Bladder capacity as a benchmark for patient stratification in interstitial cystitis/bladder pain syndrome.

Unsupervised clustering and PCA of the 48 whole genome expression data profiles identified three apparent patient clusters that were comprised primarily of: (1) IC/BPS with BC between 200–500 cc (n = 19), (2) IC/BPS with BC of 501–1500 cc (n = 22), and (3) controls (n = 7) ( Fig. 1 ). Although there was evidence of significance between-patient variability in the gene expression profiles within these clusters, especially among the BC 200–500 cc and control groups, the profiles from patients within the BC 501–1500 cc range were more tightly clustered. Importantly, 7 of the 8 HL+ patients had bladder capacities in the 200–500 cc range. The single remaining HL+ patient had a large bladder capacity (>800cc’s; demarcated in dark green), and is the sole large bladder capacity patient clustering with the small capacity subgroup. The samples clustering identified in the PCA plot clearly represented differences in patients’ expression data by anesthetic bladder capacity, therefore we hypothesized that including patients with a BC from 400–500 cc would more accurately describe the low BC IC/BPS subgroup. To test this hypothesis, we performed a statistical comparison of the demographic and clinical characteristics between the low BC and non-low BC IC/BPS patient sub-groups by comparing results using the former breakpoint (BC ≤ 400 cc) [ 7 ] with results using the new BC ≤ 500 cc breakpoint ( Table 1 ). In both group comparisons, low BC patients were significantly older, had higher ICSI and ICPI scores, as well as frequency and nocturia sub-scores on the PUF questionnaire (p<0.05). Patients within the low BC groups were also more likely to be male and more likely to have HL than those with a non-low BC (p<0.05). In each comparison, low BC patients also had a lower average number of NUAS in both the PPS and WPS subcategories, including lower rates of endometriosis, irritable bowel syndrome (IBS), chronic fatigue syndrome (CFS), fibromyalgia (FM), migraines, depression, panic disorder, and asthma (p500 cc comparison, low BC patients had a significantly shorter duration of IC/BPS symptoms prior to diagnosis and a greater prevalence of pelvic floor muscle dysfunction (PFMD) (p<0.05). In addition to this, allergies were no longer more prevalent in the non-low BC group, as they had been under the former ≤400 cc low BC group. A multivariable binomial logistic regression model was also created, with low BC (using the ≤500 cc breakpoint) as the outcome. Results of the model are shown in Table 2 . Older age, presence of HL, and higher ICSI score were all independent predictors of the low BC subgroup (p<0.05). Presence of migraines was associated with non-low BC (p<0.001).

All IC/BPS patients scheduled to undergo HOD (between 2014–2022) were recruited to participate in our IRB-approved prospective registry (IRB00018552). All HOD were performed according to both National Institute for Diabetes and Digestive and Kidney Diseases (NIDDK) and American Urological Association (AUA) guidelines, under general anesthesia at 100 cm sterile normal saline solution for 5 min [ 11 ]. For all patients who agreed to participate, following written informed consent, mucosal biopsies were collected at the time of HOD from the posterior bladder wall and banked at −20 °C for subsequent genomic analysis. At the time of enrollment, clinical and demographic information, including BC, HL status, scores on validated IC/BPS symptom scores (O’Leary Sant Interstitial Cystitis Symptom & Problem Indices [ICSI/ICPI], and Pain & Urgency/Frequency [PUF] patient symptom scale), were collected for each patient. In addition, common non-urological associated syndromes (NUAS), subdivided into localized pelvic pain syndromes (PPS) and neurologic, allergic, and widespread pain syndromes (WPS), were recorded via the intake questionnaire. Control patient samples were taken from women without the diagnosis of IC/BPS undergoing surgical treatment of stress urinary incontinence. Biopsies were collected under identical technique and storage parameters, obtained during cystoscopy prior to their surgical procedure. No questionnaire data or bladder capacity data were collected as they were not undergoing diagnosis nor treatment of IC/BPS. Mucosal biopsy specimens were selected from our biobank for whole genome gene expression analysis. 41 were from female IC/BPS patients; 7 from non-IC/BPS female controls. A total of 8 of these patients were HL positive. The IC/BPS samples were chosen to represent a range of BC from 200–1500 cc and processed as described in Colaco et al. [ 7 ] Briefly, total RNA was isolated using RNeasy MinElute ™ Plus ® columns and reagents (Qiagen ® ) according to manufacturer instructions. RNA quantity and purity were determined spectrophotometrically (NanoDrop ™ ND-1000) and RNA quality was determined using a bioanalyzer (Agilent Technologies, Palo Alto, California). High quality RNA (RIN≥7) was assayed on Illumina HT v4 BeadArray whole genome microarrays per standard protocols. Normalized expression data were analyzed using OMICS Explorer (Qlucore) and visualized by principal component analysis (PCA) and hierarchical clustering. Once genomic analysis was completed, the results clearly demonstrated that patients clustered above and below the 500 cc bladder capacity as a demarcating line. We then analyzed the remainder of the patients from our registry, an additional 450 patients with IC/BPS (collected between 2014–2022) that had also complete demographic, clinical, and questionnaire data. Characteristics of these 491 patients (454 females; 37 males) were compared between those who were low BC and those who were non-low BC using both the ≤400 cc (our original ‘breakpoint; n = 78) and ≤500 cc (new breakpoint suggested by the molecular data; n = 111) definition for low BC. Male IC/BPS patients were not involved in analyses that included endometriosis, dyspareunia, and vulvodynia. Independent samples t-test (continuous variables) and chi square tests (categorical variables; p ≤ 0.05 was considered significant) were performed. Variables with p≤0.1 and/or established clinical relevance were included in a backward binary logistic regression to produce an optimized model that fit the data without collinearity between variables. All statistical analysis was performed using IBM SPSS Version 27 (Armonk, NY).

In the context of IC/BPS, low anesthetic BC is a relative stratification metric. By combining newly acquired molecular data with clinical and demographic characteristics in a large cohort of IC/BPS patients, we conclude that anesthetic BC ≤ 500 cc provides a clinically meaningful biomarker for the bladder centric IC/BPS phenotypic subgroup.

Our group and others have investigated the clinical relevance of BC as a biomarker for IC/BPS patient subgroups [ 9 , 12 , 13 ]. In an earlier study, using molecular profiling of bladder biopsy tissue from a small number of IC/BPS patients and controls, we reported that gene expression profiles from patients with a BC ≤ 400 cc were similar to each other, and different from gene expression profiles from those with a BC > 400, and non-IC/BPS controls [ 7 ]. Recognizing that this was an important finding at the time, but also acknowledging that it was based on a very small sample size, we sought to reassess these findings by using a significantly larger IC/BPS patient cohort and by combining molecular findings with demographic and clinical data. The results from this larger cohort study suggest that the breakpoint to delineate low BC from non-low BC is more accurate and inclusive of patient clustering when moved from ≤400 cc to ≤500 cc. The benefit, from a patient stratification perspective, is that this shift will more accurately describe and cluster patients into their respective subgroups where their clinical characteristics more closely align, and effective treatments may differ. Moreover, from a practical standpoint, as anesthetic bladder capacity is much larger than a cystometric or functional bladder capacity, it is difficult to imagine that a 500 cc anesthetic bladder capacity in an adult would be viewed as ‘normal’. Our analysis of 491 IC/BPS patients has identified clear, statistically significant, differences between those who were low BC versus non-low BC. Low BC patients had a phenotype representative of bladder-centric disease. The results obtained in this study using the low BC breakpoint of ≤500 cc support this. For example, low BC patients had higher ICSI and ICPI scores, as well as frequency and nocturia sub-scores on the PUF questionnaire. These questionnaires have been validated in the past to evaluate bladder-centric symptoms of IC/BPS, namely urinary symptoms, and bladder pain [ 14 – 16 ]. We also found low BC patients to be older and have a greater likelihood of having HL. This aligns well with the current literature, as there are well-established relationships between bladder-centric symptoms, HL, smaller BC, and older age [ 12 , 17 , 18 ]. One unique finding present in our current analysis is that despite their older age, those with a low BC had a shorter duration of disease at diagnosis. This reflects our clinical experience that most patients do not appear to progress to a smaller bladder capacity over time, but rather often present with their disease at an older age with an already contracted bladder. Further, our molecular data suggest that low bladder capacity patients, regardless of HL status, cluster together. This finding supports grouping these patients together with the most accurate bladder capacity cutoff possible. There were some additional and compelling findings, particularly pertaining to the non-low BC patient sub-group. It has been shown that those who are non-low BC tend to display a non-bladder centric phenotype, uniquely defined by greater rates of NUAS [ 10 ]. The overall number of NUAS was significantly greater in those who were non-low BC, as well as when looking specifically at PPS and WPS. This has been affirmed by other groups. For example, Warren et al. reported on the high rate of functional somatic symptoms in IC/BPS patients, such as IBS, CFS, migraines, and depression [ 19 ], all of which were also found to be significantly more prevalent in non-low BC patients from our cohort. In their 12-year longitudinal cohort study Chang et al. reported that IBS increased the risk of developing IC/BPS [ 20 ], and we saw greater rates of IBS in non-low BC patients. Further, Overholt et al. reported higher rates of co-occurring endometriosis in non-bladder centric IC/BPS patients, who were also younger [ 21 ] – both findings consistent with this study. Interestingly, we found the non-bladder-centric, non-low BC sub-group was more likely to be composed mostly of women. Research on gender differences is lacking for IC/BPS, but Windgassen et al. conducted a study on this topic and found differences in disease experiences between men and women. One finding of relevance to the current study was the longer duration of symptoms before diagnosis in women compared to men with IC/BPS [ 22 ]. There are limitations to this study that warrant comment. First, the numbers and types of NUAS we elected to capture and document for each patient were extensive, but not exhaustive and there may be additional associations that are relevant, but which were not included in these analyses. Moreover, many of the diagnoses (e.g., FM, IBS, CFS) were not independently verified but were charted based on an affirmative response from the patient. Second, we did not have enough males represented in this cohort to make generalizations about sex-specific characteristics. Finally, the low BC breakpoint was determined based on molecular profiles from a relatively small number of patients (N = 41), however the distribution of BC ‘bins’ (N = 7; Fig. 1 ), each containing two or more patient profiles, covered a broad range that is representative of the larger patient population. The addition of demographic and clinical data, including BC, from an additional 450 patients supported this stratification.

Interstitial cystitis/bladder pain syndrome (IC/BPS) is a chronic and debilitating condition characterized by pelvic pain of unspecified origin, combined with one or more associated urologic symptoms (e.g., urgency, frequency), that impacts an estimated 3.3–7.9 million women and 2 million men in the United States [ 1 – 3 ]. This highly heterogeneous condition, with a poorly understood pathophysiology, continues to be difficult to diagnose and treat. Stratification of patients into clinically meaningful subgroups has emerged as one important strategy to mitigate the challenges associated with symptom heterogeneity. For example, it is widely recognized that patients with Hunner lesions (HL), referred to as Hunner lesion IC/BPS (HIC), represent a subgroup of IC/BPS patients with unique characteristics and response to treatment [ 4 – 6 ]. While HIC has been well documented, this sub-group represents a relatively small proportion (~10%–15%) of the total IC/BPS patient population, and it is plausible that the identification of additional subgroups within the other 80%–90% of patients could facilitate the development of more efficacious treatment strategies. One active area of exploration by our group is the association between anesthetic bladder capacity (BC), determined during therapeutic hydrodistension (HOD), and IC/BPS patient characteristics (demographic, clinical, and molecular). In an earlier study we performed whole genome gene expression analysis of bladder mucosal tissue, collected during HOD from a modest number of IC/BPS patients and controls, and identified a clear separation of patients’ molecular profiles based on BC [ 7 ]. In that pilot study, patients with BC ≤ 400 cc (which we identified as low BC) had a characteristic gene expression profile that was distinct from patients with BC >400 cc (which we identified as non-low BC), and controls [ 7 ]. Subsequent histological evidence showed bladder mucosa from patients with a low BC had more signs indicative of chronic inflammation than those with a non-low BC and provided support for this initial molecular-based stratification [ 8 ]. Continuing with BC ≤ 400 cc as the benchmark, further examination of clinical and demographic information in our large patient registry revealed a bladder-centric IC/BPS patient subgroup characterized by older age, higher symptom scores on validated questionnaires, higher prevalence of HL, and fewer non-urologic associated symptoms and syndromes (NUAS) [ 9 , 10 ]. The ‘low BC’ IC/BPS subgroup was initially based on the only molecular data available at the time, which had been derived from gene expression analysis of bladder mucosal samples from a single, modest sized (N = 16) patient cohort [ 7 ]. With the understanding that this definition of ‘low BC’ was only an approximation, the objective of the present study was to undertake an updated analysis of gene expression data and clinical characteristics from a large and heterogeneous patient cohort to reassess the parameters that define the low BC IC/BPS patient subgroup.