large bladder contractions by microstimulation in normal animals (Fig. 6) were similar to those sites in chronic SCI animals (Fig. 5). Voiding could be evoked in either lateral ventral horn or ventral funiculus of the S2 spinal cord in normal cats (Fig. 6). The maximal evoked bladder pressure was 46.1 cm H2O mean F 4.6 SEM. It was also possible to induce a large amplitude reflex bladder contraction and voiding (which outlasted the time of stimulation) in normal animals with the microelectrode in the dorsal horn (Fig. 6A). The effective locations for eliciting urethral sphincter contractions were also in the lateral dorsal horn (Fig. 7B second column). The maximal evoked urethral pressure was 25 cm H2O mean F 5 SEM. Fig. 7 shows the mapping of the bladder pressure (Fig. 7 first column), and urethral pressure or voiding response
Fig. 6. Bladder pressure and voiding evoked by microstimulation of S2 spinal cord in a normal male cat. Five microelectrode penetrations (marked as A, B, C, D, and E on the spinal cord drawing) were made. The corresponding responses evoked by these five penetrations are also marked by A, B, C, D, and E, respectively. The black bars under each response mark the stimulation durations (10 s). The black bars in sequence also represent a 200 Am ventral advancement of the microelectrode in the spinal cord. The first black bar always represents the electrode at the surface of the spinal cord. The most effective locations to evoke a bladder contraction are marked by symbol b*Q along the histologically identified microelectrode tracks on the spinal cord drawing. The corresponding bladder contractions are also marked by the same symbol. The voiding volume traces are integrated (cumulative) volume levels. Microstimulation: 50 AA intensity, 15 Hz frequency and 0.2 ms pulse width.
178
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
groups of animals (Fig. 7 second column). There are more locations in S2 spinal cord of the normal cats for inducing small bladder contractions (pressure less than 30 cm H2O) than in the SCI cats (Fig. 7 first column), but there are more locations in the SCI cats for inducing small urethral contractions (pressure less than 10 cm H2O) than in the normal cats (Fig. 7 second column). However, the interpretation of the mapping results should always keep in mind that the electrode penetrations were targeting the sacral parasympathetic nucleus, which resulted in the central area of the S2 spinal cord being unprobed (as shown in Fig. 7 third column where all tested microstimulation sites are presented). Influence of microstimulation parameters
Fig. 7. Mapping of the bladder, urethral pressure and voiding response in S2 spinal cord. (A) From five SCI animals. (B) From five normal animals. Microstimulation: 50–100 AA intensity, 15 Hz frequency, 0.2 ms pulse width and 10–30 s duration.
(Fig. 7 second column) in S2 spinal cord for SCI (Fig. 7 first row) and normal (Fig. 7 second row) animals. Large bladder contractions (pressure greater than 30 cm H2O) were induced by microstimulation in either lateral ventral horn or ventral funiculus of the S2 spinal cord (Fig. 7 first column) with small urethral contractions (pressure less than 10 cm H2O) or no/incomplete (b90%) voidings (Fig. 7 second column) in both SCI and normal animals. Large urethral contractions (pressure greater than 10 cm H2O) were mostly induced in the lateral dorsal horn in both
The influence of the microstimulation parameters on the evoked responses was tested with an electrode implanted in the S2 lateral ventral horn or ventral funiculus at locations similar to those marked by symbol b*Q in Figs. 5 and 6. Fig. 8 shows the influence of the stimulation parameters on bladder and urethral sphincter responses in a chronic SCI cat. Repeated stimulation evoked large bladder contractions with little attenuation of amplitude (Fig. 8A). Long duration stimulation (4 min) evoked an initial sustained large amplitude bladder contraction lasting over 1 min, which then converted into a series of rhythmic contractions (Fig. 8B). A higher stimulation frequency (15 Hz) evoked larger sustained bladder contractions than a lower frequency (3.5 Hz) (Figs. 8B and C). There was no obvious influence of changing the stimulation parameters on intra-urethral pressure responses (Figs. 8A, B and C) that were very small. The influence of the stimulation parameters and stimulation location on bladder contractions is shown in Fig. 9. At a stimulus intensity of 50 AA and duration of 10 s, only small bladder contractions were evoked at sites in the ventral horn (2.2–2.8 mm depth). However, at a higher intensity (100 AA) and longer stimulation duration (30 s),
Fig. 8. Influence of the stimulation parameters on bladder and urethral sphincter responses in a SCI female cat. (A) Influence of repeated stimulation (frequency 15 Hz, duration 10 s). (B) Low frequency (3.5 Hz) for a longer duration (4 min). (C) High frequency (15 Hz) for a longer duration (1 min). Stimulation durations are marked by the black bars at the bottom of each trace. Other stimulation parameters: intensity 50 AA, pulse width 0.2 ms.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
179
Fig. 9. Influence of stimulation parameters and stimulation location on bladder contractions and voiding in a SCI male cat. The black bars under the bladder pressure trace mark the stimulation durations. The numbers on top of the black bars indicate the stimulus intensities. The numbers below the bars indicate the electrode depths (mm) from the surface of the spinal cord. The microelectrode was advanced ventrally into the spinal cord in 0.2 mm increments to a depth of 2.8 mm, then the stimulus intensity was changed from 50 to 100 AA and the electrode was withdrawn back to 2.4 and 2.2 mm depths, respectively. The voiding volume trace is an integrated (cumulative) volume level. Other stimulation parameters: 15 Hz frequency, 0.2 ms pulse width, 10- or 30-s duration.
Fig. 10. Influence of stimulation parameters on bladder, urethra and voiding responses in normal cats. (A) Different stimulation intensities (female). (B) Different stimulation durations (male). The voiding volume trace is an integrated (cumulative) volume level. The black bars at the bottom of each trace mark the stimulation durations. Other stimulation parameters: frequency 15 Hz, pulse width 0.2 ms. Electrodes in S2 lateral ventral horn.
180
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
larger bladder contractions were evoked at the same locations (2.2–2.8 mm depth). The results shown in Fig. 9 indicated that lower stimulation intensity (50 AA) might be better to determine more precise locations for effective microstimulation, but a higher intensity (100 AA) with longer duration (30 s or 1 min as shown in Fig. 8C) can be used to compensate for a non-optimal placement of the microelectrode. The small bladder contractions evoked by stimulation in the dorsal horn (0.2–1.6 mm depth in Fig. 9) were not all synchronized with the stimulation, indicating that they were not due to direct activation of efferent pathways but were more likely due to indirect activation of preganglionic neurons via stimulation of other elements of a reflex circuit (e.g., afferent axons or interneurons). Similar to chronic SCI cats, higher stimulus intensities and longer stimulation durations also induced larger amplitude, longer lasting bladder contractions in normal cats (Figs. 10A and B). Voiding evoked by microstimulation Voiding was evoked in both spinal intact and chronic SCI cats by a single microelectrode implanted in the S2 lateral ventral horn or ventral funiculus at locations similar to those shown in Figs. 5 and 6. Stimulation with a higher intensity and longer duration was also better in inducing voiding (see Fig. 9). Fig. 11 shows the voided volumes and the corresponding bladder pressures generated by stimulation at an intensity of 100 AA, and duration of 1 min in a chronic SCI cat. The stimulation was repeated three times at approximately 4-min intervals to allow the bladder to accommodate to the smaller volume after each void. Although sustained large amplitude bladder contractions (over 60 cm H2O) were evoked, only small amounts of fluid (about 5 ml) were voided during each stimulation. With a total of 3 min of stimulation, only 15 ml of the 48-ml bladder volume was voided (31%) in this experiment.
Similarly, complete voiding was never observed in other SCI animals with the stimulation strategy shown in Fig. 11. This indicated that the urethral sphincter must have been contracting reflexively in concert with every bladder contraction (Fig. 2A) or activated by every stimulation (Figs. 5C, D and E). Voiding induced by microstimulation in spinal intact cats was also incomplete (b90%) even with a 30 second stimulation at intensity of 100 AA (see Fig. 10B) due to the co-activation of the bladder and urethral sphincter (see Fig. 10A), although a complete (N90%) voiding did occur during reflex micturition induced by bladder distention (see Fig. 3C). The average voiding efficiency during a single void evoked by microstimulation was 13.0 F 0.5% (mean F SEM) in SCI cats and 14.1 F 5% in normal cats.
Discussion This study revealed that microstimulation of the sacral spinal cord with a single electrode in S2 lateral ventral horn or ventral funiculus could induce large amplitude bladder contractions with minimal activation of the urethral sphincter in a-chloralose anesthetized cats with the spinal cord intact or with the spinal cord chronically transected. However, microstimulation elicited relatively inefficient voiding in both groups of animals. SCI animals also exhibited DH, DSD and inefficient voiding in response to bladder distention. Normal animals did not exhibit DH and DSD, and also had efficient reflex voiding in response to bladder distention. This study shows that spinal neural circuitry after SCI is sufficient to allow a single microelectrode to induce large bladder contractions for voiding, but additional procedures are needed to promote urethral relaxation during voiding. The bladder and urethral sphincter responses evoked by microstimulation of S2 sacral spinal cord are similar to those evoked by S2 ventral root stimulation. Both stimulations
Fig. 11. Voiding evoked by microstimulation of S2 sacral spinal cord in a SCI male cat. The black bars mark the stimulation duration (1 min). Total 48 ml saline in the bladder (about 80% of the volume to induce rhythmic bladder contractions). Electrode at 2.2 mm depth in the spinal cord. The voiding volume trace is an integrated (cumulative) volume level. Stimulation: 100 AA intensity, 15 Hz frequency, 0.2 ms pulse width, and 4-min interval between each stimulation.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
evoked large bladder contractions (50–100 cm H2O) with small urethra contractions (Figs. 4–6), and also evoked incomplete voiding (Figs. 4B, 10B and 11). With the stimulus intensity less than 100 AA delivered through a single microelectrode, the probable area of direct activation in the spinal cord is only about 1.0 mm in diameter (Bagshaw and Evans, 1976; Yeomans et al., 1986). Thus, the number of directly activated neurons/axons may still be very small. However, the stimulation via a single microelectrode induced responses similar to those evoked by stimulation of the entire S2 ventral root. This suggests that microstimulation via a single electrode may indirectly activate a complex spinal neural circuit which may spread rostrally and caudally within the same spinal segment and probably into adjacent spinal cord segments through axon collaterals, interneurons and inter-segmental circuits along the spinal cord. Spread to the opposite side of the spinal cord is also possible via axon collaterals (Morgan et al., 1991). In both chronic SCI (Figs. 5 and 7A) and normal (Figs. 6 and 7B) cats, the most effective locations to induce large bladder contractions without a significant contraction of the urethral sphincter were in the S2 lateral ventral horn or ventral funiculus where the efferent axons from the sacral parasympathetic nucleus passed through to exit from the cord (de Groat et al., 1996; Morgan et al., 1993; Nadelhaft et al., 1980). Therefore, it is likely that the axon, not the neuron, was activated by microstimulation (McIntyre and Grill, 1999; Rattay, 1998), although the microelectrodes passed through the sacral parasympathetic nucleus in this study. This result also agrees well with other studies on cats with an intact spinal cord (Carter et al., 1995; Grill et al., 1999). Large contractions of the urethral sphincter were evoked in the S2 lateral dorsal horn in both SCI cats (Figs. 5 and 7A) and normal cats (Fig. 7B), where the afferent axons from the pelvic and pudendal nerves project (Morgan et al., 1981; Thor et al., 1989; Ueyama et al., 1984). The activation of visceral and somatic (urethral) primary afferent axons could induce large amplitude reflex bladder contractions and voiding in normal cats (Fig. 6A), but only induced small bladder contractions in SCI cats that were not well synchronized with the stimulation (Figs. 5A and B, and 9). This indicates that microstimulation in S2 dorsal horn area might involve different reflex mechanisms in normal and chronic SCI cats. Only a spinal reflex mediated by C fiber afferents can be evoked in chronic SCI cats, whereas a spino-bulbospinal reflex mediated by lower threshold Ay fiber afferents might be elicited in normal cats (de Groat, 1995; de Groat et al., 1993). This study revealed different types of bladder and urethral sphincter activity existing among different chronic SCI cats and even at different times in the same animal. Bladder hyperactivity was seen in the animals shown in Figs. 3A and B, but not seen in another animal shown in Fig. 2A. For the same animal shown in Fig. 2, the activity of
181
urethral sphincter could be enhanced by increasing bladder volume or by bladder contractions (Fig. 2A), but it could also be inhibited by very large bladder contractions when bladder volume was increased (Figs. 2B and C). The different types of bladder and urethral sphincter activity in the chronic SCI cats indicates that the neural re-organization or plasticity after SCI (Cheng et al., 1999; de Groat, 1995; de Groat et al., 1993) might occur differently for different animals, since their neuro-urological systems might experience different external stimuli (i.e., How much urine was accumulated in the bladder between the twice daily manual expressions? How much urine was expressed each time? How many times was urethral catheterization required to empty the bladder since this could damage the urethra and increase the urethral resistance?). Thus, there is a need for a more detailed understanding of the mechanisms underlying the synaptic re-organization and recovery of lower urinary function after SCI. Inhibition of the urethral sphincter activity with very large reflex bladder contractions (Figs. 2B and C) indicated that the neural circuitry to inhibit urethral sphincter activity might still be present in some of the spinal animals as noted by Rampal and Mignard (1975). Since the spinal cord was transected at T9–T10 spinal segment, this inhibitory neural circuit must exist in the lumbosacral spinal cord. One inhibitory circuitry could consist of bladder afferent projections to g-aminobutyric acid (GABA) immunoreactive inhibitory neurons in the dorsal commissure area (Blok and Holstege, 1997; Sie et al., 2001) which in turn make synaptic connections with pudendal motor neurons in the Onuf’s nucleus (Blok and Holstege, 1997; Blok et al., 1998; Nadelhaft et al., 1980; Sie et al., 2001; Thor et al., 1989). Although this putative inhibitory pathway was not able to effectively coordinate bladder and sphincter activity after SCI in anesthetized or awake cats (Rampal and Mignard, 1975), it is possible that microstimulation in the dorsal commissure area might be more effective in inhibiting urethral sphincter activity (Blok et al., 1998; Carter et al., 1995). However, in this study, we only targeted lateral sites containing sacral parasympathetic neurons or their axons to activate bladder. Due to the significant re-organization of the neural circuitry after SCI (Cheng et al., 1999; de Groat, 1995; de Groat et al., 1993), the feasibility of facilitating voiding in chronic SCI cats by inhibiting the urethral sphincter with microstimulation of the GABA-ergic inhibitory neurons needs to be tested in the future. Even though small urethral pressures were recorded during large bladder contractions evoked by microstimulation (Figs. 5C, D and E, Fig. 8 and Fig. 10A), an efficient voiding was never observed (Figs. 9, 10B and 11). This indicated that the pressure recording from the urethra using a catheter with a very slow infusion (0.1 ml/min) might not be a good measurement to represent the entire activity along the urethra. A recent study (Wang et al., 1999) in cats showed that the urethral pressure could change as much as 5-fold at different locations along the length of urethra.
182
C. Tai et al. / Experimental Neurology 190 (2004) 171–183
a-Chloralose anesthesia can significantly reduce the autonomic reflex of the lower urinary tract (Rudy et al., 1991), which might have influenced the evoked bladder or urethral reflex responses in this study. However, this study also showed that electrical stimulation (both ventral root and spinal cord) could evoke large bladder or urethral responses, which indicates that a-chloralose anesthesia may have only modest effects on the electrically evoked efferent responses (i.e., stimulation of efferent axons) where reflexes are not dominant. In addition to bladder and urethral contractions, a variety of somatic responses were also seen with ventral root stimulation. These include tail and hindlimb movements and contraction of perineal muscles. These somatic responses were especially large with S1 ventral root stimulation and smaller with S2 or S3 ventral root stimulation. The hindlimb movement with microstimulation of the S2 spinal cord was often absent or very small at the stimulation intensities used in this study. Some tail and perineal activity was seen with microstimulation deep in the S2 ventral horn. A stimulation strategy to control micturition by microstimulation of sacral spinal cord needs to be developed in a chronic SCI animal model before any long-term study with chronic implantation in humans should be carried out. This study suggests that one electrode implanted in the S2 lateral ventral horn or ventral funiculus might be sufficient to generate the necessary bladder pressure for voiding (Figs. 9 and 11). However, an efficient/complete voiding will also require relaxation of the urethral sphincter during voiding. If we assume that microstimulation of the GABA-ergic inhibitory neurons around the dorsal commissure area could still inhibit the urethral sphincter activity in chronic SCI animals, one or several microelectrodes will also need to be implanted in this area. Even with the microelectrodes implanted both in the S2 lateral ventral horn or ventral funiculus and in the dorsal commissure area, the stimulation strategy to control micturition after SCI might still be incomplete. Brindley employed sacral posterior root rhizotomy to prevent both DSD and DH (Brindley, 1994; Brindley and Rushton, 1990; Creasey, 1993; Van Kerrebroeck et al., 1996). To improve on Brindley’s method, microstimulation has to be able to inhibit hyperactive bladder activity during urine storage. Although the dorsal horn where the pudendal afferents terminate (Thor et al., 1989; Ueyama et al., 1984) is a promising location (Chartier-Kastler et al., 2001; Kirkham et al., 2002) for microstimulation to inhibit bladder hyperactivity, it needs to be tested in chronic SCI animals since visceral afferents also terminate in this area (Morgan et al., 1981). To identify effective stimulation locations for bladder activation/inhibition and/or urethral sphincter relaxation in chronic SCI animals, experiments using a single microelectrode to acutely probe the entire sacral spinal cord will be more efficient than studies using chronically implanted electrodes, since fewer locations could be tested in one animal with the chronically implanted electrodes. Therefore,
more acute experiments on chronic SCI animals should be performed to identify those effective stimulation locations. Although many questions need to be answered before microstimulation of sacral spinal cord can be successfully applied to control micturition after SCI, the results from this study are encouraging. Bladder pressure suitable for voiding was generated by a single microelectrode. Voiding was also evoked even with a single microelectrode although it was incomplete. Microstimulation of sacral spinal cord is a promising technique to restore the function of the lower urinary tract after SCI.
Acknowledgments This work was supported by the NIH/NINDS under contract N01-NS-2301 and NIH grants 1 R01 DK 068566-1 and 1 P01 HD 39768-02.
References Bagshaw, E.V., Evans, M.H., 1976. Measurement of current spread from microelectrodes when stimulating within the nervous system. Exp. Brain Res. 25, 391 – 400. Bhadra, N., Grunewald, V., Creasey, G., Mortimer, J.T., 2002. Selective suppression of sphincter activation during sacral anterior nerve root stimulation. Neurourol. Urodyn. 21, 55 – 64. Blaivas, J.G., 1982. The neurophysiology of micturition: a clinical study of 550 patients. J. Urol. 127, 958 – 963. Blok, B.F.M., Holstege, G., 1997. The pontine micturition center projects to sacral cord GABA immunoreactive neurons in the cat. Neurosci. Lett. 233, 109 – 112. Blok, B.F.M., van Maarseveen, J.T.P.W., Holstege, G., 1998. Electrical stimulation of the sacral dorsal gray commissure evokes relaxation of the external urethral sphincter in the cat. Neurosci. Lett. 249, 68 – 70. Brindley, G.S., 1994. The first 500 sacral anterior root stimulator implants: general description. Paraplegia 32, 795 – 805. Brindley, G.S., Rushton, D.N., 1990. Long-term follow-up of patients with sacral anterior root stimulator implants. Paraplegia 28, 469 – 475. Burns, A.S., Rivas, D.A., Ditunno, J.F., 2001. The management of neurogenic bladder and sexual dysfunction after spinal cord injury. Spine Suppl. 26, s129 – s136. Carter, R.R., McCreey, D.B., Woodford, B.J., Bullara, L.A., Agnew, W.F., 1995. Micturition control by microstimulation of the sacral spinal cord of the cat: acute studies. IEEE Trans. Rehabil. Eng. 3, 206 – 214. Chancellor, M.B., Gajewski, J., Ackman, C.F.D., Appell, R.A., Bennett, J., Binard, J., Boone, T.B., Chetner, M.P., Crewalk, J.A., Defalco, A., Foote, J., Green, B., Juma, S., Jung, S.Y., Linsenmeyer, T.A., MacMillan, R., Mayo, M., Ozawa, H., Roehrborn, C.G., Shenot, P.J., Stone, A., Vazquez, A., Killorin, W., Rivas, D.A., 1999. Long-term follow-up of the North American multicenter UroLume trial for the treatment of external detrusor–sphincter dyssynergia. J. Urol. 161, 1545 – 1550. Chartier-Kastler, E.J., Denys, P., Chancellor, M.B., Haertig, A., Bussel, B., Richard, F., 2001. Urodynamic monitoring during percutaneous sacral nerve neurostimulation in patients with neurogenic detrusor hyperreflexia. Neurourol. Urodyn. 20, 61 – 71. Cheng, C.L., Liu, J.C., Chang, S.Y., Ma, C.P., de Groat, W.C., 1999. Effect of capsaicin on the micturition reflex in normal and chronic spinal cats. Am. J. Physiol. 277, R786 – R794.
C. Tai et al. / Experimental Neurology 190 (2004) 171–183 Creasey, G.H., 1993. Electrical stimulation of sacral roots for micturition after spinal cord injury. Spinal Cord Inj. 20, 505 – 515. de Groat, W.C., 1975. Nervous control of the urinary bladder of the cat. Brain Res. 87, 201 – 211. de Groat, W.C., 1995. Mechanisms underlying the recovery of lower urinary tract function following spinal cord injury. Paraplegia 33, 493 – 505. de Groat, W.C., Booth, A.M., Yoshimura, N., 1993. Neurophysiology of micturition and its modification in animal models of human disease. In: Maggi, C.A. (Ed.), The Autonomic Nervous System, Nervous Control of the Urogenital System, Auton. Nerv. Syst. vol. 3. Harwood Academic Publishers, London, U.K, pp. 227 – 289. Chap. 8. de Groat, W.C., Vizzard, M.A., Araki, I., Roppolo, J.R., 1996. Spinal interneurons and preganglionic neurons in sacral autonomic reflex pathways. Prog. Brain Res. 107, 97 – 111. de Groat, W.C., Fraser, M.O., Yoshiyama, M., Smerin, S., Tai, C., Chancellor, M.B., Yoshimura, N., Roppolo, J.R., 2001. Neural control of the urethra. Scand. J. Urol. Nephrol. (207), 35 – 43. Grill, W.M., Bhadra, N., Wang, B., 1999. Bladder and urethral sphincter pressures evoked by microstimulation of the sacral spinal cord in cats. Brain Res. 836, 19 – 30. Iwatsubo, E., Koyanagi, T., Wyndaele, J.J., El Masri, W.S., Diokno, A.C., 1999. Urological management of tetraplegic patients. Spinal Cord 37, 1 – 5. Jamil, F., 2001. Toward a catheter free status in neurogenic bladder dysfunction: a review of bladder management options in spinal cord injury (SCI). Spinal Cord 39, 355 – 361. Kirkham, A.P.S., Knight, S.L., Craggs, M.D., Casey, A.T.M., Shah, P.J.R., 2002. Neuromodulation through sacral nerve roots 2 to 4 with a Finetech–Brindley sacral posterior and anterior root stimulator. Spinal Cord 40, 272 – 281. McIntyre, C.C., Grill, W.M., 1999. Excitation of central nervous system neurons by nonuniform electric fields. Biophys. J. 76, 878 – 888. Morgan, C.W., Nadelhaft, I., de Groat, W.C., 1981. The distribution of visceral primary afferents from the pelvic nerve to Lissauer’s tract and the spinal gray matter and its relationship to the sacral parasympathetic nucleus. J. Comp. Neurol. 201, 415 – 440. Morgan, C.W., de Groat, W.C., Felkins, L.A., Zhang, S.J., 1991. Axon collaterals indicate broad interspinal role for sacral preganglionic neurons. Proc. Natl. Acad. Sci. 88, 6888 – 6892. Morgan, C.W., de Groat, W.C., Felkins, L.A., Zhang, S.J., 1993. Intracellular injection of neurobiotin or horseradish peroxidase reveals separate types of preganglionic neurons in the sacral parasympathetic nucleus of the cat. J. Comp. Neurol. 331, 161 – 182. Nadelhaft, I., de Groat, W.C., Morgan, C., 1980. Location and morphology of parasympathetic preganglionic neurons in the sacral spinal cord of the cat revealed by retrograde axonal transport of horseradish peroxidase. J. Comp. Neurol. 193, 265 – 281. North, N.T., 1999. The psychological effects of spinal cord injury: a review. Spinal Cord 37, 671 – 679. Rampal, G., Mignard, P., 1975. Behaviour of the urethral striated sphincter and of the bladder in the chronic spinal cat. Pflugers Arch. 353, 33 – 42.
183
Rattay, F., 1998. Analysis of the electrical excitation of CNS neurons. IEEE Trans. Biomed. Eng. 45, 766 – 772. Rihkhoff, N.J.M., Hendridx, L.B.P.M., van Kerrebroeck, P.E.V., Debruyne, F.M.J., Wijkstra, H., 1997. Selective detrusor activation by electrical stimulation of the human sacral nerve roots. Artif. Organs 21, 223 – 226. Roppolo, J.R., Booth, A.M., de Groat, W.C., 1993. Microstimulation of the sacral cord with two microelectrodes produces additive and synergistic bladder contraction (BC). Abstr. - Soc. Neurosci. 19, 510. Roppolo, J.R., Booth, A.M., de Groat, W.C., 1994. Bladder and external urethral sphincter responses elicited by microstimulation of the sacral spinal cord in chronic spinal cord transected cats. Abstr. - Soc. Neurosci. 20, 24. Rudy, D.C., Downie, J.W., McAndrew, J.D., 1991. alpha-Chloralose alters autonomic reflex function of the lower urinary tract. Am. J. Physiol. 261 (6 pt. 2), R1560 – R1567. Shaker, H.S., Tu, L.M., Robin, S., Arabi, K., Hassouna, M., Sawan, M., Elhilali, M.M., 1998. Reduction of bladder outlet resistance by selective sacral root stimulation using high-frequency blockade in dogs: an acute study. J. Urol. 160, 901 – 907. Sie, J.A.M.L., Blok, B.F.M., de Weerd, H., Holstege, G., 2001. Ultrastructural evidence for direct projections from the pontine micturition center to glycine-immunoreactive neurons in the sacral dorsal gray commissure in the cat. J. Comp. Neurol. 429, 631 – 637. Sievert, K.D., Gleason, C.A., Junemann, K.P., Alken, P., Tanagho, E.A., 2002. Physiologic bladder evacuation with selective sacral root stimulation: sinusoidal signal and organ-specific frequency. Neurourol. Urodyn. 21, 80 – 91. Thor, K.B., Morgan, C., Nadelhaft, I., Houston, M., de Groat, W.C., 1989. Organization of afferent and efferent pathways in the pudendal nerve of the female cat. J. Comp. Neurol. 288, 263 – 279. Tyroch, A.H., Davis, J.W., Kaups, K.L., Lorenzo, M., 1997. Spinal cord injury: a preventable public burden. Arch. Surg. 132, 778 – 781. Ueyama, T., Mizuno, N., Nomura, S., Konishi, A., Itoh, K., Arakawa, H., 1984. Central distribution of afferent and efferent components of the pudendal nerve in cat. J. Comp. Neurol. 222, 38 – 46. Van Kerrebroeck, P.E.V., Koldewijn, E.L., Rosier, P.F.W.M., Wijkstra, H., Debruyne, F.M.J., 1996. Results of the treatment of neurogenic bladder dysfunction in spinal cord injury by sacral posterior root rhizotomy and anterior sacral root stimulation. J. Urol. 155, 1378 – 1381. Wang, B., Bhadra, N., Grill, W.M., 1999. Functional anatomy of the male feline urethra: morphological and physiological correlations. J. Urol. 161, 654 – 659. Yeomans, J., Prior, P., Bateman, F., 1986. Current-distance relation of axons mediating circling elicited by midbrain stimulation. Brain Res. 327, 95 – 106. Yoshimura, N., Smith, C.P., Chancellor, M.B., de Groat, W.C., 2000. Pharmacologic and potential biologic interventions to restore bladder function after spinal cord injury. Curr. Opin. Neurol. 13, 677 – 681.