Anesthetic Bladder Capacity is a Clinical Biomarker for Interstitial Cystitis/Bladder Pain Syndrome Subtypes

Abstract

Objective:

To further examine anesthetic bladder capacity as a biomarker for interstitial cystitis/bladder pain syndrome (IC/BPS) patient subtypes, we evaluated demographic and clinical characteristics in a large and heterogeneous female patient cohort.

Material and Methods:

This is a retrospective review of data from women (n=257) diagnosed with IC/BPS who were undergoing therapeutic bladder hydrodistention (HOD). Assessments included medical history and physical examination, validated questionnaire scores, and anesthetic BC. Linear regression analyses were computed to model the relationship between anesthetic BC and patient demographic data, symptoms, and diagnoses. Variables exhibiting suggestive correlations (p ≤ 0.1) were candidates for a multiple linear regression analysis and were retained if significant (p ≤ 0.05).

Results:

Multiple regression analysis identified a positive correlation between BC and endometriosis (p=0.028) as well as negative correlations between BC and both ICSI score (p<0.001) and the presence of Hunner’s lesions (p<0.001). There were higher average numbers of pelvic pain syndrome (PPS) diagnoses (p=0.006) and neurologic, autoimmune, or systemic pain (NASP) diagnoses (p=0.003) in IC/BPS patients with a non-low BC, but no statistical difference in the duration of diagnosis between patients with low and non-low BC (p=0.118).

Conclusions:

These data, generated from a large IC/BPS patient cohort, provide additional evidence that higher BC correlates with higher numbers of non-bladder-centric syndromes while lower BC correlates more closely with bladder-specific pathology. Taken together, the results support the concept of clinical subgroups in IC/BPS.

INTRODUCTION

Despite the existence of national and international guidelines for the diagnosis and classification of interstitial cystitis/bladder pain syndrome (IC/BPS), the lack of a clear etiology, as well as a lack of consensus regarding terminology and clinical management¹, presents a severe impediment to the development of effective treatment strategies for patients diagnosed with this chronic and debilitating condition. While there is general agreement regarding the significant heterogeneity in IC/BPS, the mechanisms that underlie this variation continue to be the subject of intense research. Studies have examined IC/BPS patient heterogeneity by comparing symptoms,² immunological mediators,^3,4 histology,⁵ neurological pathways,³ psychological diagnoses,^6,7 treatment efficacy,⁸ and symptom triggers.⁹ Results from these studies suggest that systemic pain profiles, mental health status, bladder-specific histologic pathology, and the presence of neurological disease may help characterize divergent IC/BPS patient subgroups.

An earlier pilot study that evaluated mucosal gene expression in bladder biopsies from IC/BPS patients found a clear segregation of expression profiles based on a low (≤ 400 cc) versus a non-low (> 400 cc) anesthetic BC.¹⁰ In that study the low BC group was found to have increased expression of genes involved in inflammation and the immune response as well as decreased expression of genes important for bladder mucosal barrier integrity. This study was followed by publication of an IC/BPS patient case series that identified specific urologic and non-urologic factors that correlated with BC.¹¹ Factors that were found to be inversely correlated with BC were age, Interstitial Cystitis Problem Index (ICPI) score, and urinary frequency, while a diagnosis of depression was positively correlated with BC. These molecular and clinical data supported the framework for differing phenotypes of IC/BPS: a low BC subtype with bladder-centric disease and a non-low BC subtype with generalized pain and psychosomatic disease. This original case series of 110 patients did not evaluate the diagnosis of endometriosis or the duration of IC/BPS diagnosis. The concomitant prevalence of endometriosis and IC/BPS has been an interesting research finding in chronic pelvic pain (CPP) patients, with high co-prevalence rates reported.^12,13

The current study is an expanded case series of IC/BPS patients who received therapeutic hydrodistention (HOD). The goal of this study was to further investigate the initial correlational findings in a large patient cohort and to enable the identification of new relationships as they relate to anesthetic BC. The primary hypothesis was that the correlations that were identified in the original study would persist in this larger patient cohort and that endometriosis, a common underlying diagnosis in chronic pelvic pain (CPP) patients, would exhibit a positive correlation with BC. Secondary hypotheses were that duration of IC/BPS diagnosis would be inversely correlated with BC and that patients with non-low BC would be found to have a greater number of pelvic pain syndrome (PPS) and neurologic, autoimmune, and systemic pain (NASP) diagnoses.

METHODS

Subject Recruitment

Patients from the Urology Clinic at Wake Forest Baptist Medical Center were recruited to participate in an interstitial cystitis/bladder pain syndrome patient registry as described in Walker et al.¹¹ Females between ages 18 and 80 years old who were undergoing therapeutic HOD treatment for IC/BPS were prospectively enrolled in this study (IRB00018552). Exclusion criteria included: a history of pelvic urinary or gynecologic cancers, urethral diverticulum, spinal cord injury, stroke, Parkinson’s disease, multiple sclerosis, spina bifida, pelvic radiation, cyclophosphamide treatment, an active genital herpes outbreak, and active pregnancy. In addition to providing demographic data, all subjects included in the registry permitted access to their detailed medical history.

Patients were asked to complete validated questionnaires, including Interstitial Cystitis Symptom Index (ICSI) and ICPI¹⁴ prior to HOD. Information regarding lower urinary tract symptoms (LUTS), pelvic pain syndromes, and neurologic/autoimmune/systemic pain diagnoses was collected from the patient chart. LUTS included a history of urinary incontinence, nocturia (>2 episodes/night), and urinary frequency (>10 voids/day). PPS diagnoses included a history of vulvodynia, endometriosis, dyspareunia, and pelvic floor dysfunction. NASP diagnoses included a history of irritable bowel syndrome (IBS), a sleep disorder, chronic fatigue syndrome, fibromyalgia, Sicca/Sjogren’s syndrome, migraine headaches, depression, anxiety, allergies, and asthma. Many of the PPS and NASP diagnoses were made outside of the Wake Forest Baptist Health system and therefore many of these diagnoses are based upon patient description of their established diagnoses.

Bladder Hydrodistention

Therapeutic bladder HOD was performed according to our standard protocol with the patient under general anesthesia. Anesthesia was delivered via laryngeal mask airway or endotracheal tube intubation without muscle relaxation utilizing a bispectral index monitor to ensure complete general anesthesia. To ensure that there was not leakage around the cystoscope during the procedure, digital pressure was placed periurethrally and the bladder was filled with sterile water to a pressure of 100 cm and held for 5 minutes. Following HOD the bladder was emptied, and the anesthetic BC was recorded. The bladder was then filled with 200–300cc of sterile water and inspected for Hunner’s lesions. All surgical procedures were performed under the supervision of a single urologist (RJE).

Statistical Analysis

Clinical characteristics and demographic variables were examined for outliers and range checks were performed. A single upper outlier for BC was winsorized to 2000mL (which was at the 99.95^th percentile for the data set) to improve data modeling fit. BC was analyzed as a continuous measure in regression analyses. Age, BMI, BC, questionnaire scores, and duration of IC/BPS diagnosis were treated as continuous variables while the remaining clinical diagnoses were treated as binary variables. To test for an association of BC with demographic and clinical characteristics a linear regression model was calculated and all associations with a p-value of ≤ 0.1 were included in a multiple regression model. The multiple regression model was developed using stepwise methods (forward selection with backward elimination) with exit criteria of p<0.05. Standard linear regression diagnostics (e.g., collinearity) and examination of residuals (conditional normality) were examined to verify model fit to these data.

Duration of diagnosis data was available for a subset of patients (206) within the study and a Wilcoxon rank-sum test was performed to assess if there was a relationship between BC and duration of diagnosis. Fisher’s exact tests and Wilcoxon rank-sum tests were performed to ensure that the subset of patients with duration of diagnosis data was not dissimilar to the subset without these data (Supplementary Table 1). Wilcoxon rank-sum tests were also used to evaluate the relationship between BC and average number of PPS and NASP. For the Wilcoxon rank-sum tests the patient cohort was divided into low BC (≤400mL) and non-low BC (>400mL), as this dichotomous classification was used in our prior studies.^10,11

RESULTS

Data were collected from 295 female patients who underwent HOD for symptom management and/or to further characterize IC/BPS cystoscopic findings (e.g., presence of glomerulations, Hunner’s lesion, and anesthetic BC). Complete demographic, symptom, and diagnosis data was available for 257 patients, of which 206 (80%) had data regarding duration of IC/BPS diagnosis available (Figure 1). The demographic information for patients included in the analyses is listed in Table 1. The average age of patients was 45.70 (±14.10), average BMI was 29.10 (±8.06), and 228 (89.11%) of patients were Caucasian. The average duration of IC/BPS diagnosis was 125.00 months (±84.48), the average BC was 820.23mL (±319.07), and 28 (10.89%) of patients were found to have Hunner’s lesions on cystoscopic analysis following HOD (Table 1).

Figure 1.

Patient recruitment flowchart.

Table 1. Demographics/descriptive statistics.

Data on patient demographics, diagnoses, and HOD finding.

	Mean (SD)	Range	% (#)
Demographics
Age	45.70 (14.09)	19 – 81
BMI	29.10 (8.06)	15 – 54.9
Smoking Hx			42.41 (109)
Caucasian Race			89.11 (229)
Duration of IC/BPS Diagnosis	125.00 (85.48)	3 – 558
PPS
Endometriosis			21.01 (54)
Vulvodynia			16.34 (42)
Dyspareunia			46.30 (119)
Pelvic Floor Dysfunction			27.24 (70)
Total PPS Dx	1.11 (0.99)	0 – 4
NASP
IBS			40.47 (104)
Insomnia			12.84 (33)
Chronic Fatigue			17.90 (46)
Fibromyalgia			29.96 (77)
Sicca/Sjogren’s			3.11 (8)
Migraines			31.13 (80)
Depression			35.02 (90)
Anxiety/Panic Disorder			30.35 (78)
Allergies			82.10 (211)
Asthma			14.79 (38)
Total NASP Dx	2.98 (1.85)	0 – 8
PPS & NASP
Total PPS & NASP Dx	4.09 (2.36)	0 – 10
LUTS
Urinary Incontinence			33.85 (87)
Urinary Frequency			61.87 (159)
Nocturia			59.53 (153)
ICSI Score	13.71 (3.76)	1 – 28
ICPI Score	12.61 (3.04)	1 – 17
HOD Findings
Hunner’s Lesion			10.89 (28)
BC (mL)	820.23 (319.07)	100 – 1700
BC ≤ 400mL			13.62 (35)

Linear regression analysis showed a negative correlation between both ICSI and ICPI scores and BC (p <0.0001) (Table 2). Consistent with our previous report, there was also a negative correlation between age and BC in IC/BPS patients (p <0.001). There was a statistically significant positive correlation between endometriosis and BC (p = 0.013). Both urinary frequency (>10/day) and nocturia (>2/night) negatively correlated with BC (p = 0.002 and <0.001, respectively). The presence of Hunner’s lesions was also negatively correlated with BC (p <0.001). Smoking history and BMI, two demographic parameters not investigated in our earlier case series, were found to be positively correlated with BC with p-values that were not statistically significant but were within parameters for inclusion into the multiple regression analysis (p = 0.095 and 0.050, respectively).

Table 2. Simple linear regression analysis results.

Linear regressions assessing the impact of individual parameters with BC. Inclusion criteria for the multiple regression analysis is p < 0.1. Parameter estimate is the magnitude of the parameter’s effect on the outcome variable, BC.

	Parameter Estimate	Standard Error	p-value
Demographics
Age	−6.00	1.37	<0.001
BMI	−4.84	2.46	0.050
Smoking Hx	67.22	40.13	0.095
Caucasian Race	50.76	63.92	0.428
PPS
Endometriosis	123.84	48.33	0.011
Vulvodynia	−0.71	53.93	0.990
Dyspareunia	46.44	39.89	0.245
Pelvic Floor Dysfunction	16.86	44.78	0.707
NASP
IBS	−7.34	40.63	0.857
Insomnia	−20.61	59.6	0.730
Chronic Fatigue	32.28	51.98	0.535
Fibromyalgia	11.44	43.53	0.793
Sicca/Sjogren’s	−4.76	114.83	0.967
Migraines	45.94	42.97	0.286
Depression	51.37	41.68	0.219
Anxiety/Panic Disorder	26.17	43.34	0.547
Allergies	−72.00	51.83	0.166
Asthma	26.44	56.16	0.638
LUTS
Urinary Incontinence	1.99	42.14	0.962
Urinary Frequency	−126.87	40.28	0.002
Nocturia	−155.79	39.44	<0.001
ICSI Score	−23.09	5.11	<0.001
ICPI Score	−26.00	6.37	<0.001
HOD Findings
Hunner’s Lesion	−437.55	57.84	<0.001

Multiple regression analysis identified a statistically significant relationship between BC and endometriosis, ICSI score, and the presence of Hunner’s lesions (p = 0.028, <0.001, and <0.001, respectively). The directionality of these correlations was consistent with the individual linear regressions, with ICSI score and Hunner’s lesions negative correlation with BC while endometriosis correlated positively with BC. An analysis of the differences in age between patients with and without a history of an endometriosis diagnosis showed no significant difference (43.0 ± 11.58 v. 46.0 ± 14.77, p = 0.123). The lack of a statistical correlation of age with BC, a parameter that was found to correlate strongly in our original analysis, prompted additional analyses. When factoring the presence of just Hunner’s lesions into a linear regression of age with BC, there was a large change in the p-value of the age and BC correlation, from <0.001 to 0.157. Figure 2 depicts the relationship of Hunner’s lesions with age and BC.

Figure 2.

Presence of Hunner’s lesions according to age and BC.

Bivariate analysis showed higher average numbers of PPS diagnoses and NASP diagnoses in the non-low BC group compared to the low BC group (1.17 v. 0.71, p = 0.006; and 3.11 v. 2.14, p = 0.003, respectively) (Supplementary Table 1). The greater number of pain diagnoses correlating with the non-low BC group was additionally demonstrated when PPS and NASP diagnoses were added together (4.28 v. 2.86, p = 0.001). An analysis of the subset of the patients with duration of diagnosis data did not show a statistically significant difference for low BC versus non-low BC (p = 0.118). A comparison of the patients with and without duration of diagnosis data showed a difference in age (44.62 ±13.32 v. 50.16±16.33, p = 0.021) but no other statistically significant differences.

DISCUSSION

This retrospective case series is the largest study to date assessing correlates for anesthetic BC in IC/BPS patients and provides additional statistical power when compared to the analyses performed in our previous case series study.¹¹ The primary goal of this study was to determine if the original findings persisted in a significantly expanded patient cohort. Associations between BC and LUTS symptoms, PPS diagnoses, or NASP diagnoses can help identify pathologic, phenotypic, or diagnostic overlaps between IC/BPS and these conditions. The interest and importance of identifying such connections is exemplified by the work of the Multidisciplinary Approach to the Study of Chronic Pelvic Pain (MAPP) Research Network, a group that is committed to improving the characterizations of and treatments for urologic chronic pelvic pain syndromes (UCPPS). The MAPP Network has made important strides to elucidate and acknowledge the differing phenotypes of UCPPS into those with bladder-specific symptoms, regional pelvic pain, and systemic pain. This study provides additional support for these phenotypic associations for IC/BPS, a component of UCPPS.

In IC/BPS patients, ICSI scores and the presence of Hunner’s lesions were negatively correlated with anesthetic BC. It is logical to expect that patients with a smaller BC would generally have a higher number of lower urinary tract symptoms and therefore higher ICSI scores. The lack of a statistically significant correlation of BC with ICPI score, while contrary to the original case series results¹¹, may not be surprising. ICPI scores are based on subjective symptom bother whereas ICSI scores are based on the numerical frequency of symptoms. All IC/BPS patients, by definition, have high urinary bother indicative of the uniformly high ICPI scores, but the data from this study implies that patients with low BC have greater ICSI scores, presumably due to their physically smaller bladders.

Hunner’s lesions are a manifestation of severe bladder-centric IC/BPS pathology and have been shown to be associated with lower BC and to be more prevalent in older IC/BPS patients.¹⁵ Interestingly, age was a parameter that was found to have a strong statistical correlation with BC in the multiple regression analysis in Walker et al¹¹ but was only found to have this relationship in simple linear regression analysis in this study. It is widely accepted that urinary symptoms and malfunctions increase with age. There is conflicting data on whether there are significant decreases in BC (measured using voiding diaries and urodynamic testing) with advancing age.^16,17 The statistical strength of the correlation between age and BC seen in our simple linear regression was found to be significantly decreased (from <0.001 to 0.157) when including age with the presence of Hunner’s lesions in a multi-linear regression. This indicates that much of the relationship between BC and age in IC/BPS patient may be related to the presence of Hunner’s lesions in a subset of older IC/BPS patients. This does not, however, explain the roughly 50% of low BC (bladder-centric) IC/BPS patients in our cohort that are HL–. We believe that this group may represent an important bladder-centric IC/BPS subgroup worthy of additional study.

The finding of a higher number of PPS and NASP diagnoses in non-low BC patients supports the notion of a systemic pain IC/BPS subtype. The non-low BC in this patient subgroup is hypothesized to come from a relative lack of the fibrosis and scarring typically associated with the bladder-centric phenotype. In further support of these subgroups, a divergent gene expression profile was reported in Colaco et al¹⁰ where IC/BPS patients with non-low BC were found to have gene expression profiles similar to non-IC/BPS controls, and dissimilar to low BC IC/BPS patients.

It was somewhat surprising that having a low (≤400mL) versus non-low (>400mL) BC was not correlated with the duration of IC/BPS diagnosis, however this finding further supports the notion that IC/BPS does not develop from just one single pathological etiology whereby the extent of bladder-centric scarring and BC decrease are simply a function of time. It also suggests that disease progression may be relatively uncommon, accounting for no more than 10–15% of cases, among patients with IC/BPS. Taken together with the finding that a higher number of PPS and NASP diagnoses occur in patients with a non-low BC, a more complete picture of a separate, non-bladder-centric IC/BPS subtype emerges.

While the relationship between BC and endometriosis was not explored in the earlier case series report, there have been multiple studies acknowledging the frequent co-occurrence of endometriosis and IC/BPS^12,13,18, and the current study found a positive correlation between endometriosis and BC in IC/BPS patients. Although there are histologic and grossly visual (Hunner’s lesions) findings that are specific to some subpopulations of IC/BPS patients, the nature of IC/BPS as a syndrome with a clinical diagnosis leaves an association between endometriosis and IC/BPS vulnerable to diagnostic critique. However, the high co-prevalence between these conditions has been confirmed by Chung et al through laparoscopically visualized and/or biopsy confirmation of endometriosis and with cystoscopically visualized and/or potassium test confirmation of IC/BPS.¹⁸ In their analysis, 65% of female chronic pelvic pain patients had diagnostic confirmation of both conditions.

The nature of the general relationship between endometriosis and IC/BPS is not well understood. Endometriosis can have a significant inflammatory pathophysiologic component^19,20, and although the range of inflammatory pathophysiologies in IC/BPS are still being investigated,^21–23 it is widely acknowledged that inflammation can play a significant role for a subset of IC/BPS patients. It is possible that the relationship between these conditions may be found in their specific pelvic inflammatory processes. Taking the findings from the current study into account, the inflammatory processes in non-low BC IC/BPS may be more similar to that of endometriosis as opposed to those found in low BC IC/BPS in Colaco et al¹⁰ or in Hunner-positive patients in Jhang et al.²⁴

One of the main strengths of this study is the large number of patients (~2.5 times that of the original case series). In this larger cohort, the analysis also considered the duration of time that a patient carried a diagnosis of IC/BPS, an important factor when considering the possibility that IC/BPS can develop, over time, into bladder-centric disease with a diminished BC. Additionally, this larger cohort included a larger number (27) of patients with Hunner’s lesions in the analysis.

Limitations of the study included the retrospective nature of this analysis and vulnerability to recall bias based on the reliance on patient recall for much of the diagnosis documentation. Also, non-IC/BPS controls were not included in this analysis as a comparator group however it would likely be impossible to find a sufficiently large control group of patients without IC/BPS who would have undergone HOD.

Another limitation was the lack of confirmation of a patient’s endometriosis diagnosis, grossly or histologically by a laparoscopic evaluation, as the original diagnostic information was not available for many patients. This is relevant because some female chronic pelvic pain patients may be informed of a probable endometriosis diagnosis but never undergo a laparoscopy to confirm the diagnosis. Conversely, it is also possible that endometriosis is under-reported in the IC/BPS patient population.

Finally, it is possible that the number of HOD procedures (i.e., repeat HODs) may impact BC independently, however data presented in recently published studies from Kirk et al²⁵ and Walker et al²⁶ found that BC did not change in patients that had repeated HOD procedures. Additionally, the fact that our institution is a referral center for IC/BPS patients, we may have a skewed IC/BS patient population. Referral patients would be expected to have more advanced or non-traditional presentations of IC/BPS pathology. There is no indication that this skew has more, or less, bladder-centric versus systemic pain phenotypes of IC/BPS as evidenced by the percentage of IC/BPS patients with Hunner’s lesions being consistent with previously published studies.^27,28

CONCLUSIONS

The findings from this large cohort study add important new information to the evolving evidence-based framework used to define IC/BPS patient subgroups. Having a more detailed understanding of IC/BPS patient subgroup characteristics can inform clinical decision-making and should help to improve therapeutic management of IC/BPS patients.