- Top of page
Abstract In this workshop the participants were asked to consider the role of pathology in the diagnosis and management of interstitial cystitis (IC). Currently, the NIDDK definition of IC is made on clinical criteria; bladder biopsy is not required for the clinical work-up and histology is not used as a diagnostic criterion. The literature review described the most common pathological findings to be epithelial denudation or ulceration, mononuclear inflammation, edema, congestion, hemorrhage, and mast cell activation. These pathological changes were not universal or specific with 55% of IC subjects in one study having histology that was normal and indistinguishable from control subjects.
Presentations were made which generally confirmed an association between abnormal pathology and more severe disease as determined by symptoms, cystoscopic capacity, and prognosis.
During the workshop it was clear that whether or not a biopsy was performed depended on a number of factors, for example, country of origin, the research interest of the unit, and the desire to exclude malignancy on histological grounds. The consensus, which was not unanimous, was that bladder biopsy was a-non-mandatory test in the clinical work-up of the patient with IC but that urine cytology should be performed. It was discussed that bladder biopsy be an optional test and particularly relevant when one was suspicious of other conditions and in IC research.
It was hoped that bladder biopsy pathology will give clues to contributory pathogenic processes such as epithelial dysfunction and inflammation; and in the future, with the help of molecular biology, may help determine etiology.
In this workshop the participants were asked to consider the role of pathology in the diagnosis and management of interstitial cystitis. By necessity this is restricted to the histopathological findings derived from bladder biopsies of patients diagnosed with interstitial cystitis. The literature was reviewed by the moderator. There were six other presentations during the three workshops, which then included lively discussion between international and Japanese participants.
A number of questions pertaining to the pathological findings in IC were asked at the beginning of the workshop.
How useful is the pathology report in the diagnosis of IC?
What are the histological features described in IC?
How reliable are these features (i.e. specificity for IC rather than infective, drug-induced or radiation cystitis)?
What is the reproducibility and representative validity of one biopsy for the whole urothelium?
Is there some way pathology can be used together with symptoms or other markers to improve diagnostic criteria?
Is there any role for classifying a subgroup of IC where the histology is abnormal? Which criteria should be used?
What clues do the pathological findings provide into the pathogenesis and etiology of IC?
Currently, the diagnosis of IC is made on clinical criteria and the exclusion of other causes of cystitis such as infection, radiation, or drugs.1 Bladder biopsy histology is not used as a diagnostic criterion and the performance of bladder biopsy is not required for the clinical work-up.
The histological appearance of the bladder in IC has been described by a number of authors. Messing and Stamey described non-specific submucosal edema and vasodilatation as the main finding in 19 early and 19 classic IC patients.2 Mattila described 47 IC patients where normal bladder biopsy histology was seen in 20 (43%), lymphocytic and plasma cell infiltrate in 18 (38%), and thickened telangiectatic vessels in 9 (19%).3 Gillespie et al. in their assessment of 339 patients with symptoms and in the main, cystoscopic appearance consistent with IC, found histology showed greatly decreased or absent mucous layer with mucosal ulceration. Edema and ectasia were characteristically present.4
Lynes and colleagues described the pathological findings in 22 IC and 10 control subjects and found an increase in denuded epithelium (7 IC and 0 controls) and prominent submucosal inflammation (10 IC and 1 control). They found no difference in the degree of submucosal edema and vascular ectasia with marked changes occurring in both after bladder hydrodistention.5 Johansson and Fall studied 64 subjects with classic IC, 44 with-non-ulcer IC, and 20 control subjects. The majority of the classic IC group had mucosal ulcerations (96%) and hemorrhage (86%), granulation tissue (89%) and mononuclear infiltrate (100%). Of the early IC subjects, 83% had mucosal ruptures and 89% had focal submucosal hemorrhages with minimal inflammatory infiltrate (20%).6 Denson and colleagues in 2000 analyzed biopsies from 65 females and four males with IC symptoms. They found less than 10% of specimens showed vasodilatation or submucosal edema. Moderate to severe inflammatory infiltrate was found in 29% and no overall correlation was found with cystoscopic changes. However, they found more severe cystoscopic changes along with more severe inflammatory changes in the population over 60 years.7
The most extensive histological study in IC published is the preliminary report by the IC database.8 Two hundred and eleven IC biopsies, numerous clinical features and 39 histological criteria were analyzed. Multivariate analysis was performed showing an association between nocturnal frequency and lamina propria mast cell count (tryptase stain), complete urothelial denudation, lamina propria granulation tissue, and lamina propria vascular density. Urgency and submucosal granulation tissue were associated, as was urinary pain with urothelial denudation, and submucosal hemorrhage. The IC database (ICDB) study was a very exhaustive study of which only the preliminary findings are published (i.e. the association between histology and symptoms). It is of interest but not unexpected that urothelial denudation is associated with the two most troublesome symptoms of pain and nocturnal frequency. The role of mast cells both detrusor and submucosal has been acknowledged and debated for some time. An increase in mast cells has been shown in a significant subset of IC subjects.9 The ICDB found an association between somewhat increased nocturnal frequency and lamina propria mast cell count (tryptase stain). There were no associations with the four other types of mast cell counts performed.8
The results of a histological study comparing bladder biopsies from 35 control and 34 IC subjects, six of whom were classified as severe with cystoscopic capacities < 400 mL were presented.10 Light microscopy only was assessed and six histological characteristics were graded. The bladder epithelium was described as intact, partly or completely denuded. The appearance of submucosal edema, congestion and ectasia, inflammatory infiltrate, hemorrhage, and fibrosis was graded 0 (none) to 3 (severe). Fibrosis was not evident in any of these hematoxylin and eosin stained specimens. A significant increase in epithelial denudation, submucosal edema, congestion and ectasia and inflammatory infiltrate in the IC group was found. No difference was seen in the degree of submucosal hemorrhage. Epithelial denudation may be the initial step to mucosal ulceration, the histological correlate of Hunner's ulcer or classic IC. Completely denuded epithelium never occurred in controls, in about 20% of IC but 50% of severe IC (capacity < 400 mL). Submucosal edema could result from epithelial dysfunction particularly after hydrodistention. Submucosal edema of a moderate degree (score of 2) was found in only 3% (1) of controls, 13% of early IC, and 50% of severe IC subjects. Moderate or marked submucosal congestion and/or ectasia (scores of 2 or 3) was present in only 3% (1) control subject, 10% of early IC and 50% of severe IC subjects. Submucosal inflammation may play an important role in the pathogenesis of a subset of IC patients, especially those with classic or severe disease. Moderate to severe (scores 2 and 3) submucosal inflammatory infiltrate of predominantly mononuclear cells was found in 6% (2) of controls, 13% of early IC, and 50% of severe IC subjects. Submucosal hemorrhage is most likely a result of cystoscopic biopsy trauma. Submucosal hemorrhage was found to occur in 67% of all IC, whether early or severe, and 54% of control subjects.
An abnormal composite submucosal histology score was arbitrarily defined as the sum of the individual submucosal scores and was increased in the IC subjects. Scores of 4 or greater for the composite histology score were obtained by 44% of the IC group compared with 3% (1) of the control subjects. A composite submucosal histology score of 4 or greater had a sensitivity of only 44% to detect IC (67% for severe IC) and a specificity of 97%. The positive predictive value for a composite submucosal score of 4 or greater to predict IC was 94%. Conversely, this study confirms that in more than half (55%) of IC subjects studied, the histology was normal and indistinguishable from control subjects.10
However, the severe IC subjects with cystoscopic capacity less than 400 mL had bladder biopsy pathology that showed submucosal changes in two-thirds and completely denuded epithelium in half the cases. Milder IC subjects with normal or near normal cystoscopic capacity had abnormal submucosal histology in 40% and completely denuded epithelium in 20% of the cases.
Abnormal histology of the bladder biopsy specimen can be useful in supporting a diagnosis of IC and according to the literature is useful in excluding eosinophilic cystitis and carcinoma in situ.11 These two differential diagnoses were rarely found in a combined clinical series of approximately 100 female IC subjects all of whom had a bladder biopsy where two cases of eosinophilic cystitis and no case of carcinoma in situ were diagnosed [Rosamilia, Dwyer and Scurry; unpubl. series].
It was discussed that the pattern of abnormal histology when it occurred did support a theory of pathogenesis involving epithelial dysfunction and leakage allowing submucosal edema and ectasia. In addition, there was histological evidence supporting a significant role for the inflammatory process occurring in a subset of 20–40% of IC subjects.
Also presented were the results of a further histological study that found a decrease in suburothelial vascular density in the IC group and the possibility that this may have been due to increased edema.12 In support of this, a trend to increased edema was seen in 13 IC biopsies taken after as compared with before hydrodistention, which contrasted with 13 control biopsies where no such changes occurred.
Dr Magnus Fall reported his findings supporting the differentiation of two clinical subtypes of IC. Classic IC biopsies had mucosal ulceration, hemorrhage, granulation tissue, inflammatory infiltrate, high mast cell counts, and perineural infiltrates. Biopsies from patients with non-ulcer disease had relatively normal mucosa with sparse inflammation but multiple small mucosal ruptures and suburothelial hemorrhages. He described two types of mast cell; MCCT with chymase and tryptase and MCT lacking chymase. Biopsies from classic IC subjects had a 6–10 fold increase in proteinase and mast cells (especially in the epithelium); non-ulcer IC had twice the mast cell count as controls. In classic IC mast cell growth factors were expressed in the epithelium. Classic IC could be distinguished from non-ulcer IC by epithelial mast cell recruitment and high bladder wall mast cell density.
Dr Yamada, Japan also presented briefly on the involvement of mast cells in IC.
A presentation was given by Dr Mishra, India who summarized his approach to bladder biopsy in that it was useful in detecting serious and reversible diseases in his practice. He believed that there were no light microscopic findings diagnostic of IC and that hydrodistention did not affect the mucosal or submucosal histological changes.
Dr Hanus from Prague presented his study of 84 biopsies from 112 IC patients. He reported a linear relationship between the mean bladder capacity under anesthesia and severity of glomerulations. He did not find that the most prominent and extensive histologic features were in biopsies from areas of glomerulations. There were no specific ultrastructural changes of granules of mast cells in most patients with clinical diagnosis of IC. Mast cells were confirmed in light microscopy by staining with alcian blue, however, a more specific technique was the use of antibodies against mast cell proteases. A positive reaction of mast cells with the antibody NKI C3 was seen. A higher expression of antiprotein S-100 antibody was present in neurones of the muscular and tunica propria layers in patients with a long-term history of IC. There was no relationship between the severity of clinical symptomatology and the severity of histopathologic changes observed by light or electron microscopy.
Dr Nordling of Copenhagen presented his study that correlated six different diagnostic criteria with clinical outcome in 318 patients over 32 (mean of 4) years. The criteria were pain, nocturia, petechiae or ulcer, small bladder capacity, detrusor mast cell count, and intrafascicular fibrosis. The outcome was classified ranging from 1 to 8, with 1 being symptom free after hydrodistention and 8 as requiring surgery or in severe persistent pain. There was a significant correlation between the number of positive criteria and the poor outcome. The group with mild symptoms had a significantly lower detrusor mast cell count (threefold) and percentage intrafascicular fibrosis (twofold) than the severe symptom group. Mast cell activity was thought to be related to disease activity and fibrosis a relatively late manifestation.
Lively discussion took part in each of the three workshops. Dr Theoharides explained that both mast cell number and activation particularly of epithelial mast cells was important. Also interleukin-6 played a role as a mast cell growth factor. Tryptase staining was thought to be the most useful special stain in the assessment of mast cell involvement.
All participants were asked whether they performed a biopsy; there was large variation in practice although the majority of participants took a biopsy at least some of the time. In European and Japanese centers a biopsy was usually performed; in North America a biopsy was only sometimes performed. The variation in biopsy rates may be influenced by the cost of general anesthetic, cystoscopy and biopsy in some countries so that centres where cystoscopy is performed mainly as an outpatient procedure are likely to have lower biopsy rates.
The reasons for taking a biopsy included the desire to exclude carcinoma or carcinoma in situ especially in Japan. In addition, it was recognized that the international participants had strong IC research backgrounds and this may have influenced the decision to biopsy.
There was real concern particularly among the Japanese participants that one needed to exclude carcinoma in situ. To that end the recommendation of including urine cytology to the diagnostic algorithm or work-up was agreed upon. Also discussed was the question of the unsuspected rare or unusual diagnosis. During discussion it was agreed that unsuspected CIS was seen in less than 1% of cases. Isolated unsuspected cases of eosinophilic cystitis or amyloidosis were reported and in India, tuberculosis was a differential diagnosis.
Whether or not a biopsy was performed depended on a number of factors, for example, country of origin, the research interest of the unit and the desire to exclude malignancy on histological grounds. For the general clinician, the consensus was that bladder biopsy was a non-mandatory test in the clinical work-up of the patient with IC. The main reason for this was that there exists no pathological criteria specific to IC. It was discussed that bladder biopsy ought be regarded as an optional test. There was certainly a role for bladder biopsy when one was suspicious of other conditions and in IC research.
This opinion was not unanimous however, and Dr Nordling expressed his opinion that IC should only be called IC in the presence of histological changes and his series included the assessment of features such as detrusor mast cell count and intrafascicular fibrosis. Dr Fall believed that histology was to be recommended and allowed the differentiation of two clinical subtypes of IC.
The general view was that if a biopsy were performed, it should be taken after bladder hydrodistention. The reasons for this were that initial inspection, distention to capacity, and second look were important in determining the indication for biopsy and distention following biopsy would have a greater perforation rate. There was agreement that biopsies were taken of the most diseased area. It is not known if this is ideal practice. In dermatological pathology for instance of the vulva, the adjacent area to that worst affected is ideally biopsied. There was agreement that one should collaborate with one's friendly and interested pathologist. The basic histological features to assess included the extent of epithelial denudation, and submucosal features such as congestion, edema, and inflammatory infiltrate. It was agreed that if a mast cell count was requested this would best be performed by mast cell tryptase stain. In clinical practice to date there were no other useful special stains required though they are widely used in research.
Dr Sant believed there may be a potential role for biopsy in allowing targeted treatment; for example in predicting those individuals who by virtue of having histological evidence of inflammation may respond to treatments that target inflammatory mediators such as mast cell inhibition. In the future, other targeted treatment corresponding to findings on bladder pathology may include specific antagonists to neuropeptides or growth factors. Research is needed to investigate the efficacy of this approach.
In conclusion, the role and usefulness of bladder biopsy pathology varies according to local factors such as patient population, practice variation, and research interests. There are differences even in the histological characteristics that are assessed in different units. Some of these differences are due to a lack of consensus regarding the definition of IC and because the condition is one most likely of multiple etiologies. It was agreed that a significant number of IC subjects have bladder biopsy histology that is indistinguishable from control subjects on routine light microscopy, and this fact alone limits the usefulness of bladder biopsy pathology.
During the workshop the main histological findings of IC on routine light microscopy were described although it was apparent that there are many international practice variations, for example in the analysis of mast cell involvement. In those cases where the histology is abnormal, the changes seen are non-specific although the clinical history would help differentiation from other forms of chronic cystitis. It is not clear how representative one biopsy is for the whole urothelium; the majority of biopsies are taken from the most diseased area. The IC database involved multiple biopsies and further detailed analysis may help resolve this question. Currently, there is not a role for including pathological findings as a diagnostic criterion if the aim is for the diagnosis of IC to be less restrictive than the current NIDDK definition. It is possible to speculate that IC be subclassified into those with abnormal and those with normal bladder pathology. As a general rule the group with abnormal pathology were those with more severe disease in terms of symptoms (ICDB, Fall), and prognosis (Nordling). It is hoped that long-term studies by the ICDB and other authors on disease progression in their series of patients will shed more information as to whether bladder pathology is of prognostic value. Bladder biopsy pathology will remain an important tool in IC research; giving clues to contributory pathogenic processes such as epithelial dysfunction and inflammation and in the future, with the help of molecular biology may help determine etiology.
The Kyoto International Consultation on IC was a wonderful opportunity to have a discussion toward reaching an international consensus. Future progress will be aided by input from our collaborating pathologists including those involved in the IC database study. We are looking forward to working together.