Immunomodulation—A Molecular Solution to Treating Patients with Severe Bladder Pain Syndrome?

Background Patients with bladder pain syndrome experience debilitating pain and extreme frequency of urination. Numerous therapeutic approaches have been tested, but as the molecular basis of disease has remained unclear, specific therapies are not available. Objective Recently, a systematic gene deletion strategy identified interleukin-1 (IL-1) hyperactivation as a cause of severe cystitis in a murine model. Treatment with an IL-1 receptor antagonist (IL-1RA) restored health in genetically susceptible mice, linking IL-1–dependent inflammation to pain and pathology in the bladder mucosa. The study objective was to investigate whether IL-1RA treatment might be beneficial in patients with bladder pain syndrome. Design, setting, and participants Patients diagnosed with bladder pain syndrome were invited to participate and subjected to daily IL-1RA injections for 1 wk, followed by a treatment break. Patients with other urological disorders accompanied by pain were included as controls. Outcome measurements and statistical analysis When symptoms returned, treatment was resumed and responding patients were maintained on treatment long term, with individualized dosing regimens. Symptom scores were recorded and molecular effects were quantified by neuropeptide and gene expression analysis. DNA samples were subjected to exome genotyping. Results and limitations IL-1RA treatment reduced bladder pain and the frequency of urination in 13/17 patients (p < 0.001). Substance P levels in urine were lowered, and responders returned to a more normal lifestyle. Neuroinflammatory-dependent and IL-1–dependent gene networks were inhibited, as well as regulators of innate immunity. Genotyping revealed disease-associated IL1R1, NLRP3, and IL1RN DNA sequence variants in the responders. Controls did not benefit from IL-1RA treatment, except for one patent with cystitis cystica. Conclusions In this clinical study, IL-1RA treatment is proposed to reduce chronic bladder pain, immediately and in the long term. Despite the limited number of study patients, the potent acute effect and lasting symptom relief indicate that this therapeutic approach may be worth exploring in controlled clinical trials. Patient summary Treatment with an interleukin-1 (IL-1) receptor antagonist is proposed for treating bladder pain syndrome, as it can result in symptom relief and increase quality of life. Reduced neuroinflammation and IL-1 signaling provided molecular evidence of the treatment effects.


Introduction
Immune response to infection must be controlled tightly to achieve protection without collateral damage. Immune balance is easily lost, however, as host deficiencies may cause the immune defense to underperform or overachieve. Immunotherapy has the potential of restoring balance to the immune system, by compensating deficiencies or inhibiting excessive immune activation and progression to chronic disease [1][2][3].
The bladder pain syndrome (BPS) has been associated with urinary tract infection (UTI) susceptibility [4,5] and patients experience debilitating pain at bladder filling, extreme frequency of urination, and urgency of micturition, which destroy careers, social life, and sexual health [6,7]. Current therapeutic options are unfortunately insufficient. Peroral analgesics including opioids are mostly not effective, and numerous experimental intravesical treatments have shown little or no long-term effect [8]. Bladder lesions may be removed endosurgically, but in many cases, surgery has no lasting effect on bladder pain. The quality of life is severely impaired in this patient group [6,9]. The molecular basis of the BPS has remained an enigma, however, precluding the development of more specific therapies [10].
Molecular studies have identified specific innate immune checkpoints that control the severity of UTI and serve as therapeutic targets in animal models of acute pyelonephritis and acute cystitis [11][12][13]. Immunotherapy is therefore emerging as an alternative to antibiotics in UTI [11,12]. Interleukin-1 (IL-1) hyperactivation has been identified as a molecular cause of acute cystitis and bladder pathology [11], also affecting the pain receptor NK1R and its ligand substance P [11,12]. The natural IL-1 receptor antagonist (IL-1RA) inhibits the binding of IL-1 to its receptor and is used clinically as a biological immunomodulant, with an excellent safety record for indications such as gout and rheumatoid arthritis [14,15]. In proof-ofconcept studies, recombinant IL-1RA was shown to dramatically reduce bladder inflammation and the pain response in susceptible mice during acute cystitis, and to accelerate bacterial clearance [11]. This study addressed whether similar molecular mechanisms may be activated in patients with BPS and whether IL-1RA therapy might be relevant in this patient group.

2.
Patients and methods
A diagnosis of BPS was based on the diagnostic criteria and classification defined by van de Merwe et al [18] and the European Association of Urology (EAU) [10]. All patients had persistent or recurrent pain from the bladder region, combined with worsening of symptoms during bladder filling and an abnormal frequency of micturition (median symptom duration 5.8 yr, range <1->20 yr). The patients underwent cystoscopy at enrolment and were classified according to the presence or absence of Hunner's lesions [10]. The diagnostic workup was retrospective, and patients had undergone hydrodistension under general anesthesia, in connection with earlier cystoscopy or endosurgery (Table 1 and   Supplementary Table 1). Lesions had been confirmed by histology, detecting inflammation and/or detrusor mastocytosis compatible with BPS type 3C [19]. Exclusion criteria included ongoing UTI.
Seventeen patients were included in the study (seven males and ten females, median age 66 yr, range 39-93 yr, seven with BPS type 1A and ten with BPS type 3C). For a summary of patient characteristics, see Table 1, and for detailed individual information, see Supplementary

Ethics statement
Collection of samples and analysis of data were approved by the Regional to continue IL-1RA treatment or leave the study (Fig. 1A). The frequency of treatment was adjusted to clinical need of the patient (Table 2 and Supplementary Effect Predictor. ID extraction and data conversion were carried out using R [20]. The variant coordinates used as identifiers for annotation were obtained from a strand file from https://www.well.ox.ac.uk/$wrayner/ strand/index.html#Illumina (GSAMD-24v2-0_20024620_A1-b37strand.zip file). The dbSNP IDs or coordinates not present in the annotation file were extracted from the original IDs using standard pattern matching methods. To obtain dbSNP IDs (rs numbers), the coordinates were converted to dbSNP IDs using a local copy of the dbSNP database. Moreover, the reference alleles were included in the annotations using Samtools version 1.9 and in-house Python scripts [21].
The variant allele frequency of individual SNPs within these genes were calculated based on the formula: where A is the variant allele and AC is the total allele count. The AC was calculated across the samples by subtracting the number of missing genotype calls from the theoretical maximum allele count, which is two times the number of samples. The allele and genotype counts for the variants were retrieved from the genotype dataset using Tabix and inhouse Python scripts [22]. Significantly different allele and genotype counts were determined by the Fisher's exact test implemented in R for each variant.
Patients XI-XVII were subjected to limited genotyping by pyrosequencing (PSQ96) after polymerase chain reaction amplification of chromosomal DNA using the PyroMark Gold Q96 SNP reagent.
Biotinylated primers were designed to detect polymorphisms in   Table 2). Treatment effects were evaluated as changes in pain score, voiding frequency, and quality of life, using a questionnaire in Swedish and the O'Leary score (Table 2)

Statistical analysis
The data were analyzed by Prism (GraphPad version 6.01) and were considered statistically significant at p < 0.05. Clinical outcome variables were assessed by the Mann-Whitney U test and paired data using the Wilcoxon signed rank tests. Genotype data were compared using Fisher's exact test.

Participants and treatment protocol
Off-label IL-1RA treatment was offered to patients diagnosed with BPS (Tables 1 and 2). After informed consent, patients received daily subcutaneous injections of Anakinra (100 mg) for 7 d. Treatment was transiently interrupted, resumed when symptoms returned, and continued with individual treatment regimens (Fig. 1A). Therapeutic outcome variables were recorded using a symptom journal, detailing the frequency of micturition, global scoring of pain intensity, and quality of life, and confirmed by O'Leary scores. Cystoscopy at enrolment revealed signs of mucosal inflammation in ten patients (BPS type 3C; Table 1). The remaining seven patients showed no detectable bladder pathology (BPS type 1A; Table 1). Four patients with other urological disease conditions were included as controls. The controls suffered from lower pelvic pain, prostatitis, urethritis, or cystitis cystica (Supplementary Table 4).

Treatment efficacy and safety
Treatment response rates were calculated for each patient group and the two groups combined (Fig. 1B-D, Table 2, and Supplementary Table 2). The immediate subjective response of the patients with BPS was remarkable. Significant clinical improvement was observed in 14/17 patients within the first 7 d, defined by a reduction in symptom score (pain and frequency, p < 0.001) and increased quality of life (p < 0.001; Fig. 1B). One 47-yr-old patient with very severe disease since the age of 17 yr noted a relief of symptoms within 1-2 h after the first injection and reported a return to normal life (Supplementary Table 2). In the controls with lower pelvic pain, prostatitis, or urethritis, no effect was detected and treatment was discontinued after 1 wk. The patient with cystitis cystica responded and experienced long-term symptom relief. A lasting clinical response was detected in 13/17 patients, and 11 have remained on IL-1RA treatment for a median of 222 d (range 125-365 d). Individual treatment regimens have varied from daily (four patients) to weekly (five patients) or monthly (two patients) injections. One patient stopped successful treatment after 28 d due to worsened migraine attacks, and one discontinued treatment after 80 d due to a waning effect (Supplementary Table 2).
The clinical response during the 1st week of treatment was accompanied by a reduction in Substance P levels (13/ 17 patients, p < 0.001; Fig. 1B). Substance P and its receptor NK1R are essential for pain sensing in cystitis and inhibitors of NK1R have been shown to reduce bladder inflammation and pain in mice [12]. The effect on Substance P was biphasic with an initial reduction, followed by an increase when treatment was interrupted and a second decrease when therapy was resumed (Fig. 1C). Samples obtained during long-term follow-up (days 125-365) showed that the urine Substance P levels remained low compared to day 0 (Fig. 1D). In contrast, the control group with other urological disorders had low urine Substance P at enrollment, with little or no change after IL-1RA treatment, except for the patient with cystitis cystica (Supplementary Fig. 1).

Molecular validation of treatment effects
Genome-wide transcriptomic analysis was performed to further understand the molecular basis of the treatment response. RNA samples from the first ten patients were compared at onset, when the patients were in pain and after 1 wk of treatment ( Fig. 2A). Gene expression was inhibited by treatment in eight of ten patients, including canonical pathways involved in neuroinflammation, Toll-like receptor (TLR)-dependent pattern recognition, and IL-1 signaling. IL-1R1-dependent gene expression was inhibited, including IL1R1 itself, IL1RAP, CXCL1, and CXCL3 ( Fig. 2B and Supplementary Fig. 2). The patient who discontinued treatment after 80 d did not show this gene expression pattern. One patient with a clinical response showed a decrease in Substance P, but no effect on gene expression. Gene expression analysis of the combined data set identified neuroinflammatory genes as significantly inhibited by treatment. These included the alternative inflammasome constituent NAIP, TLRs 2 and 5, IRAK3, and the IL-1 receptor 1 (Fig. 3A and 3 B). Upstream regulator analysis identified IL-1b, IL-6, IL-33, and CSF3 as inhibited (Fig. 3C). In addition, the combined gene set enrichment analysis revealed a more general inhibitory effect on innate immunity, involving JAK/STAT, IFNs, and TNF gene sets ( Fig. 3D and 3 E).

Genetic analysis of BPS
In the experimental UTI model, the most severe form of acute cystitis occurs in mice, lacking Asc or Nlrp3, which are inflammasome constituents [11,23]. Excessive pro-IL-1b processing by the metalloproteinase MMP-7 drives pathology in these mice. Mice lacking IL-1b, in contrast, are protected from acute cystitis, illustrating the importance of IL-1 for symptoms and pathology [11]. Patient DNA sequence variants were identified by wholeexome genotyping. Potential susceptibility genes were selected for analysis, based on known disease associations in the murine cystitis model (IL1A, IL1B, IL1RN, IL1R1, NLRP3, PYCARD, MMP7, TAC1, and TACR1) [11,12].
Disease-associated genetic variants were identified by comparing with the 1000 genome data for the European population [21] (Fig. 4 and Supplementary Fig. 3). Fourteen significant SNPs were detected within IL1R1, NLRP3, PYCARD, MMP7, and TACR1. PYCARD encodes the inflammasome constituent ASC, identified as a cystitis susceptibility gene in mice [11]. TACR1 encodes the Substance P receptor NK1R, which drives pain sensing from the urinary bladder to dorsal root ganglia [12]. The MMP7 variant rs17098236 has been associated with gout and HIV infection but not functionally characterized [24,25]. The detected variants were intronic, suggesting potential regulatory rather than structural effects. For analyses of gene expression and DNA sequence variants in the last seven patients, please see Supplementary Figure 4.
For side effect analysis, please see Supplementary  Table 3. Acute local skin reactions were reported in five out of 17 patients. During long-term treatment, mild to moderate neutropenia or transient liver enzyme reactions were each observed in one patient. After a treatment break and normalization of the parameters, IL-1RA treatment was resumed, according to standard practice [26]. Headache after injection was reported by one patient.

Discussion
Innate immunity controls the response to infection at mucosal surfaces, where the initial contact between microbes and hosts usually takes place [27]. The damaging potential of an imbalanced response is highly significant, not least in the urinary tract, where excessive immune activation has been identified as a cause of pathology in acute pyelonephritis and acute cystitis. Immunotherapy has therefore been tested, and found to return the balance and defensive efficiency of the immune response. Small RNAbased inhibition of the IRF-7 transcription factor was effective in a murine pyelonephritis model [13], and IL-1RA therapy prevented severe cystitis and accelerated bacterial clearance in an acute cystitis model [11]. These findings provided the rationale to initiate off-label IL-1RA therapy in patients with BPS. Significant clinical improvement was detected in 76% of the patients after 1 wk and the beneficial effects were found to continue with long-term treatment, in several patients. Placebo controlled trials might therefore be undertaken to fully evaluate the therapeutic potential of this approach. Placebo effects are prevalent in patients with BPS. Decreased symptom scores have been reported in up to 50% of patients receiving placebo in controlled clinical trials [19,28]. The clinical protocol was designed with this effect in mind. Using a stop and go approach, the patients were treated for 1 wk, taken off treatment until symptoms recurred, and then offered to continue treatment. In the responders, symptoms returned when treatment was interrupted, followed by long-term symptom relief when  treatment was reintroduced in 13 patients. The long-term symptom relief further supports a treatment effect, as the placebo effect is known to decrease with time [29].

Conclusions
Data from the murine UTI model suggest that when IL-1 homeostasis is restored, a vicious cycle of neuroinflammation and pain is interrupted [12]. Molecular effects of IL-1RA treatment further indicate that this may be the case also in patients with bladder pain. The findings point to BPS as a group of conditions where immune imbalances converge on IL-1 and neuropeptide hyperactivation. A genetic element of disease susceptibility was observed, potentially affecting the balance of the IL-1 response, but this needs further study. The findings suggest that IL-1R inhibition may constitute a novel, more specific treatment approach in patients with bladder pain, with the potential to improve the quality of life. Innate immunotherapy using IL-1RA may also be relevant in recurrent cystitis, where therapeutic options are becoming increasingly limited due to antibiotic resistance.
Author contributions: Catharina Svanborg had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Critical revision of the manuscript for important intellectual content: Kinsolving, Krintel.