International Urogynecology Journal

, Volume 14, Issue 5, pp 305–309

Sacral neuromodulation in patients with interstitial cystitis: a multicenter clinical trial


  • Kristene E. Whitmore
    • Department of UrologyGraduate Hospital
  • Christopher K. Payne
    • Department of UrologyStanford University
  • Ananias C. Diokno
    • Department of UrologyWilliam Beaumont Hospital
    • Department of UrologyGraduate Hospital
    • Colorado Gynecology and Continence Center
Original Article

DOI: 10.1007/s00192-003-1080-1

Cite this article as:
Whitmore, K.E., Payne, C.K., Diokno, A.C. et al. Int Urogynecol J (2003) 14: 305. doi:10.1007/s00192-003-1080-1


Female patients with interstitial cystitis (IC) unresponsive to standard oral and intravesical therapy were enrolled at three clinical sites for percutaneous sacral nerve root stimulation (PNS) in a prospective, observational pilot study. Evaluation was in the form of a 3-day voiding diary completed both prior to and following the commencement of sacral nerve root stimulation. Symptoms were also assessed by the O'Leary–Sant Interstitial Cystitis Symptom and Problem Indices (ICSI and ICPI). Baseline and test stimulation values for voiding diary parameters and O'Leary–Sant scores were compared to determine treatment efficacy. A total of 33 patients were enrolled. Statistically significant improvements were seen in frequency, pain, average voided volume and maximum voided volume. Significant improvements were also seen in ICSI and ICPI scores. Subacute PNS appears to be effective in reducing symptom severity and increasing voided volumes in patients with IC previously unresponsive to standard therapy.


FrequencyInterstitial cystitisPelvic painSacral neuromodulationUrgency



Interstitial cystitis


Sacral nerve root stimulation


Percutaneous sacral nerve root stimulation


Interstitial Cystitis Symptom Index


Interstitial Cystitis Problem Index


Interstitial cystitis (IC) is a clinical condition of multifactorial etiology characterized by urinary frequency, urgency, debilitating episodic bladder discomfort and chronic pelvic pain. As many as 10% of patients with IC exhibit no appreciable response to standard oral and intravesical therapy [1]. Major surgical intervention, including subtrigonal or supratrigonal cystectomy with substitution cystoplasty, has typically been employed in cases of intractable IC. Reported success rates for surgical therapy vary widely from 25% to 100%, with initial enthusiasm for such procedures having waned as many patients experience recurrent symptoms. In addition, long-term complications, including urinary retention and persistent pain, greatly limit the acceptance of major operative intervention [2, 3, 4].

Sacral nerve root stimulation (SNS) has proven experimental and clinical benefit in the treatment of urinary frequency and urgency, urge incontinence and non-obstructive urinary retention [5, 6, 7, 8, 9, 10, 11]. Although IC is characterized by frequency and urgency, the utility of SNS has not been established in the IC population, nor have patients with significant pelvic pain typically been included in clinical trials of sacral neuromodulation. The purpose of this study was to evaluate the efficacy of percutaneous sacral nerve root stimulation (PNS) in the alleviation of symptoms in patients with severe IC who would otherwise have been candidates for major surgical intervention.

Materials and methods

Between 14 April 2000 and 6 March 2001, patients with intractable IC were assessed at three clinical sites in an investigator-initiated Medtronic-sponsored prospective observational pilot study to determine the efficacy of SNS in IC symptom reduction. Entry criteria included the presence of urinary frequency, urgency and/or bladder pain, in addition to cystoscopic findings of IC at the time of bladder distension as defined by the National Institute of Arthritis, Diabetes and Digestive and Kidney Diseases [12]. Additional criteria for inclusion included the following: age greater than 18 years; willingness and competence to complete a voiding diary and questionnaires; discontinuation of intravesical dimethyl sulfoxide at least 6 weeks prior to PNS; and clinical stability on IC medications (patients must have been on an established medical regimen, with stable but persistent symptoms). Exclusion criteria included: pregnancy; severe or uncontrolled diabetes; urinary tract infection; primary stress incontinence or a preponderant stress component in the presence of mixed incontinence; prior neurolysis; and previous substitution cystoplasty. This study was approved by the Institutional Review Boards of each participating institution. All patients signed an Institutional Review Board sanctioned consent form before enrollment.

Data including age, symptom duration and previous treatments were recorded. Baseline symptoms were measured by a 3-day voiding diary, which recorded frequency, urgency and bladder pain in addition to average and maximum voided volumes. Frequency was defined as the number of voids within a 24-h period.

Urgency and bladder pain were rated on a 4-point scale (0 – none, 1- mild, 2 – moderate, 3 – severe). Symptoms were also assessed by the O'Leary–Sant Interstitial Cystitis Symptom and Problem Indices (ICSI and ICPI) prior to PNS (total maximum scores of 20 and 16, respectively) [13].

Percutaneous nerve evaluation was performed in the operating room with the patient in the prone position under monitored intravenous sedation (or in an ambulatory surgical setting employing local anesthesia only). The buttocks were taped apart to allow for adequate visualization of a bellows response, and the feet were exposed to allow observation of great toe flexion at the time of nerve root stimulation. Following appropriate sterile preparation and draping, the third sacral foramina were identified fluoroscopically employing bony landmarks as previously described. [14] The skin and subcutaneous tissue overlying the foramina were infiltrated with local anesthesia, and a 3.5-inforamen needle (Model 041828, Medtronic, Minneapolis, Minnesota) was used to isolate the third sacral nerve root (S3). Proper needle placement was confirmed fluoroscopically. Following the application of current to the foramen needle using a patient cable (Model 041831) attached to an external impulse generator (Model 3625), proper contact with S3 was confirmed through the generation of appropriate motor and sensory responses (bellows and plantar flexion of the ipsilateral great toe; throbbing or tugging sensation in the vagina, perineal body or rectum). A test stimulation wire (Model 3057) was threaded through the foramen needle and secured to the skin. In most cases bilateral wires were placed. The fourth sacral nerve roots were employed in the absence of appropriate motor and/or sensory responses at the time of S3 stimulation. Patients were discharged home with an external impulse generator attached to a single wire at an amplitude that elicited appropriate sensation without discomfort. Subacute sacral nerve root stimulation was delivered for 7–14 days, with the application of current for a minimum of 3 days to each side. Subjects completed 3-day voiding diaries for each wire tested, in addition to ICSI and ICPI during the treatment phase. Voiding diary values and O'Leary–Sant scores recorded during stimulation with the wire producing the best clinical response were used in data analysis.

Baseline and test stimulation values for frequency, average voided volume and maximum voided volume were compared using a paired t-test, with significance being set at P<0.05. Baseline and test stimulation values for urgency, bladder pain and O'Leary–Sant scores were compared using Wilcoxon's signed ranks test, with significance at P<0.05.


A total of 33 patients were included in the study [15 (45.5%) from Graduate, 10 (30.3%) from William Beaumont and 8 (24.2%) from Stanford]. Mean patient age was 44.0±15.7 years. Average symptom duration was 4.1±4.7 years. Thirty-two patients (97.0%) had current frequency or urgency. Thirty subjects (90.9%) had current bladder pain. All patients had had previous non-surgical therapy, including oral pentosan polysulfate (90%), amitriptyline (81%), hydroxyzine (65%) and intravesical dimethyl sulfoxide (59%). Three patients were not included in the statistical analysis as 1 became pregnant and did not undergo test stimulation; 1 showed symptom improvement and withdrew prior to nerve evaluation; and 1 completed the test stimulation but did not complete the voiding diaries correctly.

The baseline and test stimulation averages for each voiding diary variable are reported in Table 1. Statistically significant differences between pretreatment and treatment values were seen in frequency, bladder pain, average voided volume and maximum voided volume. No significant difference was found between baseline and test values for urgency. Statistically significant improvements were seen in ICSI and ICPI scores, with decreases from 16.4±3.0 to 10.3±5.4 and 13.8±2.4 to 8.6±5.3, respectively. Twenty-three patients (76.7%) had a subjective improvement of ≥50% in one of their major presenting symptoms, the usual threshold above which one is considered eligible for a permanent implant. Seventeen patients elected to undergo neurostimulator placement, yielding an implant rate of 51.5%.
Table 1.

Baseline and treatment values for voiding diary variables


Baseline average

Test average

P value

Frequency (voids/24 h)


19.7 ± 10.1

12.3 ± 4.8


Urgency (0none–3severe)


1.8 ± 0.8

1.6 ± 0.7


Pain (0none–3severe)


2.2 ± 0.7

1.6 ± 0.8


Average voided volume (ml)


92 ± 73

134 ± 106


Maximum voided volume (ml)


235 ± 174

320 ± 242


* Number of patients who answered question

**P <0.05

Logistic regression analyses were used to analyze whether previous treatments, specifically oral pentosan polysulfate and intravesical dimethyl sulfoxide, were associated with success of sacral neuromodulation, but no statistical associations were found.

Two adverse events were reported. One patient experienced severe pain at the time of percutaneous nerve evaluation under local anesthesia, precluding the placement of a test stimulation wire (this subject became pregnant prior a reattempt with intravenous sedation, and was one of the 3 exclusions). Another subject complained of a buzzing and an unpleasant tapping at the skin site during subacute stimulation, which resolved when the external pulse generator was turned off (included in the sample as a failure).


The bladder wall contains visceral parasympathetic afferents in the form of lightly myelinated Aδ fibers and unmyelinated C fibers, possessing predominantly mechanosensitive (tension) and chemosensitive (nociception) properties, respectively [15]. Upon bladder filling, Aδ fibers respond to physiologic low-threshold intravesical pressure, whereas C fibers are typically silent. In animal models, unmyelinated bladder afferents exhibit impulse transmission following chemical irritation, and in the presence of significant epithelial inflammation may exhibit both spontaneous activity and novel mechanosensitivity [16, 17]. Such C fiber plasticity may play a role in the evolution of IC symptoms. Physiologic bladder filling in the presence of chronic inflammation could produce an afferent barrage resulting in frequency, urgency and pain.

Sacral neuromodulation is thought to function in part through the inhibition of C fiber impulse transmission to the central nervous system. Experimental evidence in support of this mechanism has been provided by Shaker et al. [18], who administered SNS to spinalized animals with (C fiber-mediated) bladder hyperreflexia. Thirty-nine female Sprague–Dawley rats were divided equally into three groups: normal controls; spinally transected at T10; and spinally transected and electrically stimulated. Three weeks post transection, bladder hyperreflexia and C fiber activity were confirmed by urodynamics and increased L6 dorsal root ganglia levels of neuropeptide (substance P, neurokinin A, and calcitonin gene-related peptide), respectively. Electrostimulation of S1 in group 3 resulted in the absence of appreciable detrusor activity upon filling, in addition to significantly less neuropeptide within the L6 dorsal root ganglia than in group 2, consistent with decreased C fiber impulse delivery.

Inhibition of C fiber transmission by sacral neuromodulation may occur as a result of primary somatic afferent activation, as typically employed SNS impulse parameters exhibit an affinity for somatic (versus visceral) and afferent (versus efferent) nerve fibers [19, 20, 21] Somatic afferent inhibition of C fiber transmission may be explained by the 'gate theory' introduced by Melzak and Wall in 1965, in which well myelinated large-diameter (somatic) fibers impair impulse delivery by poorly myelinated small-diameter (visceral) fibers to higher centers [22].

Several investigators have recently studied the use of sacral neuromodulation in patients with IC [23, 24]. Maher et al. [23] prospectively evaluated 15 consecutive IC patients with a mean disease duration of 5.2 years who underwent a 7–10- day trial of percutaneous third nerve root stimulation. Evaluation was in the form of voiding diary, pain score (0–10) and quality of life questionnaire. Test stimulation resulted in statistically significant improvements in mean daytime frequency, nocturia, urgency, mean voided volume and mean bladder pain (decrease from 8.9 to 2.4). The quality of life variables of general health, social health and bodily pain also exhibited significant improvement. Chai et al. [24] evaluated 6 consecutive patients with IC who underwent 5 days of percutaneous third nerve root stimulation. Statistically significant improvements were reported for frequency, urgency and pain (decrease from 7.0 to 2.3 using a scale of 1–10).

The efficacy of sacral neuromodulation in IC patients having undergone permanent neuroprosthetic implantation was reported recently by Peters [25]. Twenty-two patients with a mean postoperative follow-up of 5.2 months exhibited statistically significant improvements in daytime voids, nocturia and 24-h voids. Sixty-five percent of patients reported at least a 50% improvement in pelvic pain, with 80% reporting similar improvement in pelvic pressure. In a smaller study by Feloney and Culkin [26], 6 of 8 patients with IC who underwent permanent implant were reported to exhibit a 'durable response' (mean follow-up of 18 months), and 5 (62.5%) no longer required IC medications.

The results obtained in our multicenter clinical trial are similar to those reported by the aforementioned researchers with regard to statistically significant reductions in frequency, pain and voided volumes. Our study offered the additional perioperative quantification of symptoms (frequency, nocturia, urgency and pain), and the degree to which these symptoms were problematic, through the use of the validated ICSI and ICPI, respectively. Both indices showed statistically significant improvements at the time of test stimulation. No severe adverse events were observed.

Deserving of comment are the 6 patients eligible for implant who chose not to undergo placement of a neurostimulator. These subjects made this choice of their own accord, with or without a measure of bias offered by the investigator given patient enthusiasm for the therapy and expectations with regard to clinical improvement.


Subacute sacral nerve root stimulation was shown to provide benefit in patients with interstitial cystitis, with a decrease in all symptom measures. The exact mechanism of action has yet to be defined. Our results should be tempered with the understanding that response to test stimulation does not necessarily predict implant success in IC which, in contrast to detrusor overactivity, is a disease with a fluctuating and unpredictable course and a variable response to therapy. We believe that these (and other) results do suggest that a multicenter prospective trial of chronic SNS with long-term follow-up is warranted.

Copyright information

© International Urogynecological Association 2003