Reflexes That Promote Bladder Storage

Two important reflexes may play an important role in modulation of bladder function: the guarding reflex and the bladder afferent loop reflex. Both reflexes promote urine storage (guarding reflex under somatic influence and bladder loop reflex under sympathetic tone). The guarding reflex guards or prevents urine loss from times of cough or other physical stress that would normally trigger a micturition episode. Suprapontine input from the brain turns off the guarding reflex during micturition to allow efficient and complete emptying. The bladder afferent reflex works through sacral interneurons that then activate storage through pudendal nerve efferent pathways directed toward the urethral sphincter. As such, the activity has truly been realized only in cats but has been postulated to exist in humans and to function the same. Similar to the guarding reflex, the bladder afferent reflex promotes continence during periods of bladder filling and is quiet during micturition (Fig. 70–1).

Figure 70–1 The guarding reflex promotes continence and allows the outlet to contract the urinary sphincter during periods of stress (e.g., cough). The brain can turn this reflex off during voiding.

(Modified from Leng WW, Chancellor MB. How sacral nerve stimulation neuromodulation works. Urol Clin North Am 2005;32:11–8.)

Reflexes That Promote Bladder Emptying

Signals from the bladder that may modulate the need to void with fullness, pain, pressure, or stretch may elicit bladder afferent activity through the Aδ or even C fibers. These bladder afferent nerve fibers then synapse with both parasympathetic efferents (bladder-bladder reflex) and parasympathetic urethral efferents (bladder-urethral) reflex. The urge to void may then be translated as an initial activity (inhibitory) of the bladder-urethral reflex to allow the pressure in the urethral outlet to drop immediately before a bladder contraction ensues and simultaneously permit the bladder-bladder reflex to allow a smooth bladder contraction to occur as the reflex is maintained throughout the entire void (de Groat, 1978; de Groat et al, 1981, 1996).

Putative Mechanism of Action of Sacral Neuromodulation

Although our knowledge of how neuromodulation works is evolving, two main theories exist: (1) direct activation of efferent fibers to the striated urethral sphincter reflexively causes detrusor relaxation and (2) selective activation of afferent fibers causes inhibition at spinal and supraspinal levels. Accumulating evidence suggests that activation of somatic sacral afferent inflow at the sacral root level that in turn affects the storage and emptying reflexes in the bladder and central nervous system accounts for the positive effects of neuromodulation on both storage and emptying functions of the bladder (Yoshimura and de Groat, 1992, 1997; Leng and Chancellor, 2005). Malaguti and coworkers (2003), using detection of somatosensory evoked potentials during sacral neuromodulation, concluded that sacral neuromodulation therapy works by sacral afferent activity and concomitant activation of the somatosensory cortex. Because sacral neuromodulation has been clinically proven for both storage (urgency/frequency and urgency urinary incontinence) and emptying (nonobstructive urinary retention) dysfunctions of the bladder, isolating the mechanism of action to the micturition reflex pathway of sacral afferent and efferent pathways alone is challenging. However, by understanding the reflexes that influence the promotion of urine storage or emptying of the sacral micturition reflex pathway one begins to realize how neuromodulation may affect these reflexes and elicit symptomatic and functional improvement of voiding function (Fowler et al, 2000). Animal models of sacral neuromodulation for detrusor overactivity are now being developed and may shed more light on how this technology achieves its benefits (Riazimand and Mense, 2004; Vignes et al, 2009)

Putative Mechanism of Action of Sacral Neuromodulation in Overactive Bladder

The bladder storage and emptying reflexes are modulated by several centers in the brain and may be altered by neurologic injury that effectively unmasks involuntary bladder contractions. Thus, sacral neuromodulation of these primitive reflexes may restore normal micturition (de Groat, 1976). Animal data exist to support the fact that somatic afferent input to the spinal cord can affect the guarding and bladder-bladder reflexes (de Groat, 1978; Chen et al, 1993). It is believed that suppression of interneuronal transmission in the bladder reflex pathway may be how sacral neuromodulation affects detrusor overactivity (de Groat, 1976; Kruse and de Groat, 1993; Leng and Chancellor, 2005). The inhibition by electrical neuromodulation may, in part, modulate the sensory outflow from the bladder through the ascending pathways to the pontine micturition center, thereby preventing involuntary contractions by modulating the micturition reflex circuit but allowing voluntary voiding to occur (Fig. 70–2). The preservation of voluntary voiding may be due to selective avoidance of normal sensory ascending outflow pathways of the bladder from Aδ fibers to the pontine micturition center as well as initiation of the descending pathways from the pontine micturition center to sacral efferent outflow pathways. Therefore, as is seen in clinical practice, sacral neuromodulation may affect and improve the abnormal bladder sensations, involuntary voids, and detrusor contractions but still maintain normal bladder sensations and voluntary voiding patterns.

Figure 70–2 Pudendal nerve afferent firing can modulate and accordingly inhibit the bladder micturition reflex. SNS, sacral nerve stimulation.

(Modified from Leng WW, Chancellor MB. How sacral nerve stimulation neuromodulation works. Urol Clin North Am 2005;32:11–8.)

Putative Mechanism of Action of Sacral Neuromodulation in Urinary Retention

Sphincteric activity can be turned off by brain pathways to allow efficient and complete bladder emptying. If the suprasacral pathways are altered, the guarding and urethral reflexes still exist and cannot be turned off. This may cause retention, as in the spinal cord–injured patient who in turn has detrusor-sphincter dyssynergia resulting in urinary retention. Thus, inhibition of the guarding reflexes may allow urinary retention states to be improved (Fig. 70–3). Originally for urinary retention due to functional outlet obstruction, sacral neuromodulation was thought to directly turn off excitatory flow to the urethral outlet and facilitate bladder emptying. This may be an efferent effect of the primary afferent mechanism of sacral neuromodulation as reported by Dasgupta and colleagues (2005), who described novel neuroimaging evidence for the existence of abnormal interaction between the brainstem and cortical centers in women with urinary retention due to Fowler syndrome or overactive sphincteric activity. The direct therapeutic effect of sacral neuromodulation in patients with sphincter overactivity or functional outlet obstruction appears to be in the afferent restoration of a normal pattern of midbrain activity and decreased cortical activity leading to inhibition of the guarding reflexes.

Figure 70–3 In neurologic disease the supraspinal circuitry is “disconnected” and therefore cannot turn off the spinal guarding reflex, and thus retention occurs. Sacral neuromodulation (SNS) can restore the normal voluntary pattern of micturition by inhibition of the spinal guarding reflex.

(Modified from Leng WW, Chancellor MB. How sacral nerve stimulation neuromodulation works. Urol Clin North Am 2005;32:11–8.)

Criteria for Selection of Patients

Because many lower urinary tract symptoms and dysfunctions are secondary to a neuromuscular etiology, a thorough history and physical examination will often reveal the nature (acute vs. chronic) and help classify the cause (neurogenic, anatomic, postsurgical, functional, inflammatory, or idiopathic). In addition to the unique evaluation of pelvic floor muscle dysfunction (Siegel, 2005), a urinalysis is routinely performed; urine cytology should be considered in patients who present with refractory symptoms of dysuria, urgency, or frequency of urination, because carcinoma in situ and bladder tumors may present as irritative bladder symptoms without hematuria. Further assessment of the bladder function, urodynamic studies including cystometrography, pressure-flow studies, and electromyography of sphincters and pelvic floor muscles are performed on a selected basis, because most non-neurogenic assessments are completed routinely with the use of a voiding diary and a focused physical examination of the pelvis. Electromyography is recommended in suspected cases of neurogenic bladder dysfunction, detrusor-sphincter dyssynergia, or Fowler syndrome and may be considered for evaluation of inappropriate pelvic floor muscle behavior (Dasgupta and Fowler, 2003). As of now, urodynamics have not yielded adequate predictive factors that allow enhanced selection for patients undergoing sacral neuromodulation (Groenendijk et al, 2007). The characteristics of neurogenic bladders, as seen in patients with multiple sclerosis and spinal cord injury, can change with time and disease progression. Therefore, reevaluation with urodynamics and assessment of the upper urinary tracts may be needed when symptoms change despite active medical intervention.

Cystourethroscopy may yield information helpful in making a diagnosis. Anatomic lesions such as urethral stricture, bladder neck fibrosis, trabeculation, and bladder lesions have been found even in women with bladder outlet obstruction. Baseline upper tract imaging is performed in patients with neurologic disease or, if indicated, by physical or baseline studies or a patient’s history.

Sacral neuromodulation is frequently attempted in patients in whom traditional conservative measures (e.g., bladder retraining, pelvic floor biofeedback, and medications) have failed and before more invasive surgical procedures (e.g., enterocystoplasty and urinary diversion). Despite all the studies done to date there are no defined preclinical factors, such as urodynamic findings, that can predict which patients will or will not respond to sacral neuromodulation.

Although most patients are considered candidates for neurostimulation and neuromodulation therapies when more conservative treatment has failed there are some clinical considerations for excluding patients from this therapy. These include significant anatomic abnormalities in the spine or sacrum that may present challenges to gaining access; mental incapacitation of patients who cannot manage their device or judge the clinical outcome; physical limitations that prevent the patient from achieving normal pelvic organ function, such as functional urinary incontinence; and noncompliance of the patient.

Relative contraindications for patients who may be considering or who have an implantable electrical stimulation device are magnetic resonance imaging (MRI) and pregnancy. Magnetic fields produce currents in neuroelectrodes, and there is some concern that the magnetic field from MRI may damage the pulse generator, as discussed later. Many radiologists are reluctant to provide MRI services for patients with implantable electrical stimulation devices despite the anecdotal evidence that no adverse event has occurred when MRI has inadvertently or purposefully been done for emergent reasons or in small trials (Hassouna and Elkelini, 2005). For patients who have InterStim devices in place (see later), the authors advocate removal of the device in preparation for elective MRI based on current manufacturer recommendations. After the MRI procedure, a new neuroelectrode and generator may be placed. Anecdotal reports of patients safely undergoing the study (MRI) with the InterStim implant in place have occurred, but patients should have the devices turned off in anticipation. Because of the potential for teratogenicity or abortion from the effect of electrical stimulation, electrical stimulation has been considered contraindicated in pregnant women with various voiding dysfunctions. However, whether electrical stimulation can cause abortion or malformation is not known. Wang and Hassouna (1999) reported no adverse effects of electrical stimulation on pregnant rats and that termination of pregnancy is not advised for prospective mothers when electrical stimulation has been performed unknowingly in early pregnancy. Women with electrical stimulation devices for pelvic health conditions who become pregnant may simply turn off their devices during pregnancy.

Electrical Stimulation of the Bladder

Transurethral electrical bladder stimulation (TEBS) has been pursued not only for initiating sensory awareness of bladder filling and stimulating detrusor contractility (see later section) but also for increasing bladder capacity at low pressure in pediatric patients with myelomeningocele (Kaplan et al, 1989; Decter et al, 1994). The authors in these two cited references carried on an interesting point-counterpoint discussion in publications regarding the practical benefit of TEBS (Kaplan, 2000; Decter, 2000), which remains controversial.

Both authors seem to be saying the same thing with respect to results but attach a totally different significance to the practical implications of the results. Kaplan (2000) points out that when the procedure was initiated the goal was to provide children with neurogenic bladder dysfunction, mostly secondary to spina bifida, enough sensation to detect a filling or full bladder and to have them synergistically void or catheterize in a timely manner. As the results of treatment have been evaluated over the years, the real benefit of this program, to Kaplan at least, was the potential to increase bladder capacity while maintaining or decreasing end-filling bladder pressure (in essence, improving compliance). Decter (2000) gives what seems to be a reasonable summation of the largest multi-institutional report of this therapy involving 568 patients who underwent TEBS at 11 institutions, only 335 of whom had adequate pretreatment and post-treatment urodynamics for evaluation. Bladder capacity increased by 20% or more in 56% of the 335 patients, whereas pressure at bladder capacity decreased by 25% or more in 16% of those in whom the bladder capacity increased. Decter (2000) calculated that only 30 of 335 patients had both a 20% or more increase in bladder capacity and a 25% decrease in compliance. He reports his experience with 25 patients during a 4-year period as showing that bladder capacity increased more than 20% (values referred to a comparison to age-adjusted bladder capacity) and end-filling bladder pressures showed clinically significant decreases in 29% of patients. Putting this in perspective, he believes “the practical benefits our patients derive seem limited…the urodynamic improvements we achieved after stimulation did not materially alter daily voiding routine (i.e., clean intermittent catheterization) of these children.” Pugach and associates (2000) also reported the results of TEBS in a group of pediatric patients; only 7 of 44 (16%) had safe storage pressures with continence after treatment. There exist limited new data since 2000 on this treatment modality probably owing to other better potential nerve targets.

Sacral Rhizotomy

In most cases, bilateral anterior and posterior sacral rhizotomy or conusectomy converts a overactive detrusor to an areflexic one. This alone may be inappropriate therapy because it also adversely affects the rectum, anal and urethral sphincters, sexual function, and the lower extremities. In an attempt to leave sphincter and sexual function intact, selective motor nerve section was originally introduced as a treatment to increase bladder capacity by abolishing only the motor supply responsible for involuntary contractions. The initial use of this procedure followed the observation that the third anterior (ventral) sacral root provided the dominant motor innervation of the human bladder. To enhance the clinical response and to minimize side effects, differential sacral rhizotomy should always be preceded by stimulation and blockade of the individual sacral roots with cystometric and sphincterometric control.

Although technique refinements, such as percutaneous radiofrequency selective sacral rhizotomy and cryoneurolysis, have occurred, there is still controversy about the role of anterior rhizotomy procedures within a treatment plan for detrusor overactivity. Torrens (1985) summarized successful results in collected groups of patients that ranged from 48% for idiopathic overactivity to 81% for patients classified as having a “paraplegic bladder.” However, as he astutely pointed out, the definition of success varies from one series and from one patient to another. When these procedures are used, they should certainly be preceded by urodynamics and urologic evaluation of the effects of selective nerve blocks before performance, especially in patients without fixed neurologic disease or injury. Even then, unintended effects on pelvic and lower extremity sensory or motor functions may occur with disastrous medical and legal sequelae.

Both Tanagho and Schmidt (Tanagho and Schmidt, 1988; Tanagho et al, 1989) and Brindley (Brindley and Rushton, 1990) have popularized the concept of sensory deafferentation by dorsal or posterior rhizotomy to increase bladder capacity as part of their overall plan to simultaneously rehabilitate storage and emptying problems in patients with significant spinal cord injury or disease. These are patients in whom electrical stimulation also was used to alleviate emptying deficits (see section on emptying disorders). McGuire and Savastano (1984) also mentioned dorsal root ganglionectomy alone in such patients to increase bladder capacity.

Gasparini and colleagues (1992) reported durability of the deafferentation response to selective dorsal sacral rhizotomy up to 64 months after section. The technique involves selecting nerve roots whose intraoperative stimulation provokes an adequate detrusor response. The dorsal and ventral components of these roots are then separated and the dorsal root or roots severed. An increase in bladder capacity of 259 to 377 mL was noted in 16 of 17 patients studied (24 in the original series), with an increase in the volume to the first contraction of 99 to 270 mL. Fourteen patients were cured of incontinence, and 2 improved; the technique failed in 1 patient. Of seven potent men, two experienced a decrease in erectile frequency but were still able to achieve penetration. Bowel and sphincter function were unaffected. Koldewijn and associates (1994) reported on the effects of intradural bilateral posterior root rhizotomies from S2 to S5 with implantation of an anterior root stimulator in a group of patients with suprasacral spinal cord injury. All showed persistent detrusor areflexia afterward, although 2 required subsequent secondary rhizotomy at the level of the conus. A majority showed decreased bladder compliance up to 5 days postoperatively, followed by a rapid increase thereafter.

Brindley (1994) summarized the advantages of bilateral posterior sacral rhizotomy in treatment of voiding dysfunction after spinal cord injury as abolishing reflex incontinence, improving compliance, and abolishing striated sphincter dyssynergia without altering resting tone. Partial or selective procedures are considered only in such patients who retain some sensation or have excellent reflex erections. Madersbacher (2000) comments that posterior sacral rhizotomy for sacral deafferentation of the bladder is best achieved by the intradural approach, which has the advantage that, in this location, motor and sensory fibers can easily be separated, whereas distal to the spinal ganglion, motor and sensory fibers are intermingled and a clear separation is no longer possible. If an intradural procedure is not possible, he believes that a deafferentation at the level of the conus medullaris is preferable to an extradural sacral approach. He mentions that he has treated 65 tetraplegic or paraplegic patients with post–spinal cord injury reflex urinary incontinence who were resistant to all other means of conservative treatment. Incontinence was abolished in 90% of these patients.

Sacral Neuromodulation

Neuromodulation is an innovative treatment of lower urinary tract symptoms and dysfunctions of bladder storage secondary to neuromuscular causes. In addition to the application of evolving technologies for sacral neuromodulation therapy, expanding clinical indications such as neurogenic detrusor overactivity, post–sling-related voiding dysfunction, interstitial cystitis, pelvic pain, pediatric voiding dysfunction, and bowel disorders as well as novel forms of transcutaneous and implantable neuromodulation devices for different nerve roots are under investigation.

Technique

Sacral nerve stimulation (SNS) by the InterStim (Medtronic, Minneapolis, MN) procedure is performed in two stages: stage I, a clinical trial of a temporary or permanent lead for external stimulation; and stage II, implantation of a subcutaneous implantable pulse generator (IPG). Each stage can be performed with monitored anesthesia care supplemented by local anesthesia. During the initial introduction of sacral neuromodulation therapy, patients underwent a percutaneous nerve evaluation by the placement of a unilateral percutaneous lead in the S3 foramen with use of local injectable anesthesia. The lead was connected to an external pulse generator and worn by the patient for several days. A large number of false-negative results with therapy are attributed to improper lead placement and migration. Whereas some physicians still prefer to perform the first stage by a percutaneous nerve evaluation approach, most have adopted a permanent tined lead placement for the first stage in an attempt to avoid the issues related to high false-negative results with the first stage and high false-positive results with the second stage. Changes in lower urinary tract symptoms and postvoid residuals are recorded in a detailed bladder diary. If improvement is minimal or absent, revision or bilateral percutaneous lead placement may be attempted. If more than 50% improvement in symptoms of urgency/frequency or urgency urinary incontinence is attained, a permanent IPG is implanted. The length of the trial with the external pulse generator may vary slightly from patient to patient, by the indication, and by the surgeon’s practice preference. In patients with urgency/frequency and urgency urinary incontinence, a 1- to 2-week trial is generally adequate. For retention, a longer trial of 3 to 4 weeks or more may be necessary before a desired clinical response is obtained.

Previous lead placement required a more time-consuming surgical dissection of the layers above the sacral foramen and unreliable lead fixation with anchors. Recent technical advances have made implantation of the percutaneous lead easier and less prone to migration. Spinelli and colleagues (2003) were the first to present the advantages of the tined lead that used a percutaneous approach for placement and fixation (Fig. 70–4). Subsequent large-scale clinical experience worldwide confirms that the tined lead is less prone to migration and has decreased false-negative results with the screening trial. Furthermore, the false-positive rate of the screening trial is reduced when placement of a permanent lead with reliable fixation during the screening trial ensures that the same location of stimulation is achieved when the IPG is implanted. With a percutaneous nerve evaluation or similar temporary lead electrode during the screening trial, a different clinical outcome may occur when the permanent lead is placed at the time of the IPG implantation.

Figure 70–4 The tined lead is introduced typically into the S3 nerve foramen. The “tines” allow the lead to be fixed into the fascial layers above the sacrum. This lead has a quadripolar configuration (four contact points).

(Courtesy of Medtronic, Minneapolis, MN.)

For the first stage of the procedure, preoperative intravenous antibiotics are given before the procedure and aseptic techniques of foreign body implants are implemented. The patient is placed in the prone position, and the buttocks are held apart by wide tape retraction so that the anus is visible during test stimulation. The anus and tape are prepared in a sterile fashion and then covered with a separate plastic drape until visualization is needed during the procedure. The sterile drape covering the feet must be folded back such that the feet can be visualized during the procedure as well.

The location of the S3 foramen is approximated by measuring 9 cm cephalad to the dropoff of the sacrum and 1 to 2 cm lateral to the midline on either side. Alternatively, the site may also be localized by palpating the cephalad portions of the sciatic notches bilaterally and drawing a connecting line that intersects the midline of the sacrum; one fingerbreadth on either side of the midline of the sacrum at this intersection will define the location of the S3 foramen (Fig. 70–5). The foramen needle is then inserted into the S3 foramen. The pelvic plexus and pudendal nerve run alongside the pelvis, and therefore the needle should be placed just inside the ventral foramen. The position of the needle is confirmed by fluoroscopy. The nerve is tested for the appropriate motor response, which is dorsiflexion of the great toe and bellows contraction of the perineal area, which represents contraction of the levator muscles (bellows reflex). Simultaneous sensory responses determined at the time of lead placement help optimize positioning. Although this is controversial, if one can localize the stimulation at the time of lead positioning to the vagina-rectum juncture in females and perineoscrotal area in males this is proper localization of the device. The foramen needle stylet is then removed and replaced with the introducer sheath. The distal aspect of the lead consists of four electrodes numbered 0 through 3. The lead is placed into the introducer sheath as directed to expose the electrodes. Typically, electrodes are positioned such that electrodes 2 and 3 straddle the ventral surface of the sacrum (Fig. 70–6). Test stimulation is repeated on each electrode, and the responses are observed. An S3 response should be noted on at least two of the electrodes (see Table 70–2). Once the surgeon is satisfied with the position, the sheath is removed, releasing the tines that anchor the lead. A sensory response, sensation of stimulation in the perineum, is not needed to confirm proper placement if the correct S3 motor response is observed. However, when a motor response is absent, raising the conscious level of the patient during the procedure and detecting the correct sensory response will confirm proper localization, and a clinical response may still be obtained during the screening trial period despite the absence of the motor response.

Figure 70–5 A, The percutaneous test stimulation is performed in an outpatient setting. A small lead wire is placed into S3 and connected to an external stimulator for 1 week to administer stimulation to the nerve roots. B, Measurement of the S3 nerve foramen is typically 9 cm to the coccyx and 11 cm to the anal verge. One measures 1 to 2 cm lateral to this mark to find the rough site of the S3 nerve foramen. The “cross hair” technique can be used to find the midline fluoroscopically and the lower aspect of the sacroiliac joints laterally to find the S3 nerve foramen as well.

(Courtesy of Medtronic, Minneapolis, MN.)

Figure 70–6 Fluoroscopy is used to confirm lead placement, typically with the lead configurations to obtain optimal muscle (bellows response and ipsilateral toe contractions) and sensory (vaginal, penile, or scrotal “pulsating” feeling) responses. The lead may then be deployed.

A 3- to 4-cm incision into the subcutaneous tissues in the upper lateral buttock is made below the beltline or below the level of the ischial wings for connecting the permanent lead to the percutaneous extension lead wire. If the screening trial is successful, this connection site will be the site of implantation for the IPG. With use of the tunneling device provided in the commercial kit, the permanent lead is transferred to the medial aspect of the lateral buttock incision. The lead is then connected to the extension wire, and the tunneling device is used again to transpose the extension wire from the medial aspect of the incision to an exit point on the contralateral side of the back. This transfer and long tunnel reduce the occurrence of infection from the percutaneous exit site of the wire. The extension wire is connected to the external pulse generator. Patients are able to resume their normal activities immediately but are advised to limit excessive movement-related activities, such as high-impact exercises, for the duration of the trial period.

The external generator can be flexibly programmed for the duration of the intended trial while the patient records symptoms and bladder function in a voiding diary. If there is more than 50% improvement in the symptoms or voiding function, a stage II procedure is performed.

A stage II procedure entails placement of the IPG. No fluoroscopy is required during stage II when a permanent neuroelectrode has been placed for the stage I procedure; however, if a percutaneous nerve evaluation was performed for stage I, fluoroscopic confirmation of the neuroelectrode placement is advised. The patient may be placed in the prone position or a lateral position with the site of the previous lateral incision for the lead connections placed upward (Fig. 70–7). The lateral position may improve ventilation during sedation. The previous buttock incision overlying the lead connections is opened, the percutaneous extension wire is removed, and the extension lead is secured to the permanent lead and subsequently to the IPG. A pocket is made in the subcutaneous tissue that is large enough to avoid tension on closure and at a depth to provide a covering layer of subcutaneous tissue anterior to the pulse generator to prevent erosion.

Figure 70–7 A 3- to 4-cm counterincision in the upper gluteal crease is made for a deep subcutaneous pocket to allow implantation of the implantable pulse generator.

Bilateral Stimulation and Neuromodulation

The current technique for sacral neuromodulation involves a unilateral lead at the S3 nerve foramen to achieve results in cases of urgency, frequency, urgency urinary incontinence, and idiopathic nonobstructive urinary retention. Bilateral stimulation has been suggested as an alternative, particularly in failed unilateral lead placements, for potential salvage or added benefit as the bladder receives bilateral innervation (van Kerrebroeck et al, 2005). The initial consideration of bilateral stimulation was based on animal studies demonstrating that bilateral stimulation yielded a more profound effect on bladder inhibition than did unilateral stimulation (Schultz-Lampel et al, 1998a, 1998b). An animal model of unilateral versus bilateral stimulation has suggested that bilateral stimulation may be more effective overall (based on reduction of detrusor overactive contractions) than unilateral stimulation (Kaufmann, 2008).

There has been only one prospective clinical study to demonstrate the differences in unilateral versus bilateral stimulation (Scheepens et al, 2002). This study was a prospective randomized crossover design in which all patients underwent unilateral as well as bilateral test stimulations to assess the benefits of bilateral stimulation. Both unilateral and bilateral test stimulation was continued for 72 hours, and the patients were randomly assigned to start with unilateral or bilateral stimulation. No significant difference was found in the unilateral versus bilateral group with regard to urgency urinary incontinence, frequency, or severity of leakage in the overactive bladder group, although overall results were impressive in both categories. The retention group had better parameters of emptying (volume per void) in bilateral compared with unilateral stimulation. Still, the numbers were too small in the retention group for adequate conclusions to be made. It appears that the data as presented, at least from a clinical perspective, do not suggest a large role for routine bilateral stimulation for most patients. Perhaps there will be subgroups that may benefit more than others (e.g., retention patients), but larger-scale studies with good methodology as shown in the Scheepens study will be required. Still, if one could increase the overall success rates of patients undergoing sacral neuromodulation, one could ultimately help more patients. Accordingly, Pham and colleagues (2008) examined 124 patients undergoing stage I sacral neuromodulation and stratified patients into unilateral and bilateral groups, retrospectively. Successful stage 1 trials were noted in 58% of unilateral patients and 76% of bilateral patients. An important component that still needs to be evaluated is whether it is cost effective to “routinely” place bilateral leads in the setting of most unilateral lead success rates approaching 70% and 80%.

Selective Nerve Stimulation

Transcutaneous Electrical Stimulation

Other methods of electrical stimulation have been used that seem to occupy a place midway between anal, vaginal, or perineal stimulation and sacral root stimulation. TENS devices have been used to limited degrees to achieve better tolerability to bladder filling and may have some efficacy in postponing voiding. Fall and associates (1980) described TENS use suprapubically in patients with interstitial cystitis, and subsequent studies have been done to gain wider use of this modality (Lindstrom et al, 1983; Fall and Lindstrom, 1991). The exact stimulation parameters are not agreed on because different frequencies have been used; 2 Hz may stimulate pudendal afferents, whereas 50 Hz may stimulate striated paraurethral musculature. Similarly, low-frequency TENS may have some use in abolishing detrusor contractility (Bower et al, 1998). As such, this technology is easy to do and apply, but it may be required for extended periods to gain treatment benefits. S2 or S3 TENS may make some sense, because direct stimulation transcutaneously of this area may yield better results than suprapubic stimulation. Positive results have been demonstrated on the basis of urodynamic data, with improved bladder capacity, delay in first urge to void, and reduced detrusor instability (Bower et al, 2001; Hoebeke et al, 2001). For adequate maintenance of the benefits of this therapy it must also be continued for longer durations. McGuire and associates (1983) described 16 patients with involuntary bladder contractions of varying etiology who were treated with common peroneal or posterior tibial nerve patch electrode stimulation: 12 patients initially were dry, 3 were improved, and 1 was “possibly improved.” Vereecker and colleagues (1984), however, were unable to suppress hyperactivity by this method in patients with suprasacral spinal cord injury or disease. Okada and associates (1998) reported a positive experience with transcutaneous stimulation of the thigh muscle in 19 patients with detrusor overactivity; the maximal cystometric capacity was increased by 57% in 11 of 19 patients.

Noninvasive magnetic stimulation of the sacral roots will also inhibit bladder contractions and cause effects that will persist for short times beyond the period of stimulation. This type of stimulation at present cannot be applied for prolonged periods and is currently unsuitable for long-term treatment, although it may be helpful for preliminary assessment of candidates for chronic sacral root neuromodulation. The extracorporeal magnetic stimulation provided by the “chair” has been used clinically for potential use in overactive bladder, stress incontinence, and pelvic pain (Galloway et al, 2000). No other prospective data on this technology are available for overactive bladder and, furthermore, the exact mechanism of action, if effective, remains to be explained, namely, magnetic, nerve root or peripheral nerve, or intramural nerve stimulation.

Complications and Troubleshooting of Sacral Neuromodulation

With the widespread adoption of sacral neuromodulation, an increasing need has developed to understand complications of this therapy and to learn how to troubleshoot the devices when responses change. It appears that the introduction of the tined lead concept has changed the frequency and profile of the complications that were once only technology related while keeping the patient-related complications at the same frequency.

Published Series

The SNS study group has published several reports on the efficacy and safety of the procedure for individual indications (Siegel et al, 2000). The complications were pooled from the different studies on the basis of the fact that the protocols, devices, efficacy results, and safety profiles were identical. The studies recruited 581 patients, 219 of whom underwent implantation of the InterStim system. The complications were divided into percutaneous test stimulation–related and postimplant-related problems. Of the 914 test stimulation procedures done on the 581 patients, 181 adverse events occurred in 166 of these procedures (18.2% of the 914 procedures). Most complications were related to lead migration (108 events, 11.8% of procedures). Technical problems and pain represented 2.6% and 2.1% of the adverse events, respectively. For the 219 patients who underwent implantation of the InterStim system (lead and generator), pain at the neurostimulator site was the most commonly observed adverse effect at 12 months (15.3%). Surgical revision of the implanted neurostimulator or lead system was performed in 33.3% of cases (73 of 219 patients) to resolve an adverse event. This included relocation of the neurostimulator because of pain at the subcutaneous pocket site and revision of the lead for suspected migration. Explant of the system was performed in 10.5% for lack of efficacy. One should consider the fact that, at the time, the generator was implanted in the lower abdomen.

Everaert and associates (2004) reported specifically on the complications with SNS. This was a retrospective study of 53 patients who had undergone implantation of the quadripolar electrode (Medtronic InterStim, model 3886 or 3080) and subcutaneous pulse generator in the abdominal site (Medtronic Itrel 2 IPG) between 1994 and 1998. Device-related pain was the most frequent problem and occurred equally in all implantation sites (sacral, flank, and abdominal). This occurred in 18 of the 53 patients (34%) and was more frequent in patients with dysuria and retention or perineal pain. Pain responded to physiotherapy in 8 patients, and no explantation was done for pain reasons. Current-related complications occurred in 11%. They performed 15 revisions in 12 patients. Revisions for prosthesis-related pain (n = 3) and for late failures (n = 6) were not successful. Similar series have been published by White and associates (2009) and show relatively low rates (30%) of adverse events and may have been predicted by trauma, body mass index and enrollment into a pain clinic, and history of adverse events.

A review (Hijaz and Vasavada, 2005) of the tined lead approach was performed at the Cleveland Clinic from June 2002 to June 2004 when 167 patients underwent sacral neuromodulation for indications of refractory overactive bladder, idiopathic and neurogenic urinary retention, and interstitial cystitis. In this cohort, 180 stage I operations used the tined lead approach. After 2 to 4 weeks of test stimulation, 130 (72.2%) patients proceeded to stage II implantation of the IPG.

Stage I complications can lead to explantation or revision of the tined lead. The reasons for either fall under response related, mechanical, or infection related. In this series, 50 tined leads were explanted (27.8%). The majority of lead explantations were performed for unsatisfactory or poor clinical response (46 of 50; 92%). The rest of the explantations were done for infection (4 of 50; 8%). Explantation for response reasons is not truly considered a complication as much as it is an integral part of the procedure. Stage I revisions totaled 22 of the 180 operations (12.2%). Revisions were done for marginal response (13 of 22), frayed subcutaneous extension wire (6 of 22), lead infection (3 of 22), and improper localization of stimulus (1 of 22). Eleven (50%) of the revisions proceeded to stage II generator implantation. When the revision was done for a marginal response (13 of 22), the response was ultimately clinically satisfactory in 5 of 13 (38.5%), and patients proceeded to generator implantation. Typically, when the patient reported a marginal or equivocal response during the test stimulation in the absence of infection or mechanical problems, a lead revision was offered with intraoperative sensory testing. Because 38.5% of these revisions eventually were successful and patients proceeded to stage II, this is an option to keep in mind in motivated patients with equivocal response.

As in stage I, stage II complications can be divided into explantation (generator and lead) or revision. Explantation was performed in 16 of 130 (12.3%). Explantations were done for infection and failure to maintain response in 56.3% and 43.7%, respectively. Revisions were done for infection, mechanical (generator related), and response causes. The revision rate of stage II in this series was 20% (26 of 130).

When infection at the generator site is diagnosed, the best management is explantation of the whole system.

Response-related complications necessitating revision are more common (18 of 26). The algorithm for management of a patient who presents with a decreased or absent response after a successful interval is outlined in Figure 70–8.

Figure 70–8 Diagnostic algorithm and troubleshooting for recurrent symptoms. IPG, implantable pulse generator.

The outlined algorithm includes testing of impedances (Fig. 70–9). Impedance describes the resistance to the flow of electrons through a circuit. Impedance or resistance is an integral part of any functioning circuit. However, if there is too much resistance, no current will flow (open). If there is too little resistance, excessive current flow results in diminished battery longevity (short). The electrical circuit that we are referring to starts at the neurostimulator’s circuitry and goes through the connectors to the extension wires, through the extension connector to the lead wires, through the lead’s electrodes to the patient’s tissue, and back either through another electrode and up the same path to the circuitry (bipolar) or to the neurostimulator case and into the circuitry (unipolar).

Figure 70–9 Intraoperative algorithm for impedance problem management. IPG, implantable pulse generator.

If the circuit is broken somehow, electrons cannot flow. This is called an open circuit, and impedance measurements are high. Open circuits can be caused by a fractured lead or extension wires and loose connections. Patients generally feel no stimulation if an open circuit is present. In measurement of impedances by the programmer, unipolar measurements are most useful for identifying open circuits because they take one lead wire measurement at a time, immediately identifying which connection or wire has the problem.

Short circuits, which are reflected in low impedance measurements, can be caused by body fluid intrusion into the connectors or crushed wires that are touching each other. The electrons will always follow the path of least resistance. Patients may or may not feel stimulation, or stimulation may not be present in the correct area (i.e., the generator site) or may vary in strength (i.e., a surging sensation). In measurement of impedances by the programmer, bipolar measurements are most useful for identifying shorts between two wires.

Therefore, impedance measurement is used as a troubleshooting tool to check the integrity of the system when a patient presents with a sudden or gradual disappearance of stimulation. Many measurements fall within the 400- to 1500-ohm range. High levels (>4000 ohms) identify open circuits, and low levels (<50 ohms) identify short circuits. Medtronic (manufacturer of the InterStim) recommends performing the impedance measurements at the time of closure of the incision, at the first programming session, to get a baseline measurement and at any time a problem is suspected. These measurements will identify which electrodes, if any, are intact and allow the programmer to proceed with programming of only those with acceptable impedance measurements. If all electrode measurements read above 4000 ohms, a revision may be necessary.

The intraoperative algorithm for management of impedance problems includes initial testing of impedances (see Fig. 70–9). First, the tined lead is disconnected from the extension to the IPG and then dried. The connection may be irrigated with sterile water, and then the connection is dried with the 3-Fr ENT suction device before it is reconnected again. At this stage the impedances are repeated. If they normalize, the revision can be concluded at this stage. If impedances continue to be abnormal, one can evaluate or change the 10-cm extension and retest impedances. If they continue to be abnormal, the lead is revised. It has been the authors’ experience that the connections to the IPG and the IPG itself seldom have anything to do with abnormal impedance values.

Troubleshooting Algorithm

After successful completion of stage II, a number of events can occur, and the treating physician should develop an algorithm to handle these events in a timely and efficient manner. These events with their probable causes and the troubleshooting algorithm are covered in this section.

Pocket (IPG Site) Discomfort

The probable causes of this symptom are pocket related or output related. Pocket-related causes of discomfort include infection, pocket location (waistline), pocket dimension (too tight, too loose), seroma, and erosion. Output-related causes include sensitivity to unipolar stimulation if this mode is used or current leak. To troubleshoot this problem the evaluating specialist is advised to do the following (Fig. 70–10):

1. Turn off the device and ask the patient if the discomfort is still present to differentiate pocket-related from output-related causes.

2. If the discomfort persists, the cause is not related to the device output. In the absence of clinical signs of infection, pocket-related causes such as pocket size, seroma, and erosion must be considered.

3. If the discomfort disappears, device output is probably causing discomfort. If the stimulation program is unipolar, switch to bipolar and see whether that eliminates discomfort. Some patients are sensitive to the unipolar mode of stimulation, because the positive pole is the neurostimulator itself. Another possibility is leakage of fluid into the connector. This somehow creates a short circuit whereby the current from the device now follows this fluid pathway out to the patient’s tissue. Most patients report this as a burning sensation. Even though current is following this fluid out to the patient’s tissue, some of the current may also be getting to the electrodes as well, so some patients feel both burning in the pocket and stimulation in the perineum. One can try reprogramming around this by using different electrode combinations. If reprogramming is unsuccessful, the patient is asked if the “burning” sensation is tolerable (it will not harm the patient’s tissues); if it is not tolerable, a revision may be necessary to dry out the connection sites.

Figure 70–10 Troubleshooting algorithm for implantable pulse generator site discomfort.

Recurrent Symptoms

When the patient presents with recurrent symptoms, evaluation of the stimulation perception is necessary. The possibilities are that the patient perceives the stimulation in a wrong location compared with baseline, has no stimulation, or has intermittent stimulation.

Wrong Location

If the patient reports that the stimulation location or pattern has changed, it is best to go back to each unipolar setting and “map” out where the patient feels the stimulation. The device is set to 0−, case+ and the patient is asked where she or he feels the sensation; next it is set to 1−, case+ and the patient again is asked about the sensation; next it is set to 2−, case+ and finally to 3−, case+. If these combinations do not confirm the target area, the next step is to start programming bipolar combinations. When those are exhausted, sometimes increasing the pulse width widens the stimulation area. If the programming possibilities are exhausted, then revision for lead repositioning or relocation to the other side may be necessary.

No Stimulation

The obvious is checked first. The device parameters must be set high enough, an inadvertent on-off is checked (set magnet switch off to avoid inadvertent magnet activations), and whether the IPG is nearing the end of its life is checked. Next impedance readings are performed, paying close attention to unipolar impedances. These impedances measure one lead wire with the case, so it is easy to isolate a problem. Using unipolar impedances, it is possible to tell which lead wires are still intact and which ones are not, as mentioned previously. Programming of the electrodes is then continued with acceptable impedance measurements. Bipolar measurements are checked to rule out short circuits as well (very low impedance measurements). If programming around the malfunctioning lead does not restore the stimulation, the patient will often need to undergo revision.

Intermittent Stimulation

Again, inadvertent on-off is checked. Intermittent stimulation can be caused by either a loose connection or positional sensitivity. If a loose connection is suspected, palpating the connection site and re-creating the intermittency is a good clue as to where the problem lies. Taking impedances while the patient reports the stimulation intermittently determines whether the problem is positional (acceptable impedances are still present) or mechanical (when the patient feels stimulation go off, the impedances are high). With positional sensitivity, the lead position shifts when a patient moves in a certain direction (for instance, the patient reports that the stimulation goes away on standing). The lead position may have moved farther from the nerve during standing, and the amplitude may just need to be increased. Intermittent stimulation represents a challenging dilemma to troubleshoot.

Electrical Stimulation Directly to the Bladder or Spinal Cord

Clinical trials of direct electrical stimulation of the bladder to facilitate emptying originated in 1940 but have met with only partial success and intermittent enthusiasm since then (Wein and Barrett, 1988). Direct electrical stimulation was most effective in patients with hypotonic and areflexic bladders. Initial success, defined as low postvoid residual urine volume with sterile urine, was achieved in only 50% to 60% of patients, and secondary failure often supervened, usually related to fibrosis, electrode malfunction, bladder erosion, or other equipment malfunction. The spread of current to other pelvic structures with a stimulus threshold lower than that of the bladder often resulted in abdominal, pelvic, and perineal pain; a desire to defecate or defecation; contraction of the pelvic and leg muscles; and erection and ejaculation in male patients. It was also noted that the increase in intravesical pressure was generally not coordinated with bladder neck opening or with pelvic floor relaxation and that other measures to accomplish these ends could be necessary. Direct electrical stimulation of the sacral spinal cord was also performed as an attempt to take advantage of the remaining motor pathways to initiate micturition. Although some short-term success was noted, many of the side effects seen with direct bladder stimulation occurred as well because the stimulus applied in this way was also unphysiologic. Enthusiasm for both of these approaches has waned considerably, and resurrection seems unlikely.

Electrical Stimulation to the Nerve Roots

For the past 25 to 30 years, Brindley (1993) and the Tanagho group (Tanagho and Schmidt, 1988; Tanagho et al, 1989) have pursued neurostimulation for the treatment of voiding dysfunction. The use of electrical stimulation for storage disorders and pelvic floor dysfunction has been covered. The focus of this section is on the use of anterior root electrical stimulation to facilitate emptying.

The Brindley device is the one most commonly used. Prerequisites for such use are described by Madersbacher and Fischer (Fischer et al, 1993) as (1) intact neural pathways between the sacral cord nuclei of the pelvic nerve and the bladder and (2) a bladder that is capable of contracting. The chief applications are in patients with inefficient or nonreflex micturition after spinal cord injury. Simultaneous bladder and striated sphincter stimulation is obviated by sacral posterior rhizotomy, usually complete, which eliminates reflex incontinence and improves low bladder compliance, if it is present (Brindley, 1994). The stimulation sequences and parameters themselves and their neurophysiologic consequences lead to less striated sphincter dyssynergia, even without posterior rhizotomy, than is seen in reflex micturition in a spinal cord–injured patient. Complete sacral deafferentation is usually performed, however, with the exception listed by Brindley as those patients who have genital sensation or useful reflex erections. Electrodes are applied intradurally to S2, S3, and S4 nerve roots, but the pairs can be activated independently (Fig. 70–11). The detrusor is usually innervated primarily by S3 and to a smaller extent by S2 or S4. Rectal stimulation is by means of all three roots equally. Erectile stimulation is chiefly by S2 with a small contribution from S3 and none from S4. Micturition, defecation, and erection programs are possible, with stimulus patterns set specifically for each patient (Brindley, 1994). The ventral sacral roots, however, carry both parasympathetic fibers to the bladder and somatic fibers to the striated sphincter. Ventral root stimulation therefore occasionally results in detrusor–striated sphincter dyssynergia.

Figure 70–11 Brindley-Finetech system for sacral root stimulation.

(Courtesy of NDI Medical, Cleveland, OH.)

The current Brindley stimulator uses the principle of post-stimulus voiding, a term first introduced by Jonas and Tanagho (1975) to obviate this. Relaxation time of the striated sphincter after a stimulus train is shorter than the relaxation time of the detrusor smooth muscle. When interrupted pulse trains are used, voiding is achieved between the pulse trains because of the sustained high intravesical pressure. This concept and other methods that are available to overcome the stimulation-induced sphincter dyssynergia and allow low-pressure emptying are nicely reviewed by Rijkhoff and colleagues (1997). Post-stimulus voiding has a few shortcomings, described in this article as follows: voiding occurs in spurts at above-normal bladder pressures; when the stimulus parameters are not properly adjusted the detrusor pressures can become too high, putting the upper tracts at risk; and movement of the lower limbs occurs during stimulation because the nerve roots also contain fibers innervating leg musculature, and this movement can be cumbersome for the patient.

Brindley (1994) carefully reviewed the experience in the first 500 patients treated with his prosthesis with a total follow-up, at that time, of 2033.5 years. Of the total, 2 patients were lost to follow-up and 21 had died. Of the deaths, two were from septicemia (one definitely unrelated to the implant and the specifics of the other unmentioned) and one from related renal failure; the causes of five were unknown. Ninety-five reoperations were required for repair, six stimulators were removed (four infected), and two were awaiting repair. In 45 patients the stimulator was believed to be intact but not used for various reasons. In all others the stimulators were in use (411 for micturition and in most for defecation and in 13 for defecation alone) and the users were believed to be “pleased.” Upper tract deterioration was reported in only 2 of 365 patients with full deafferentation and in 10 of 135 with incomplete or no deafferentation. Two of these 10 had impaired renal function, and 1 died of this condition.

van Kerrebroeck and associates (1997) reported the results of use of the Finetech-Brindley stimulator in 52 patients. These patients were selected by screening 226 patients; complete posterior sacral root rhizotomies were performed in all. Thirty-seven of the patients had 6 months of follow-up; in these patients, complete daytime continence was achieved in 73% and night-time continence in 86%. There were significant increases in bladder capacity and bladder compliance, and residual urine was reduced significantly. Complications included cerebrospinal fluid leaks, which resolved spontaneously in 23 patients; nerve damage that resolved in 1 patient; and one implant failure caused by a cable fracture, which was successfully repaired. Sauerwein and colleagues (1999) reported in abstract form the results of sacral deafferentation and implantation of an anterior sacral root stimulator in 294 patients with spinal cord injury. Bladder spasticity was relieved in all. In 50%, micturition was achieved by stimulating S4 and S5 sacral ventral roots; in the remaining cases, it was achieved by stimulating the S2 and S3 roots. Variations in surgical approaches designed to achieve stimulation of only bladder contraction are described by Dahms and associates (2000), who also summarize overall success rates for sacral ventral root stimulation in spinal cord–injured patients at about 75%. Recent laparoscopic access techniques have minimized the invasiveness of this procedure and may have a role in the future in this select group of patients (Possover, 2009).

Extradural stimulation has been used by the Tanagho group (Tanagho and Schmidt, 1988; Schmidt, 1989; Tanagho et al, 1989) in the treatment of 19 patients with serious and refractory neuropathic voiding disorders. Extensive dorsal rhizotomy was performed, and a stimulator was implanted on the ventral component of S3 or S4 with selective peripheral neurotomy (Tanagho et al, 1989). In eight patients (42%), complete success was achieved with reservoir function, continence, and low-pressure/low residual voiding with electrical stimulation. Ten patients qualified as achieving partial success, regaining reservoir function and obtaining continence.

Electrical stimulation of the ventral sacral roots with some techniques to reduce detrusor hyperactivity and obviate striated sphincter dyssynergia has become an accepted treatment modality for lower urinary tract dysfunction in spinal cord–injured patients. Although more detailed follow-up is necessary, and further evolution will doubtless occur, these techniques have achieved remarkable improvements and success rates, which now seem to have stabilized at a high level.

Transurethral Electrical Bladder Stimulation

Intravesical electrotherapy is an old technique that has been resurrected with some interesting and promising results. The use of this technique to increase bladder capacity and to increase compliance has previously been discussed. This section deals with the use of TEBS to facilitate bladder emptying by establishing conscious control of the initiation and completion of a micturition reflex. Fischer and colleagues (1993) describe their concept of the basis for this use as follows. In patients with incomplete central or peripheral nerve lesions—and only these patients are suitable for this method—at least some nerve pathways between the bladder and the cerebral centers are preserved but are too weak to be efficient under normal circumstances. TEBS in this situation is hypothesized to activate specific mechanoreceptors in the bladder wall. With depolarization of these receptors, activation of the intramural motor system is said to occur, resulting in small local muscle contractions that further depolarize the receptor cells. As soon as this local motor reaction reaches a certain strength, “vegetative afferentation” begins, meaning that stimuli travel along afferent pathways to the corresponding cerebral structures with the occurrence of sensation. This, in time, reinforces efferent pathways, and their stimuli create centrally induced and more coordinated and stronger detrusor contractions. Ebner and coworkers (1992) simply conceptualize the mechanism as involving an artificial activation of the normal micturition reflex and further suggest that repeated activation of this pathway may “upgrade” its performance during voluntary micturition.

Children with congenital neurogenic bladder dysfunction who have never experienced the urge to void require a biofeedback system to realize the nature and meaning of this new sensation induced by TEBS. This exteroceptive stimulation is also important for other groups of patients because it signals detrusor contractions and whether, and to what degree, voluntary detrusor control is or has become possible and, by demonstrating progress, serves as positive feedback.

This technique involves direct intraluminal monopolar electrical stimulation with a special catheter equipped with a stimulation electrode. Saline solution is used as the current-leading fluid medium in the bladder. Exteroceptive reinforcement is achieved by visual recording of detrusor contractions on a water manometer connected to the stimulation catheter. An intensive bladder training program has to be combined with TEBS and must be highly individualized. Only patients with an incomplete spinal cord lesion and with receptors still capable of reactivity and with a detrusor still capable of contractility will benefit from this technique. The achievement of conscious control requires, in addition, an intact cortex.

Fischer and colleagues (1993) have used this technique in patients with incomplete spinal cord injury and other incomplete central or peripheral lesions of bladder innervation, in pediatric patients with congenital neurogenic lower urinary tract dysfunction, and in patients, especially children, with non-neurogenic dysfunction voiding. Only patients with preserved pain sensation in sacral dermatomes S2 through S4 improved with this technique. The technique is time consuming because stimulation must be performed on a daily basis for weeks and months, with an individual treatment time of about 90 minutes. Kaplan and Richards (1988) reported on such therapy in myelodysplastic children, performing the treatment for 60 minutes (during a 90-minute catheterization), 3 to 5 days a week for 15 to 30 daily sessions. Of 62 patients evaluated, 42 completed at least one series of treatment. “Success” was defined differently for infants than for older children. For infants, success implied a decrease in filling pressure, an increase in the quality of bladder contraction, and a decrease in residual urine. For older children, this type of result implied a heightened awareness of detrusor contractions before and during a contraction, maintenance of low-pressure filling, effectively emptying detrusor contractions with low residual urine, and either a conscious urinary control or timely enough sensory input to allow clean intermittent catheterization for continence. Of children who initially had some detrusor contraction on initial evaluation, 80% were said to have achieved some or all of the success parameters. Of those with no initial detrusor activity, 33% achieved some success.

Other reports have been less optimistic. Lyne and Bellinger (1993) reported the results of TEBS treatment of 17 patients with neurologic dysfunction, 10 with myelomeningocele, and 2 with lipomeningocele. Ultimately, all patients showed detrusor contraction during therapy (12 did so initially), but results related to increased bladder capacity and improved continence were disappointing. Five patients showed minor positive changes in continence. After completion of therapy in 12 patients who had serial cystometry, 5 experienced an increase in capacity (14% to 158%) and 4 a decrease (7% to 37%). Decter and associates (1994) used TEBS in 25 patients with neurogenic voiding dysfunction. TEBS was correlated with an increase from 18 to 24 in the number of patients who manifested contraction on stimulation and from 3 to 12 in the number who sensed contraction during stimulation. However, cystometry showed a more than 20% increase in the age-adjusted bladder capacity in only 6 of 18 patients with serial studies and clinically significant improvements in end-filling pressures in 5 of these. A telephone questionnaire revealed that 10 of 18 patients or parents perceived an improvement in bladder function, but the authors stated that “the limited urodynamic benefits our patients achieved have not materially altered the daily voiding regimen and, because of these factors, we are not enrolling any new patients in our…program.” This technique is certainly controversial. Some question the theoretical basis and the definitions of “success” applied to patients treated. Currently, even Kaplan (2000) does not seem enthusiastic about the use of this technique to attain the goal of volitional voiding, and Decter (2000), as previously pointed out, stated that, in his opinion, TEBS is a modality with limited clinical efficacy for facilitation of filling-storage and emptying.

Sacral Neuromodulation of Emptying Disorders

Sacral neuromodulation has been successful in patients with idiopathic nonobstructive retention, in patients with retention secondary to deafferentation of the bladder after hysterectomy, and in patients with Fowler syndrome (Dasgupta et al, 2004; De Ridder et al, 2007). Patient factors predictive of success have been sought, and increasing numbers of reports are differentiating results of sacral neurostimulation in urinary retention based on the functional disorders causing the condition, detrusor acontractility, and functional outlet obstruction. Bross and coworkers (2003) evaluated the predictive ability of the carbachol test and concomitant diseases in patients with an acontractile bladder. Whereas 33% of patients had a successful bilateral percutaneous nerve evaluation, a positive carbachol test result was not predictive of success. Goh and Diokno (2007) in a retrospective study of patients implanted for nonobstructive urinary retention found a statistically significant difference in predicting success of test stimulation for patients with a preimplantation ability to void (>50 mL) versus non-voiders. Further evidence of the predictive role of detrusor function was reported by Bertapelle and colleagues (2008) in developing a detrusor contractility test using urodynamics in conjunction with sacral neurostimulation in urinary retention as an exclusion criteria for sacral neurostimulation; this test appears to be a reliable tool to rule out detrusor acontractility due to irreversible bladder myopathy or complete neurogenic lesion and to predict success of permanent sacral neuromodulation.

A large, prospective, randomized multicenter trial to evaluate the efficacy of SNS for urinary retention was performed by Jonas and colleagues (2001). After a percutaneous nerve evaluation period of 3 to 7 days, 68 patients (38% of those evaluated) with chronic urinary retention qualified for permanent implantation. Patients were randomly assigned to the treatment or control group, in which treatment was delayed for 6 months. Successful results were initially achieved in 83% of patients who received the implant, with 69% able to discontinue intermittent catheterization completely. At 18 months, 71% of patients available for follow-up had sustained improvement. These results have been corroborated by others with longer follow-up (Datta et al, 2008; White et al, 2008).

Aboseif and coworkers (2002) evaluated the efficacy and change in quality of life in patients with idiopathic, chronic, nonobstructive functional urinary retention. Thirty-two patients with idiopathic retention requiring intermittent catheterization underwent percutaneous nerve evaluation. Permanent implants were placed in 20 patients (17 women) who showed more than 50% improvement in symptoms. Eighteen patients were subsequently able to void and no longer required intermittent catheterization; one patient required bilateral SNS implants. Average voided volumes increased from 48 to 198 mL, and postvoid residual volume decreased from 315 to 60 mL. Eighteen patients reported more than 50% improvement in quality of life, although the questionnaire used in the study was not described. Significant score improvements in the Beck Depression Inventory and SF-36 after sacral root neuromodulation for retention have been demonstrated by Shaker and Hassouna (1998). Overall success rates of the percutaneous nerve evaluation range from 33.3% to 100% (Koldewijn et al, 1994; Scheepens et al, 2002; Spinelli et al, 2003). Improvement in patients with retention may not be as rapid as in patients undergoing sacral root stimulation for other reasons. A percutaneous nerve evaluation period of at least 2 to 3 weeks and a permanent implant lead evaluation of 4 weeks or more have generally been recommended. Furthermore, bilateral or caudal SNS for urinary retention may be considered initially or for unilateral SNS trial failures, although this technique is performed infrequently and is less studied to date (van Kerrebroeck et al, 2005; Maher et al, 2007; Pham et al, 2008).