Introduction

Bladder pain syndrome/interstitial cystitis (BPS/IC) is a chronic pelvic pain syndrome (> 6 months) perceived to be related to the urinary bladder accompanied by at least one other urinary symptom such as urgency or increased frequency, in the absence of urinary tract infection or other obvious pathology [1, 2]. Its primary aetiology remains unknown, and it probably has a multifactorial origin. Several mechanisms have been proposed that seem to be initially linked with a chemical, biological or autoimmune aggression:

  • Glycosaminoglycans layer defect [3,4,5];

  • Mast cell-induced inflammation of the bladder wall [6, 7];

  • Autoantibodies [8,9,10,11,12,13];

  • Infectious mechanisms: as a proper urinary tract infection or as a disturbance of the normal bladder microbiota [14,15,16,17];

  • Crossed organ sensitisation [18,19,20,21];

  • Peripheral and central nerve system sensitisation [22, 23].

Due to the difficulties underlying this condition’s diagnosis, which is done mainly by exclusion of confusable diseases, there is a huge variability in the definition of BPS/IC used in the literature, and consequently, it is difficult to establish the real incidence and prevalence of this syndrome. One of the first population studies, performed in Finland, set BPS/IC prevalence in 10.6 cases per 100,000 inhabitants with a higher rate in women (18.1 cases per 100,000 women). Men account only for 10% of the population diagnosed with BPS/IC [24]. However, several years after, another Finnish group reported higher figures, more than 200 cases per 100,000 inhabitants [25]. In the USA, epidemiological studies reported a variable prevalence. One of the first studies published [26] estimated BPS/IC global prevalence in 43,500–90,000 cases (doubling Oravisto’s Finnish results). In Spain, the first epidemiological study was published recently by members of IFU Group (Group of Investigation of Health Outcomes in Functional and Urodynamic Urology) with the collaboration of 37 Functional Urology units from all over Spain. The results obtained showed that 5.4% of the patients visiting these units were diagnosed with BPS/IC, with 9 women for each man [27]. When the diagnosis includes a cystoscopic confirmation, prevalence falls. Nevertheless, it is well known that this syndrome is heavily underdiagnosed and, for this reason, population-based surveys might be closer to the real prevalence of this entity [28].

Management of BPS/IC is complex and requires a multidisciplinary approach involving urologists, gynaecologists, physiotherapists, nurses, psychotherapists, and chronic pain specialists. Stepwise approach from less invasive to more invasive therapies must be followed. Conservative and pharmacological treatments have been already treated in this issue, but it is well known that up to 10% of BPS/IC patients are refractory to these therapies. Before the development and introduction of non-invasive methods, the options available for those patients included major surgeries such as augmentation cystoplasty; supratrigonal and subtrigonal cystectomy; urinary diversion without cystectomy; and, finally, simple cystectomy with formation of an ileal conduit. But even after the complete removal of the bladder and the urethra, pain may persist [29], a point that should be thoroughly clarified to patients before considering surgical options. Furthermore, some groups have reported that patients with BPS/IC without Hunner’s lesions, with a greater bladder capacity [30, 31] and localising the main painful point at the urethra [32] could have a worse response to reconstructive surgery. Although it is not yet approved by the FDA for this indication, sacral neuromodulation (SNM) has been used for refractory BPS/IC by multiple authors with an acceptable effectiveness, considering that in this group of patients, behavioural, oral, and intravesical treatments have previously failed.

Role of Sacral Nerve Neuromodulation in BPS/IC

Mechanisms of Action

Neuromodulation can be defined as a physiological process in which the influence of activity in one neural pathway modulates the pre-existing activity in another through synaptic interaction [33]. SNM is supposed to balance excitatory and inhibitory impulses from and to the pelvic organs at sacral and suprasacral centres through the stimulation of afferent nerves in the pelvis. The electric pulses are supposed to modulate not only the spinal cord reflexes but also brain networks. The latter has been investigated with functional magnetic resonance. In a study with overactive bladder patients, brain activity decreased in the left anterior cingulate cortex, the bilateral insula, the left dorsolateral prefrontal cortex and the bilateral orbitofrontal cortex after sacral neuromodulation [34•]. It has been also studied that changing SNM stimulus intensity alters the patterns of brain activity changes [35•].

The effect of SNM on pain disorders is usually explained by the gate theory proposed by Melzack and Wall [36]. They suggested that pain perceived to have a visceral origin, which stimulates primary afferent fibres and travels to the brain via transmission cells, could be blocked by converging impulses arising from a somatic origin (by non-nociceptive fibres at the same dermatome) that activate inhibitory interneurons located in the dorsal horn of the spinal cord. Impulses from the dorsal horn are controlled by a descending system containing fibres from the brainstem, thalamus, and limbic lobes, and thereby, SNM controls the pain sensations at the spinal segmental gate and modulates pain sensation at higher brain centres [33].

Birder and de Groat demonstrated that certain spinal areas (specially the dorsal commissure and the lateral laminae near the sacral parasympathetic nucleus) showed increased c-fos expression after both noxious (irritative intravesical stimulus) and non-noxious (bladder distension with saline) inputs, converging on the same dermatome [37]. It was also shown that a nociceptive afferent input from the pudendal nerve coming from the urethra activated cells in similar regions of the cord as does a nociceptive input from the bladder. This finding supports Ruch’s convergence theory of visceral referred pain [38]: nociceptive input from visceral and somatic structures converge onto the same central nociceptive pathways.

In 1991, Thon et al. hypothesised that SNM could reduce the perceived intensity of the pain by masking or changing it through the sensation of the electrical stimulation [39].

Clinical Findings

We can find in the literature small case series and retrospective and prospective studies showing the efficacy of SNM in patients with BPS/IC refractory to conservative treatments, including both oral and intravesical drugs. Some studies also include a heterogeneous patient population (pelvic pain syndromes and bladder pain syndromes with and without typical cystoscopic findings), making the comparison between results suboptimal. However, the evidence regarding the effectiveness of SNM in BPS/IC is increasing. Table 1 summarises the main findings of those studies focused on the use of SNM for the treatment of BPS/IC or associated chronic pelvic pain conditions.

Table 1 Summary of the main findings of those studies focused on the use of SNM for the treatment of BPS/IC or associated chronic pelvic pain conditions
Full size table

SNM for the treatment of BPS/IC first appeared in the report by Shaker et al. in 1999 on patients implanted for urgency-frequency syndrome that also complained of pelvic pain, although no specific data and figures are given [40]. Zermann et al. published the first case report on the use of sacral neuromodulation in a patient with severe pain due to interstitial cystitis, with a significant improvement 6 months after implantation [41]. At the same year, Chai et al. reported their initial experience with the percutaneous nerve evaluation (PNE) in 6 patients, showing an improvement in both objective (voiding frequency, urinary heparin-binding epidermal growth factor–like growth factor concentration, urinary antiproliferative factor activity) and subjective (pelvic pain and urinary urgency scores, patient’s impression of improvement) data [42]. One year later, Maher et al. reported their data on 15 women diagnosed with refractory IC who underwent PNE. The patients showed a significant increase of + 53 ml in the mean voided volume and also in patients’ quality of life (improvement in the Short Urinary Distress Inventory and SF-36 Health Survey). They also had a decrease in − 9 episodes of day-time micturition, − 4 nocturia episodes, and − 6.5 points in the visual analogue pain scale [43]. However, in these first reports, no follow-up information is given.

Siegel et al. published in 2001 their experience with a prospective series of 10 patients with chronic pelvic pain who underwent InterStim® implantation after successful PNE, showing a decrease in pain scores from an average of 9.7 to 4.4. At the end of the follow-up (average of 19 months), 9 patients had a decrease in the severity of the worst pain compared with baseline. Seven patients answered to a pelvic pain questionnaire mailed by the authors, and although they all would undergo again the procedure, only 6 still were using the InterStim® with a median improvement of 85%. They also reported 27 adverse events; 3 patients asked for device explantation due to lack of efficacy, another one required implantable pulse generator (IPG) removal due to wound infection with successful re-implantation, and a fifth one needed IPG relocation due to pain at IPG site [44].

Comiter published a prospective study in 2003 evaluating 25 patients with BPS/CI refractory to conservative treatments. The evaluation of results was done through voiding diary, pain score on a scale of 0–10, and the Interstitial Cystitis Symptom Index-International Cystitis Problem Index (ICSI-ICPI) questionnaires. The response rate to the test phase was 68% (40% in patients tested with PNE and 87% in patients tested with quadripolar lead insertion). With a mean follow-up of 14 months, there are significant improvements in voiding diary parameters and ICSI-ICPI scores and an average decrease in pain of 4.2 points. Loss of effect of SNM therapy occurred in one patient [45].

In the same year, Peters et al. reported their experience in 37 tested patients, from whom 26 (21 women and 15 men) underwent permanent quadripolar lead and IPG implant. They also reported a better response to test phase in patients with the quadripolar lead than with the monopolar PNE lead [46]. The same observation has been made in the study by Powell and Kreder, although long-term efficacies in both groups are similar [47].

Whitmore et al. also published a multicentre study sponsored by the manufacturer in which 33 women with BPS/IC were tested for SNM. It was a clinically homogeneous population refractory to conservative treatment with oral drugs (hydroxyzine, amitriptyline, and pentosan polysulfate) and intravesical instillations (dimethyl sulfoxide). Improvements ≥ 50% in urgency and bladder pain scales, parameters of the voiding diary, and/or ICSI-ICPI questionnaires were found in 76% of the women tested, but the device was implanted in 51.6% of the patients. No relationship was found between previous failure to oral pentosan polysulfate or intravesical dimethyl sulfoxide and patient’s response to SNM [48].

The Swiss Sacral Neuromodulation Working Group published in 2007 the results of a nationwide registry of SNM for refractory lower urinary tract dysfunction including 17 tested patients with chronic pelvic pain syndrome (11 women and 6 men). Bilateral implantation was done in two patients and unilateral in 5 patients. With a mean follow-up of 10 months, patients maintained a 65% of improvement at the last visit [49].

Regarding mid- and long-term results of SNM in BPS/IC patients, Marinkovic published in 2011 a retrospective series of 30 patients with a mean follow-up of at least 6 years. Their results showed significant improvements in voiding diary’s variables (frequency, nocturia, voided volumes) and urinary symptom scales, and an average decrease in VAS from 6.5 to 2.4. The reintervention rate not related to battery changes in this series is high (27%), but in all cases, they were related to trauma of various kinds that caused electrode migration or skin erosion at the pulse generator site [50].

Gajewski and Al-Zahrani published two studies in 2011 regarding the long-term follow-up of patients treated with SNM, but only one of them focused exclusively on patients with BPS/IC [51]. Seventy-eight patients underwent the test phase (70 PNE and 8 quadripolar leads) and 46 proceeded to the implantation of the InterStim® device (59%). Long-term success rate (average follow-up 62 months) was 72% and the average Global Response Assessment scale improvement in these patients was 80%. Reasons for device removal were poor outcome (9 patients), painful stimulation (3), and radiation of the stimulation to the leg (1 patient). Average durability of the battery was 93 months (± 25.1).

Ghazwani et al. also reported their results with SNM specifically in patients with BPS/IC. They evaluated the improvement in pain score and in the urinary distress inventory short form (UDI-6). They observed that the response for the voiding diary variables and bladder pain scale was not only evident in the 1-year follow-up, but it continued to improve thereafter with a mean follow-up of 71.5 months [52].

The most concerning problem with SNM besides reprogramming is the reoperation rate, which has been addressed in large series studying all indications. Shih et al. analysed possible predictors of reintervention in a series of 142 implanted patients, including age, sex, obesity, diabetes, chronic pain, type of urinary tract dysfunction, and use of non-TinedLead electrode versus TinedLead. A total of 55 (38.2%) patients required reoperation, and the overall explantation rate was 24.6% (35/142 patients) with an average time to device removal of 44 months (range 3–124 months). They observed an increased risk of reintervention in those patients with BPS/IC or a history of chronic pain (including fibromyalgia, chronic spinal pain, chronic pelvic pain and use of narcotic medication), although it did not reach statistical significance. However, the multivariate analysis did not show any association between any of the factors studied and the risk of reintervention [53]. In a study with 407 implanted patients and a median follow-up of 28.9 months (range 1.6–121.7) designed specifically to assess predictive factors of reintervention, Peters et al. found 134 patients (32.9%) requiring reoperation at (median) 22.9 months (25th and 75th percentiles, 8.6 and 45.1 months, respectively), including 78 patients (19%) who had their lead and/or IPG revised and 56 (14%) explantations at median 18.4 months (range 1–87.5) since implant. After excluding battery replacements, the overall reoperation rate was 24%. They found a higher proportion of reinterventions among patients with a diagnosis of BPS/IC. However, when it was analysed with other variables like independent factors (gender, body mass index, follow-up time, medical comorbidities, and presence of complications), a longer follow-up time and the presence of any complications with the procedure were the only statistically significant factors of reoperation [54].

Some groups have reported their overall experience with SNM including patients with BPS/IC. For example, Donon et al. published a series with 12 patients referring BPS/IC and 8 patients with bladder hyperactivity and pain, and they found that a 58.3% were improved, having 2 considering themselves as cured [55]. Peeters et al., in a retrospective study with 217 implanted patients for different indications (70 urgency incontinence, 34 urgency-frequency syndrome, 94 idiopathic urinary retention, 11 neurogenic bladder, 8 BPS/IC, 1 nocturnal enuresis) and with a mean follow-up of 20 months in the subgroup of BPS/IC, showed that all these patients reported a subjective improvement in pain of between 70 and 79% with no postoperative complications [56]. But not all groups have had a good experience with SNM in BPS/IC patients. For example, Elhilali et al. reported their long-term experience including two patients with BPS/IC and two patients with chronic pelvic pain, but only one of the patients with pelvic pain reported improvement [57].

Significant reductions in the use of narcotics have been reported in several studies, but it was the main focus in the study by Peters et al. in 2004. Eighteen of the 21 patients included used chronic narcotics before the InterStim®, and the other three took non-narcotic analgesics. With a mean follow-up of 15.4 months after implantation, the mean narcotic use decreased from 81.6 mg/day of morphine dose equivalents (MDE) before implantation to 52.0 mg/day (36%; P = 0.015), having four patients that ceased using all narcotics [58]. In the study by Powell and Kreder, medication use also decreased after implantation with 46.2% of patients dependent on amitriptyline stopping this medication completely, 54.5% on hydroxyzine stopping, 60% stopping pentosan polysulfate, 60.0% no longer requiring DMSO, and 20% no longer requiring narcotics postoperatively [47]. A reduced intake of different drugs has been reported in other studies [52].

Other formulas to objectively assess the effect of SNM have been explored. We have already seen that in year 2000, Chai et al. identified a seven-fold increase in urinary heparin-binding epidermal growth factor–like growth factor concentration and a decrease in the urinary antiproliferative factor activity besides a clinical response to SNM [45]. Fourteen years after, Peters et al. studied the changes on the urinary secretion of different chemokines. Before treatment, urine levels of CXCL-1, sIL-1ra, monocyte chemotactic protein-1 (MCP-1), and CCL2 positively correlated with clinical variables (pain score, urgency, Interstitial Cystitis Symptom Problem Index [ICSPI], daily voids). After 24 weeks of treatment, the chemokines’ urinary level decreased [59].

Mahran et al. [60••] have performed a systematic review independently analysing patients with BPS/IC and those with chronic pelvic pain of different aetiologies (such as chronic anal pain or chronic postsurgical pelvic pain).They reported that, although patients with pelvic pain not specifically due to BPS/IC show slightly better response to SNM, patients with pure BPS/IC also have statistically significant improvement in pain according to visual analogue scales (− 4.13 in BPS/IC versus − 5.72 in non-BPS/IC).

Wang et al. have also reviewed 17 studies including more than 500 patients with follow-ups of up to 86 months [61]. Main variables analysed were reduction of pain, reduction of the ICSI-ICPI questionnaire scores, and success rates of SNM therapy. Twelve studies evaluated pain using a visual analogue scale (VAS), with an average reduction of 3.99 points on a 0–10 scale, somewhat lower than that reported in our series (5.85). The ICSI-ICPI questionnaire scores were also significantly reduced, and success rates reported were between 60 and 98%. Complication rates ranged between 0 and 56%, with a mean explantation rate of 8%. Medium and long-term results (76% success) did not differ significantly from the short-term results (88% success), thus not confirming findings previously reported by other authors according to which the effect of SNM therapy in BPS/IC could be less durable than in other indications [62, 63].

Other implantation routes have been proven. Zabihi et al. reported the short-term follow-up (mean of 15 months) of 21 women and 9 men treated with bilateral lead implantation accessing the sacral epidural space through the sacral hiatus and placing the quadripolar tined leads in a retrograde fashion under fluoroscopy over the S2–S4 sacral nerve roots (caudal epidural SNM). Patients not only refractory to conservative management but also those who have not previously responded to S3 InterStim® implantation were included and evaluated with the O’Leary IC symptom and problem index (ICSI-ICPI), the short form of the Urogenital Distress Inventory (UDI-6), and the RAND 36-item health survey (SF-36). They reported an implantation rate of the 77%. There were four infections; three patients underwent revisions and one had the device removed. In total, five devices were explanted (four for failure and one for infection). One other patient underwent revision for device malfunction. The authors blame this high infection and reintervention rate to the learning curve [63].

Some authors have reported that other stimulation routes like posterior tibial nerve stimulation [64, 65], chronic pudendal nerve stimulation [66], or the laparoscopic implantation of neuroprosthesis to the sacral plexus [67] can be effective in the neuromodulating treatment of different chronic pelvic pain syndromes, but these are not the focus of this review.

Position of SNM in Clinical Practice Guidelines

Based on growing evidences, major international clinical practice guidelines currently consider SNM as a therapeutic alternative in patients with refractory BPS/IC before considering more invasive therapies [68]. The International Consultation on Incontinence of 2016, the guideline of the International Continence Society (ICS) and the International Consultation on Urological Diseases (ICUD) [69••], and the American Urological Association (AUA) propose to perform a test with SNM in selected patients with symptomatology refractory to oral, intravesical, and hydrodistension treatments, considering SNM within the fourth-line treatments, before considering treatment with oral cyclosporine or invasive surgeries such as augmentation cystoplasty or urinary diversions with or without cystectomy [70]. This decision should be left to the individual clinician and patient. The European Association of Urology (EAU), in its 2019 update of its Guidelines on Chronic Pelvic Pain, also recommends to offer SNM before performing more invasive interventions [71].

Our Experience with Sacral Neuromodulation for Bladder Pain Syndrome/Interstitial Cystitis

Study Population and Methods

We have retrospectively analysed our results with sacral neuromodulation in the management of patients with BPS/CI refractory to third-line treatments. After local research ethics committee approval, medical records of patients who underwent sacral root testing (stage 1 of sacral neuromodulation) and those implanted with the InterStim® I and II devices (Medtronic, Inc., Minneapolis, MN, USA) between December 1999 and January 2017, with at least 1-year follow-up, have been reviewed. Also, for the evaluation of the quality of life and overall satisfaction with SNM therapy, a telephone survey was carried out by a nurse not linked to the SNM procedure with three questions: health-related quality of life before and after the implant (“being 0 the worst health status you can imagine and 100 the best health status you can imagine”), satisfaction with the SNM procedure (“score your satisfaction with the whole sacral root neuromodulation procedure from 0 to 10), and if they would recommended SNM to a friend or relative.

Variables analysed were perception of pain on a Numerical Pain Rating Scale (NPRS) from 0 to 10 before and after treatment, subjective global response to neuromodulation using a Global Response Assessment tool of 0–100%, complications (device-related pain, infection, migration), reinterventions, device explants, and battery life. Categorical variables are described with frequencies and percentages; quantitative variables, on the other hand, are described with means, standard deviations, median, and percentiles, as appropriate.

Results

Nineteen patients with refractory BPS/IC were tested. Patients operated before year 2003 (three patients) underwent the open procedure with percutaneous nerve evaluation, and after year 2003, patients underwent implantation of the TinedLead® percutaneous electrode. Women comprised the 89.5% of patients, and average age at the test time was 53 years (range 35 to 77 years). Successful test phase was reported by 15 patients, and they underwent the InterStim® implant. However, three patients were lost from follow-up, so they were excluded from the analysis, considering that 63.15% of patients were successfully implanted and followed (12 of 19 patients). The average follow-up was 96.25 months (range 12–204 months). During this period, loss of therapeutic effect was seen in 4 patients (33% of implanted patients), between 6 and 90 months after the implant.

Approximately four out of 5 patients reported a subjective clinical improvement after the InterStim® implantation between 50 and 90% (83.4%), and the other 16.6% of patients expressed a complete resolution of symptoms (subjective clinical improvement of 90–100%). The reduction in the numerical pain scale score was statistically significant (t12 = 9.45; p < 0,001), with an average change of − 5.85 points after SNM therapy (Table 2).

Table 2 Numerical Pain Rating Scale (NPRS)
Full size table

Mean score on the health-related quality of life scale 0–100 improved significantly from 17.86 (± 16.04) before the intervention to 75.71 (± 24.9) after SNM therapy. Satisfaction with the procedure was high, with an average score of 7.71 (± 2.81) on a 0–10 scale, and 100% of implanted patients would recommend SNM to a friend or relative.

Focusing on SNM complications, 41.7% of the patients reported pain or discomfort related to the device, being resolved in most cases by reprogramming stimulation parameters except in 2 cases, which required InterStim® explant (both associated with loss of effect). We have not had any case of lead migration, but there was a lead rupture during the test phase that required major intervention for extraction. The reintervention rate was 75% (9 of 12 patients), including 7 battery replacements and both aforementioned explants, with an average battery duration of 67.17 months (± 26.38).

In our experience, 6 out of 10 patients with refractory BPS/IC can be successfully managed with SNM therapy, with a significant improvement in their quality of life.

Conclusions

Sacral nerve neuromodulation has proven its effectiveness in the treatment of patients with bladder pain syndrome/interstitial cystitis, as well as in other painful chronic pelvic syndromes. With a fairly favourable safety profile, but at the cost of a high rate of reinterventions, it is a therapeutic alternative to consider before aggressive and/or potentially mutilating surgeries in patients with BPS/IC who are refractory to conservative and pharmacological therapies.