Bladder Pain Syndrome

FACULTY OF HEALTH SCIENCES UNIVERSITY OF COPENHAGEN Bladder Pain Syndrome Symptoms, quality of life, treatment intensity, clinical and pathological f...
Author: Austin Lee
28 downloads 0 Views 5MB Size
FACULTY OF HEALTH SCIENCES UNIVERSITY OF COPENHAGEN

Bladder Pain Syndrome Symptoms, quality of life, treatment intensity, clinical and pathological findings, and their correlations

PhD thesis Benedikte Richter

2010

Preface The present PhD thesis is based on work carried out during my employment at the Department of Urology at Copenhagen University Hospital Herlev from 2005 to 2010. The study took place as collaboration between the Department of Urology, the Department of Pathology, and the Laboratory of Endocrinology at Copenhagen University Hospital Herlev.

The supervisors of the PhD study were: Professor, DMSc Jørgen Nordling, Department of Urology Associate Professor, DMSc Alastair B. Hansen, Department of Pathology Copenhagen University Hospital Herlev

This thesis is based on the following papers, which will be referred to by their roman numerals:

I. Benedikte Richter, Ulrik Hesse, Jørgen Nordling. Patients with Bladder Pain Syndrome have disabling symptoms and severely impaired quality of life also after long-term follow-up. Manuscript for re-submission.

II. Benedikte Richter, Ulrik Hesse, Alastair B. Hansen, Thomas Horn, Svend O. Mortensen, Jørgen Nordling. Bladder Pain Syndrome/Interstitial Cystitis in a Danish population: a study using the 2008 criteria of the European Society for the Study of Interstitial Cystitis. British Journal of Urology International 2010;105(5):660-667.

III. Benedikte Richter, Anne Roslind, Ulrik Hesse, Jørgen Nordling, Julia S. Johansen, Thomas Horn, Alastair B. Hansen. YKL-40 and mast cells are associated with detrusor fibrosis in patients diagnosed with Bladder Pain Syndrome/Interstitial Cystitis according to the 2008 criteria of the European Society for the Study of Interstitial Cystitis. Histopathology 2010;57:371-383. The defence will take place October 8th 2010 at 14 hrs at Herlev University Hospital.

2

Table of contents Preface..................................................................................................................................................2 Table of contents..................................................................................................................................3 Acknowledgements..............................................................................................................................4 Abbreviations .......................................................................................................................................5 Summary in English.............................................................................................................................6 Summary in Danish/ Dansk Resumé ...................................................................................................8 Background ........................................................................................................................................10 Bladder Pain Syndrome .................................................................................................................10 Definition ...................................................................................................................................10 Epidemiology .............................................................................................................................12 Aetiology ....................................................................................................................................12 Mast cell involvement.................................................................................................................13 Fibrogenesis...............................................................................................................................14 Diagnosis ...................................................................................................................................15 Treatment ...................................................................................................................................16 Questionnaire tools ....................................................................................................................17 Biomarkers .................................................................................................................................18 YKL-40 ..........................................................................................................................................18 Aims of the study ...............................................................................................................................20 Materials.............................................................................................................................................21 Patients ...........................................................................................................................................21 Database .........................................................................................................................................22 Tissue collection (Paper III)...........................................................................................................22 Antibodies (Paper III) ....................................................................................................................23 Ethics..............................................................................................................................................24 Methods..............................................................................................................................................24 Primary patient evaluation .............................................................................................................24 Questionnaire study (Paper I) ........................................................................................................24 BPS questionnaire......................................................................................................................25 O’Leary Sant Symptom and Problem Index (OSPI) ..................................................................25 SF-36..........................................................................................................................................25 24 h voiding diary ......................................................................................................................25 Methodological evaluation for Paper I......................................................................................26 Descriptive study (Paper II) ...........................................................................................................26 Methodological evaluation for Paper II ....................................................................................27 Pathology study (Paper III) ............................................................................................................27 Immunohistochemistry ...............................................................................................................27 Immunoelectron microscopy analysis........................................................................................29 Enzyme-linked immuno sorbent assay analysis .........................................................................30 Methodological evaluation for Paper III ...................................................................................30 Statistics .........................................................................................................................................33 Results and discussion .......................................................................................................................34 Results from Papers I and II...........................................................................................................34 Discussion of Paper I .....................................................................................................................40 Discussion of Paper II ....................................................................................................................42 Results from Paper III ....................................................................................................................45 Discussion of Paper III...................................................................................................................49 Conclusions and perspectives ............................................................................................................51 References ..........................................................................................................................................53 Appendix A, Appendix B, Papers I-III……………………..……………………………………….70 3

Acknowledgements This research is in continuation of work initiated by our deeply respected, now departed colleague, Tage Hald, Professor DMSc, Department of Urology, Herlev Hospital. I wish to thank my main supervisor, Jørgen Nordling, for giving me the possibility of doing research, for introducing me to a network, and especially for engaging me in a disease that needs far more attention in the future. Hopefully, many urologists will follow and extend their expertise on this field which demands more than surgery. I am deeply grateful to my supervisor Alastair Hansen, for spending many pleasant and educational hours to teach me discipline, structure and strategy in medical research (almost in a German way ;)); for your invaluable support, enthusiasm and good advice in any situation; for sharing lots of chocolate, and your friendship! Another person who encouraged me from the very first day is Steinunn Jørgensen, project coordinator. Dear Steinunn, I thank you for your positive way of thinking and being! Your support gave me strength in many situations – it must be the very special Icelandic energy… I want to give a very special thank to Thomas Horn, for teaching me Pathology in med school in the most inspiring way and becoming my mentor. Thank you for always spending a minute with me when needed, for your brilliant analysis of any situation, and for your ethics and great heart! My deepest gratitude also to Julia Johansen, who spent valuable hours with me to explore YKL-40, for your support and guidance, for your admirable expertise and energy. My warm thanks to all the kind persons who helped me in many ways: Susanne Lenskjold for supporting me practically at any time, and for warm-hearted conversations; Jeanett Storgaard for always having a minute even when the phone was burning; Hanne Kruse and Bente Stærgaard for excellent expert work with the immunoelectron microscope; Elisabeth Freilow for teaching me counting mast cells; Lise Lotte Thyme for spending her time to optimize and perform double-labelling procedures; Hanife Dzaferi, Stine Jørgensen and Jens Buhl Iversen for skilful performance of IHC; Tonni Løve Hansen and Debbie Nadelmann for performing YKL-40 ELISA; staff at the Urology out-patient clinic and at the Pathology Department for always being open and including me. Professor, PhD Paul A. Price, University of California San Diego, USA, is thanked for the YKL-40 antibodies. A deep thank you to Jesper Rye Andersen and Beth Bjerregaard for providing research facilities during my employment. Financial support from the following sources is greatly appreciated: the Research Foundation at Herlev Hospital, Augustinus-Fonden, Beckett-Fonden, Direktør Jacob Madsens og hustru Olga Madsens Legat, Christian Larsen og dommer Ellen Larsens Legat, Denmark. The daily work would not have been the same if it wasn’t for my friends and collegues, the “Super 5” – thank you from the bottom of my heart, dear Lene, dear Delfina and dear Reza – for lots of talks, espresso, laughs and tears – our neverending friendship! Thank you, dear Martin, for sharing office several years, having nice talks and supporting me with your technical assistance in almost any situation. A warm thanks to Ulrik Hesse, the statistician who made me understand, that there is a reason why doctors, who did NOT spend 5 years on studying statistics, do NOT need to have the almighty insight in any statistical problem – and a P-value less than 0.05 is not the only thing worth going for…. I would never have made it so far, if it was not for my family and friends, giving me the best you can get through the toughest period of my life. Deepest thanks to my mother Tove, for always being there! To my children Freja and Jonas, the best reason to live for!

Copenhagen, September 2010 Benedikte Richter

4

Abbreviations AEC: APF: BP: BPS: BR: COPD: DAB+: DMSc: DMSO: ECM: EGF: ELISA: ESSIC: FFPE: GAG: GH: GRA: IASP: IC: ICDB: ICS: ICPI: ICSI: IHC: IL-6: kDa: mAb: MH: NGF: NIDDK: NIH: OSPI: PF: PUF: RE: RP: SF: SF-36: TGF-β: TH: UH: VAS: VEGF: VT:

amino-9-ethylcarbazole antiproliferative factor bodily pain Bladder Pain Syndrome Benedikte Richter chronic obstructive pulmonary disease 3,3’diaminobenzidine Doctor of Medical Science dimethyl sulfoxide extracellular matrix epidermal growth factor enzyme linked immuno sorbent assay European Society for the Study of Interstitial Cystitis formalin-fixed, paraffin-embedded glycosaminoglycan general health Global Response Assessment International Association for the Study of Pain interstitial cystitis Interstitial Cystitis Database International Continence Society interstitial cystitis problem index interstitial cystitis symptom index immunohistochemistry interleukin-6 kilo Dalton monoclonal antibody mental health nerve growth factor National Institute of Diabetes, Digestive and Kidney Diseases National Institute of Health O’Leary Sant Symptom and Problem Index physical functioning Pain, Urgency, Frequency Symptom Scale role limitations due to emotional health problems role limitations due to physical health problems social functioning Short Form-36 transforming growth factor-β Tage Hald Ulrik Hesse visual analogue scale vascular endothelial growth factor vitality

5

Summary in English Bladder Pain Syndrome (BPS), previously known as interstitial cystitis (IC) is a chronic, disabling disease with unknown aetiology. Patients suffer from chronic pain, pressure, or discomfort, perceived to be related to the bladder, accompanied by at least one other urinary symptom such as persistent urge to void or frequency. Clinical and pathological criteria for the diagnosis and classification of BPS have been published by the European Society for the Study of Interstitial Cystitis (ESSIC). The first part of this thesis (Paper I) comprises a questionnaire study conducted to evaluate symptom severity and quality of life, measured by the O’Leary Sant Symptom and Problem Index (OSPI), and the validated Short Form-36 (SF-36), in a Danish patient population with BPS. Furthermore, relationships between OSPI, SF-36 and clinical variables were investigated. Increasing severity of symptoms was reflected in significantly worsened quality of life, which was considerably impaired in BPS, also compared to other chronic diseases such as chronic obstructive pulmonary disease and back pain. Pain severity and treatment intensity were crucial to OSPI and SF-36 scores, other clinical variables were not. On this background retrospective clinical data from 349 consecutive patients with BPS were analysed (Paper II) in order to identify possible variables with prognostic value for outcome in terms of treatment intensity. Numbers of detrusor mast cells, detrusor fibrosis and nocturnal frequency at initial evaluation showed to have significant prognostic value for later treatment intensity. Cystoscopic findings did not have prognostic value for treatment intensity. Based on these results, a new biomarker for inflammation and fibrosis, YKL-40, was investigated in relation to detrusor mastocytosis and detrusor fibrosis (Paper III). Immunohistochemistry, immunoelectron microscopy and ELISA in 45 patients verified YKL-40 protein expression in detrusor mast cell granules and submucosal macrophages, and elevated YKL-40 levels in serum and urine compared to healthy subjects. Detrusor fibrosis was associated with numbers of YKL-40 immunoreactive cells and mast cells in the detrusor, and urine YKL-40 level. Reduction of bladder capacity was related to numbers of YKL-40 immunoreactive cells and mast cells in the detrusor. The ESSIC criteria were used for diagnosis and classification in all three studies. Patients with BPS type C (histopathologically changed biopsies) required more intensive treatment than those with types A and B, but all groups were equal regarding symptom severity and quality of life. In conclusion, BPS is a severe chronic disease causing substantial deterioration in quality of life. Symptom severity did not predict clinical findings, and vice versa. Numbers of detrusor mast cells, detrusor fibrosis and nocturnal frequency were associated with failure of standard urological

6

therapy, whereas cystoscopic findings were not. The biomarker YKL-40 was expressed in detrusor mast cells. YKL-40 immunoreactive cells and mast cells in the detrusor were associated with detrusor fibrosis and reduced bladder capacity, reflecting a relationship between histopathological, presumably inflammatory changes in bladder biopsies, and bladder shrinkage. YKL-40 concentration was elevated in serum and urine in patients with BPS compared to healthy subjects and may be used as non-invasive biomarker for evaluation of bladder fibrogenesis, which needs to be explored.

7

Summary in Danish/ Dansk Resumé Smertefuld blæresygdom (BPS), tidligere kendt som interstitiel cystitis (IC), er en kronisk invaliderende sygdom med ukendt ætiologi. Patienterne lider af kroniske smerter, trykken eller ubehag som føles som kommende fra blæren, ledsaget af mindst et andet urinvejssymptom som vedvarende vandladningstrang eller hyppig vandladning. Kliniske og patologiske kriterier for diagnose og klassifikation af BPS er offentliggjort af det Europæiske Selskab for Studier af Interstitiel Cystitis (ESSIC). Den første del af PhD projektet (Artikel I) omhandler en spørgeskema-undersøgelse udført for at vurdere symptomernes sværhedsgrad og livskvaliteten, målt ved hjælp af O’Leary Sant Symptom og Problem Indeks (OSPI), og det validerede Short Form-36 (SF-36) i en dansk patientpopulation med BPS. Endvidere blev sammenhænge mellem OSPI, SF-36 og kliniske variable undersøgt. Tiltagende sværhedsgrad af symptomer afspejlede sig i signifikant forværret livskvalitet, som var betydelig nedsat ved BPS, også sammenlignet med andre kroniske sygdomme som kronisk obstruktiv lungesygdom og rygsmerter. Smerteintensiteten og behandlingsintensiteten var afgørende for OSPI og SF-36 scores, andre kliniske variable var ikke. På denne baggrund blev retrospektive kliniske data fra 349 konsekutive patienter med BPS analyseret (Artikel II) for at identificere mulige variable af prognostisk værdi for sygdomsudfaldet i form af behandlingsintensitet. Detrusor mastcelle antal, detrusor fibrose og natlig vandladning ved primær evaluering viste sig at have signifikant prognostisk værdi for efterfølgende behandlingsintensitet, mens fund ved kikkert-undersøgelse ikke havde. Baseret på de foreliggende resultater blev en ny biomarkør for inflammation og fibrose, YKL-40, undersøgt i forhold til detrusor mastocytose og detrusor fibrose (Artikel III). Immunhistokemi, immunelektronmikroskopi og ELISA hos 45 patienter påviste YKL-40 ekspression i detrusor mastcelle granula og submukøse makrofager, samt forhøjede YKL-40 koncentrationer i serum og urin sammenlignet med raske personer. Detrusor fibrose havde sammenhæng med antal af YKL-40 immunreaktive celler og mastceller i detrusor, og urin YKL-40 koncentration. Reduceret blærekapacitet var relateret til antal af YKL-40 immunreaktive celler og mastceller i detrusor. ESSIC kriterierne blev benyttet til diagnose og klassifikation i alle tre studier. Patienter med BPS type C (histopatologisk forandrede biopsier) blev behandlet mere intensivt end dem med type A og B, mens alle grupper var ens med hensyn til symptomernes sværhedsgrad og livskvalitet. Sammenfattende fandtes, at BPS er en alvorlig kronisk sygdom, medførende betydelig nedsat livskvalitet. Kliniske fund kunne ikke forudsiges ud fra symptomernes sværhedsgrad, og vice versa. Detrusor mastcelle antal, detrusor fibrose og natlig vandladning havde sammenhæng med svigt af urologisk standardbehandling, cystoskopi-fund havde ikke. YKL-40 var udtrykt i detrusor mast8

celler. Antallet af mastceller, ligesom samlet antal af YKL-40 immunreaktive celler i detrusor var relateret til detrusor fibrose og formindsket blærekapacitet, tydende på en sammenhæng mellem formentlig inflammatorisk betingede histopatologiske forandringer i blærebiopsier og blæreskrumpning. YKL-40 koncentrationen var forhøjet i serum og urin hos patienter med BPS sammenlignet med raske personer og kan muligvis anvendes som non-invasiv biomarkør til vurdering af bindevævsdannelsen i blæren, hvilket må undersøges nærmere.

9

Background Bladder Pain Syndrome Definition Bladder Pain Syndrome (BPS) was previously known as interstitial cystitis (IC) and is a chronic, disabling disease. The term “interstitial cystitis” refers to a disease with inflammatory changes in deeper layers of the bladder wall. Inflammation is seen in the submucosa but not in the detrusor (1-3), opposite eosinophilic cystitis, where eosinophils in the muscle layer are mandatory for the diagnosis (4;5). IC has never been defined but has been diagnosed on the basis of bladder pain, pressure or discomfort, urgency and urinary frequency, cystoscopic and morphological findings, and exclusion of other causes of the symptoms (6-8). The term for the disease has been changed several times, from IC to IC/painful bladder syndrome, and to BPS/IC. In this thesis, it is referred to as BPS (6;8-10). In the late 1980s the National Institute of Diabetes, Digestive and Kidney Diseases (NIDDK, USA) presented a set of consensus criteria, revised in 1990, which were meant to ensure comparable patient populations for research, explicitly not to define the disease (11): Pain associated with the bladder or urinary urgency, and glomerulations or Hunner’s ulcer on cystoscopy under anaesthesia in patients with nine months or more of symptoms, at least eight voids per day, one void per night, and cystometric bladder capacity less than 350 ml. The massive lack of a clinical definition for IC resulted in the diagnostic use of the NIDDK criteria by clinicians (12). Since the NIDDK criteria were too restrictive, patients with less severe symptoms were excluded from being diagnosed with IC, thereby not receiving necessary attention, health care and insurance reimbursement (12). Another critical point of the NIDDK criteria was the inclusion of patients with urgency alone, patients who later were taken into the definition of overactive bladder syndrome by the International Continence Society (ICS) (13). At the same time the ICS introduced ‘painful bladder syndrome’ including IC, defined by suprapubic pain related to bladder filling (13). However, it was shown that only two-thirds of patients presumably having IC fulfilled this definition (14). The lack of a proper definition had led to the use of various algorithms worldwide for diagnosing IC, resulting in studies on patient populations with different characteristics (15-25). In 2008, the European Society for the Study of Interstitial Cystitis (ESSIC) proposed the change of term to bladder pain syndrome, since this would fit in well with the taxonomy of the International Association for the Study of Pain (IASP) regarding the pain syndrome complex, and supported by the growing evidence for the increased incidence of co-morbid pain conditions (e.g. fibromyalgia, chronic fatigue syndrome, irritable bowel syndrome) in patients with BPS (9;26-28). With the 10

ESSIC proposal diagnostic criteria and a classification for BPS were provided (Figure 1) (8). The ESSIC agreed that BPS would be diagnosed on the basis of chronic (>6 months) pelvic pain, pressure, or discomfort perceived to be related to the urinary bladder accompanied by at least one other urinary symptom such as persistent urge to void or frequency (8). Confusable diseases as the cause of the symptoms had to be excluded. The idea was to establish a set of diagnostic criteria, allowing subclassification with the possibility for subgrouping patients with either different prognosis and/or different response to treatment. The ESSIC proposal resulted in controversies about term definitions and the necessity to change the disease name omitting IC, but obviously a general agreement is about to be reached (10). Future research will have to focus on validation of the ESSIC criteria internationally to evaluate their utility in different clinical settings.

Figure 1. ESSIC classification of types of Bladder Pain Syndrome on the basis of findings at cystoscopy with hydrodistension and of biopsies.

Biopsy

Cystoscopy with hydrodistension 1

Not done

Normal

Not done

XX

1X

2X

3X

Normal

XA

1A

2A

3A

Inconclusive

XB

1B

2B

3B

XC

1C

2C

3C

3

Positive

1 2 3

Glomerulations Hunner’s lesion

2

cystoscopy: glomerulations grade 2-3 with or without glomerulations histology showing inflammatory infiltrates and/or detrusor mastocytosis and/or granulation tissue and/or intra-fascicular fibrosis

From van de Merwe et al. Diagnostic criteria, classification, and nomenclature for painful bladder syndrome /interstitial cystitis: an ESSIC proposal. Eur Urol 2008;53:60-67.

11

Traditionally, BPS is divided into two entities, ulcer (classic) and non-ulcer BPS, with about 90% having the non-ulcer form (29;30). Patients with ulcer BPS present cystoscopically with a mucosa lesion in the bladder wall called Hunner’s ulcer. This represents rather a distinctive inflammatory lesion than a chronic ulcer, being a characteristic deep rupture through the mucosa and submucosa provoked by bladder distension (8). The ESSIC accepted to replace the term Hunner’s ulcer by Hunner’s lesion which will be referred to in this thesis, with following definition by Magnus Fall: ‘‘The Hunner’s lesion typically presents as a circumscript, reddened mucosal area with small vessels radiating towards a central scar, with a fibrin deposit or coagulum attached to this area. This site ruptures with increasing bladder distension, with petecchial oozing of blood from the lesion and the mucosal margins in a waterfall manner. A rather typical, slightly bullous edema develops postdistension with varying peripheral extension (8).” Patients with ulcer BPS have a smaller bladder capacity under general anaesthesia (31), and higher levels of intravesical luminal nitric oxide (32) compared to those with non-ulcer BPS. Surgical treatment of Hunner’s lesions may improve bladder pain dramatically, but lesions may recur (33;34). Patients with non-ulcer BPS tend to be younger at diagnosis and at symptom onset than those with classic BPS (35), and develop multiple glomerulations and superficial mucosal bleeding at cystoscopy and hydrodistension (31). These findings support the assumption of two different entities of BPS (8;29-37), which needs to be explored further. Epidemiology Due to the inconsistency regarding diagnostic criteria and methods, and to the low incidence of BPS, epidemiological studies have been difficult to conduct (6;8-10;38). This resulted in widespread prevalence estimations ranging from 1.2 per 100.000 individuals and 4.5 per 100.000 females in Japan (19), to a questionnaire based study that suggested a figure in American women of 20.000 per 100.000 (39). A female preponderance (5-10 times more) has been demonstrated, the reason for which is still unknown (38;40). Aetiology The aetiology of BPS is still unknown and may consist of multiple coexisting and reinforcing mechanisms (9;10;41;42). Several theories have been put forward through the years, including inflammation (1-3), urothelial dysfunction and glycosaminoglycan (GAG) layer defects (43;44), inhibition of urothelial bladder cell proliferation (45), autoimmune mechanisms (46), infection (4749), pelvic neurogenic cross-talk (50;51), nitric oxide metabolism (32;52), hypoxia (53) and genetic predisposition (54), some of which are described briefly in Appendix A. Mast cell involvement and

12

fibrosis in BPS have been debated for years (55;56) and were in focus in the present thesis; both issues are outlined more detailed below. Despite previous efforts, no significant conclusive advances have been made in understanding the aetiology of BPS (10;41;57). It is today recommended to consider BPS as a part of a generalized somatic disorder (10). Mast cell involvement Mast cells are multifunctional immune cells that develop from a specific bone marrow progenitor cell (58). They play a key role in inflammatory and allergic processes and carry a variety of mediators in granules (e.g. histamine, heparin, tryptase, chymase, transforming growth factor-β (TGF-β), vascular endothelial growth factor (VEGF)) (58-61). Vasoactive and inflammatory mediators secreted by mast cells may explain various symptoms seen in BPS; as an example, tryptase causes microvascular leakage and stimulation of protease-activated receptors, causing inflammation and neuronal hyperexcitability (62;63). VEGF is vasodilatory, and its overexpression in the bladder in BPS may contribute to hypervascularity and glomerulations (64). Histamine is supposed to mediate pain as seen in BPS (65). A recent animal study in irritable bowel syndrome showed, that mucosal mast cell mediators excited nociceptive visceral sensory nerves (66), which may parallel the mechanisms in BPS. Presumably mast cells are triggered to infiltrate the bladder by non-immunologic stimuli, such as microbial infection, chemicals, neuropeptides (58). Animal studies supported the hypothesis that released mast cell mediators may elicit a neuro-inflammatory response in bladder tissue by increasing epithelial permeability, resulting in an influx of potassium ions and leading to sensory nerve up-regulation (67-69). Bladder inflammation triggers nerve growth factor (NGF) increase (70;71) and enlargement of peripheral neurons that innervate the bladder (72). This supposedly leads to a “vicious circle” with neurogenic cross-organ sensitization and further mast cell attraction. The hypothesis of a neuro-inflammatory aetiology has lately been supported by basic research (67;73;74). Since mast cells are capable to synthesize the pro-fibrotic cytokine TGF-β and are present in the bladder detrusor, they may play a role in detrusor fibrogenesis resulting in chronicity seen in some patients with BPS and comparable to other diseases (2;8;55;58-61;73;75). In Crohn disease mast cells showed to have influence on proliferation, collagen production and contractile activity of fibroblasts in vitro, but due to diverse findings the exact mechanisms were not identified (76). It was shown that mast cell chymase treatment of bladder fibroblasts in vitro resulted in dosedependent, significant increases in collagen types I and III (77). Collagen types I and III and high counts of mast cells were demonstrated in the detrusor in myelodysplastic bladder tissue (77), supporting mast cell involvement in bladder fibrogenesis. 13

There have been numerous reports on the presence of mast cells in bladder biopsies with contradictory results (1;3;75;78-80). Several studies reported a significant increase in numbers of detrusor mast cells in biopsies of patients with BPS compared to controls (75;78-80), whereas others found no difference (1;3). The Interstitial Cystitis Database (ICDB) study group reported a statistically significant association between numbers of mast cells in the lamina propria and night-time voiding frequency (81). Despite these discrepancies, mast cell quantification in detrusor biopsies is recommended by the ESSIC and performed at many centres (8;80;82). Fibrogenesis Some patients with BPS develop bladder shrinkage, and histological examination may show fibrosis in the bladder detrusor (83). Fibrosis is caused by pathophysiologically increased accumulation of extracellular matrix (ECM), consisting of collagen, proteoglycans, elastine, fibronectin and other proteins (84;85). The function of the ECM is to support the three-dimensional structure of tissues, and to connect cells and the extracellular space (85-87). In a dynamic flow a constant remodelling and destruction of matrix proteins is seen, partly dependent on surrounding factors and tissue damage (88;89). Mean turn-over rate of the ECM is 80-120 days (86;90). Damage to tissues can result from various stimuli, including infections, autoimmune reactions, toxins, radiation and mechanical injury (88;89). A regenerative repair process is initiated, mediated by pro-inflammatory cytokines, recruited inflammatory cells such as macrophages, neutrophils, monocytes and mast cells (91), producing growth factors and pro-fibrotic cytokines such as TGF-β, and activation of collagen-producing cells such as myofibroblasts and fibroblasts, which are supposed to be the key cellular mediators in fibrosis and also have been identified in the bladder (77;85;91-93). In a normal healing process injured cells are replaced with cells of the same type; under certain circumstances connective tissue replaces normal parenchymal tissue, forming a scar. Fibrosis occurs when the synthesis of new collagen exceeds the rate at which it is degraded, such that the total amount of collagen increases over time (88;89). Apart from secretion of new ECM, myofibroblasts contribute to reconstruction of injured tissue by exerting high contractile force (9496). Deregulation of this mechanism results in tissue contracture, which may take place in some patients with BPS (56;77;94). Contrary to this hypothesis, classic myofibroblast characteristics were not verified in fibroblastic cells from neuropathic human detrusor (92), which may be explained by the possible existence of fibroblasts with an atypical myofibroblast phenotype (92). Ultimately, extensive tissue remodelling due to sustained tissue damage and fibrosis, and contracture can lead to organ failure and death, which is seen in diseases such as idiopathic pulmonary fibrosis, liver cirrhosis, cardiovascular fibrosis, systemic sclerosis and nephritis (86-89;91;97-100).

14

Diagnosis The diagnosis of BPS is often difficult, since the symptoms are identical in other clinical conditions, such as urothelial carcinoma and carcinoma in situ, which have to be excluded (8). According to ESSIC, BPS is diagnosed on the basis of chronic (>6 months) pelvic pain, pressure or discomfort perceived to be related to the urinary bladder accompanied by at least one other urinary symptom like persistent urge to void or frequency (8;82). Apart from a complete medical history with focus on co-morbidity, previous pelvic diseases and operations, a clinical examination of the patient has to be performed, including inspection and palpation of the lower abdomen and the pelvic area, vaginal and rectal exploration (8). Urine dipstick and cytology have to be done to exclude urinary tract infection, urothelial carcinoma or other lesions. Basic information on voiding patterns is provided by urinary diaries completed by the patients, which is essential for making the diagnosis. Cystoscopy is performed to rule out other conditions but also with a diagnostic intention. The bladder is inspected under general anaesthesia with hydrodistension, a procedure which has advantageous effects on some patients’ symptoms, though probably short-lived (82;101). The bladder is distended at an infusion height of 80 cm above the symphysis pubis for three minutes, and the bladder mucosa is inspected for bleeding (glomerulations), Hunner’s lesion, and other pathological findings such as carcinoma, before and after this procedure (82). The cystoscopic diagnosis of Hunner’s lesion without bladder distension in general anaesthesia is often difficult or impossible (M. Fall, personal communication). After distension, deep bladder biopsies including detrusor are obtained from the two lateral walls and dome, in addition to biopsies from lesional areas (82). The bladder morphology is evaluated, including presence of fibrosis and mast cell quantification in the detrusor (55;82). For exclusion of confusable conditions, it may be necessary to perform other examinations, laboratory analyses and uroflowmetry with potassium sensitivity test, for which the ESSIC has provided guidelines (8). At the Department of Urology at Herlev Hospital, primary evaluation of patients suspected of having the disease has always been performed using six clinical evaluation criteria for BPS, introduced by the late Professor, DMSc Tage Hald, comprising the following variables: bladder pain/pressure/discomfort; nocturnal frequency; bladder capacity under general anaesthesia; mucosal glomerulations and/or Hunner’s lesion; detrusor mastocytosis; and detrusor fibrosis (Table 1) (82;102).

15

Table 1. Clinical evaluation criteria for severity of Bladder Pain Syndrome, used at Herlev Hospital.

Criteria

Positive

Pain

Yes

Nocturnal frequency

≥twice

Bladder capacity

≤500 ml

Glomerulations and/or Hunner’s lesion

Yes

Detrusor mastocytosis

≥27 cells/mm2

Detrusor fibrosis

Yes

Treatment Treatment possibilities are limited and include intravesical drug instillation, systemic (oral) medical therapy, and surgical procedures for refractory disease. Intravesical drug instillation or oral therapies are the mainstay of treatment (41;103). Intravesical agents are further described in Appendix A. In extreme cases, radical surgical treatment (e.g. urinary diversion, cystectomi) and pain clinic care are the only final options that provide relief in some patients. Curative therapy has not been found due to the unclear pathophysiology of BPS, with the exception of patients presenting with classic BPS, which is the goal for some specific, successful treatment strategies (33;34;104). At the Department of Urology at Herlev Hospital, analgetics including opioids, antiepileptics, antidepressives and tranquilizers are used for oral therapy, and GAG layer replacement or dimethyl sulfoxide (DMSO) intravesically. In severe cases systemic steroids may be prescribed. Agents used more sporadically are antihistamines and leukotrien-receptor antagonists. Years ago, also other medications were used through shorter periods (cetiprin systemically; and lapis, bacillus CalmetteGuerin, heparine, sucralfate, gentamicine intravesically). Some agents (e.g. pentosan polysulfate sodium) used in other countries are not registered for use by the Danish Medicines Agency. Neuromodulation comprising sacral nerve stimulation and posterior tibial nerve stimulation, is not routinely applied, but has been performed in refractory cases or in research with diverse results (105-110). Multimodal therapy including a psychological approach is imperative when treating

16

patients with BPS, since coping strategies are necessary to manage symptoms and problems related to the disease (103;111-115). Questionnaire tools A questionnaire is a clinical and a research instrument consisting of a series of questions for the purpose of gathering specific information from respondents. Questionnaires often used in BPS in order to evaluate symptoms are the O’Leary Sant Symptom and Problem Index (OSPI), the Pain, Urgency, Frequency Symptom Scale (PUF) and the Global Response Assessment (GRA) (116118). The OSPI, comprising a symptom index (ICSI) and a problem index (ICPI), was developed to measure voiding and pain symptoms and the degree to which patients find them problematic, and was designed and validated for patients’ self-administration (116). The index was not intended as a screening tool for BPS, but for complementary use at evaluation of symptoms during the diagnostic process, and for monitoring changes in clinical status as a result of treatment of patients with BPS (116;119). A later study validated the OSPI for correlation of symptom severity to other determinants of BPS (16). The PUF Scale quantifies severity of BPS symptoms and how bothered the patient is by these, like the OSPI (117). However, the PUF Scale was validated by comparison to the PST which showed, that the PUF Scale could reliably predict potassium sensitivity test outcomes (117). Another study investigating the validity of the PUF Scale according to findings at cystoscopy with hydrodistension showed that the tool was neither a reliable predictor of BPS nor of disease severity (120). The reliability of both questionnaires in distinguishing patients with BPS from others with urinary tract symptoms was evaluated later, showing that neither of them was sufficiently specific to be used as sole diagnostic indicator (121). As mentioned before, both scales were confirmed to be useful as complementary tools (121). Another important assessment scale is the GRA including 5-7 categories for evaluation of symptoms related to treatment effect, which is a valuable treatment outcome indicator (118;122-125). In the present thesis, the OSPI was chosen for evaluation of symptoms and associated problems. For assessment of the quality of life experienced by patients with BPS, the standardized Short Form-36 (SF-36) was used. Quality of life is impaired in many patients suffering from BPS, in some due to a depression (113;114;126;127). SF-36 comprises eight health status scales, evaluating both physical and mental health, and is a tool which can be used for general evaluation in both healthy individuals and patients with different diseases (128;129).

17

Biomarkers A NIH (National Institute of Health, USA) study group agreed to the following definition of a biomarker in 1998: "a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention" (130). Biomarkers can serve many purposes, including confirmation of diagnoses, monitoring treatment effects or disease progression, and prediction of clinical outcomes (131). Noninvasive assessment of a biomarker in blood and/or urine would be of benefit for patients with BPS, supplementing or even replacing cystoscopy with hydrodistension and bladder biopsy in general anaesthesia. The search for a specific, reproducible biomarker for BPS, which diagnosis is based on symptoms and exclusion of confusable diseases, has been intensive. Investigated urinary biomarkers include antiproliferative factor (APF) (132-135), epidermal growth factor (EGF) (124;136) and interleukin6 (IL-6) (124;136;137), besides numerous others (138-143). Mast cell secreted urinary substances such as histamine and its metabolite methylhistamine have been investigated, since BPS has been assumed to result from a chronic inflammatory condition involving mast cells in the bladder wall (8;55;58;73;75;137). However, no suitable molecule has been identified until now, because none of the evaluated biomarkers has proved to be specific for BPS or reproducible in prospective studies.

YKL-40 YKL-40 (also named human cartilage glycoprotein 39, chitinase-like-3 protein-1, or chondrex) is a 40 kDa heparin-, collagen- and chitin-binding glycoprotein without chitinase activity, and a member of “mammalian chitinase-like proteins” (144). YKL-40 is named after its three N-terminal amino acids (tyrosine (Y), lysine (K) and leucin (L)) and its molecular weight (145). Human YKL-40 consists of a single polypeptide chain of 383 amino acids (144), and the crystallographic threedimensional structure is made of two globular domains, whose folding structure is conserved among the 18-glycosyl-hydrolase family (146). The human YKL-40 gene (CHI3L1) consists of 10 exons and is located on chromosome 1q32.1 (144;147). In healthy subjects, YKL-40 can be measured in serum with a median concentration of 43µg/L by a commercial enzyme-linked immuno sorbent assay (ELISA) (148;149). Serum YKL-40 does not depend on gender, but increases with age (148;149). YKL-40 mRNA and protein expression have been demonstrated in human embryonic and fetal tissues in all three germ layers in vivo, and a high protein expression was associated with cell proliferation, differentiation, and tissue morphogenesis (150). In normal adult tissue, YKL-40 protein expression, assessed by immunohistochemistry

18

(IHC), has been demonstrated in cells and tissues originating from all three germ layers as well (151). A strong YKL-40 expression is found in cells with high proliferation/turn-over rate and cells with high metabolic activity (151). The different cell types in the bladder were therefore not or only weakly stained, except for the urothelium which was stained moderately to intensely (151). It is interesting to note, that mast cells and macrophages with YKL-40 positive staining were not identified in normal bladder tissue, but in other normal tissues, such as gall bladder, ileum, colon, fallopian tube (151). YKL-40 is expressed in inflammatory diseases in different cell types (e.g. macrophages, neutrophil granulocytes, chondrocytes, vascular smooth muscle cells, mast cells) (148;152-163). It is seen in macrophages in the inflamed synovial membrane from patients with rheumatoid arthritis, showing a relationship between high serum YKL-40 and progression in joint destruction (157). Also, the number of YKL-40 positive cells is related to the degree of synovitis in patients with rheumatoid arthritis and osteoarthritis (163). In patients with asthma, serum YKL-40 correlates with YKL-40 positive cells in lung biopsies and with disease severity (159;160;164). Excessive YKL-40 production may cause matrix accumulation, leading to tissue fibrosis (165). YKL-40 has been demonstrated to promote growth of fibroblast cell lines through activation of MAP kinase and PI3K signalling pathways propagating mitogenic signals (166), and it binds to type I collagen modulating the rate of type I collagen fibril formation (165). YKL-40 secreted from adipose tissue inhibits the degradation of type I collagen (167). Recent studies have shown a correlation between YKL-40 and fibrosis in diseases such as liver fibrosis (152;156;161), rheumatoid arthritis (154;155;157;163), atherosclerosis (153;158), asthma (162;164) and chronic obstructive lung disease (159;160). Also in cancer, YKL-40 is supposed to have a biological function not yet understood. Several studies have shown that elevated serum concentration of YKL-40 in patients with breast, colorectal, ovarian, kidney, prostate, small cell lung cancers, and malign melanoma was an independent prognostic variable of short recurrence-free interval and short overall survival with hazard ratios between 1.3 and 4.1 (168-180).YKL-40 positive mast cells were demonstrated in patients with primary breast cancer and ovarian cancer by IHC (181;182). YKL-40 positive macrophages were found in patients with breast cancer (182), ovarian cancer (181), small cell lung cancer (183) as well as squamous cell carcinoma of the head and neck (180). These studies suggested that YKL-40 positive inflammatory cells, herein tumour associated macrophages, could be involved in preserving extracellular matrix during tissue remodelling or destruction (183). Based on previous findings of YKL-40 in diseases with inflammation and fibrosis, it was hypothesized that YKL-40 may play a role in detrusor fibrogenesis in BPS. 19

Aims of the study The aims of the present PhD thesis were:

1. To describe symptoms, quality of life and clinical characteristics of Danish patients with BPS and long-term follow-up (Papers I and II).

2. To evaluate associations between quality of life, clinical variables (symptom scores, pain, nocturnal frequency, bladder capacity, mucosal glomerulations, detrusor mastocytosis, detrusor fibrosis) and outcome in terms of treatment intensity (Papers I and II).

3. To evaluate the utility of the ESSIC classification from 2008 for the identification of patient subgroups with different prognosis and requirements for treatment in a Danish patient population with BPS (Papers I and II).

4. To evaluate the utility of YKL-40 as a prognostic biomarker for bladder fibrosis in patients with BPS by YKL-40 protein expression in bladder tissue and YKL-40 levels in serum and urine (Paper III).

20

Materials Patients The thesis is based on a population of 386 patients with a clinical diagnosis of BPS, registered according to the internationally valid coding system and identified in Herlev Hospital’s data register. The population comprised 360 women (93.3%) and 26 men (6.7%), with a median age at diagnosis of 53 (range 16-88) years and follow-up duration at hospital of 3 (range 0-38) years. All patients fulfilled the diagnostic ESSIC criteria. At study baseline, 281 patients were alive, three of whom had moved to another country; 105 had died since their last hospital follow-up. Patients were included in the study as follows:

Paper I: 278 patients living in Denmark were included in a questionnaire study (Figure 2). Paper II: Data from all 386 patients were included in a descriptive study of BPS. Data from 349 patients were complete for statistical evaluation. Paper III: 55 patients were included in a Pathology study.

Figure 2. Flow-chart used for inclusion to questionnaire study.

386 patients in database followed between 1966 -2007

Patients not included:

278 patients received

105 dead 3 moved to other countries

postal questionnaire

postal reminders 3 weeks after initial sending

146 answered questionnaires used for analysis

132 not returned/answered

21

Database Before study initiation, data were obtained retrospectively from patient files at Copenhagen University Hospital Herlev. The following variables were entered onto a computerised database: gender; age at diagnosis; year of diagnosis and of last hospital follow-up; pain related to the bladder (yes/no); nocturnal frequency (number of voids); bladder capacity in general anaesthesia (ml), and presence of mucosal glomerulations and Hunner’s lesion (yes/no) evaluated at cystoscopy with hydrodistension in general anaesthesia; mast cell count and presence of fibrosis at histological evaluation of detrusor muscle in bladder biopsies. Data were originating from the primary evaluation of patients with BPS before treatment at Herlev Hospital. All patients were classified into four categories according to the regimen/intensity of treatment they received or had received through their observation period at the department, a variable described in detail in Paper II, used as an expression for disease severity (Table 2). Furthermore, all patients were classified according to the ESSIC-guidelines (Figure 1) (8). The database was updated at data collection from answered questionnaires (Paper I). Table 2. Treatment intensity (treatment regimen) for Bladder Pain Syndrome.

Category

Detail

1

Cystoscopy with hydrodistension and deep bladder biopsies as a single treatment

2

As 1, plus one treatment with one either oral or intravesical treatment cycle (equivalent to one medical treatment within the total observation period)

3

As 1, plus several treatment cycles (equivalent to two or more medical treatments within the total observation period)

4

High-dose steroids, surgery, referral to pain clinic (at least once within the total observation period)

From Richter et al. Bladder pain syndrome/interstitial cystitis in a Danish population: a study using the 2008 criteria of the European Society for the Study of Interstitial Cystitis. BJU International 2010;105(5):660-667. Used with the kind permission of Wiley-Blackwell Publishing Company and the BJU International.

Tissue collection (Paper III) Formalin-fixed, paraffin-embedded (FFPE) bladder tissue specimens from 10 patients known to have BPS with significant detrusor mastocytosis were selected from the archives of the Department of Pathology, to examine whether YKL-40 immunoreactive cells were present in the detrusor

22

muscle in BPS. For further advanced analyses a consecutive series of 45 patients with BPS or suspected of suffering from BPS was included, collecting serum and urine samples from all 45, FFPE bladder tissue specimens from 44 and fresh bladder biopsies from eight of the patients.

Controls Single- and double-labelled immunohistochemistry (IHC), immunoelectron microscopy analysis: For positive control for YKL-40, a paraffin-embedded cytoblock of the YKL-40 producing osteosarcoma cell line MG-63 and tissue from invasive breast carcinoma were included as in previous studies (184;185). For mast cell tryptase, tissue from skin, appendix and bladder with naphtol esterase positive granule-containing mononuclear cells was used, while a tonsil with macrophages and granulocytes, and appendix with histiocytosis was used for CD68. ELISA: A commercial ELISA (Quidel Corporation, San Diego, CA, USA) was used. In the ELISA kit two controls (i.e. a sample with a YKL-40 concentration of approximately 40 ng/ml and a sample with a YKL-40 concentration of approximately 150 ng/ml) are included. YKL-40 in these samples is purified from serum-free conditioned media of cell cultures of the osteosarcoma MG-63 cell line. The same YKL-40 was used for the standards included in the standard curve in a concentration of 0-300 ng/ml. The healthy controls consisted of A) 45 age- and sex-matched subjects without BPS, recruited from the staff at Herlev Hospital; and B) 245 healthy blood donors from Hvidovre Hospital and elderly persons, described earlier (149).

Antibodies (Paper III) YKL-40 antibody The non-commercial, monoclonal mouse antibody (mAb) 201.F9 (isotype IgG 2b,κ) was selected for YKL-40 analyses and donated by Professor, PhD Paul A. Price, Department of Biology, University of San Diego, La Jolla, USA. The antibody was generated by immunisation of mice with YKL-40 protein, purified from serum-free, conditioned media from monolayer cultures of the YKL-40 producing human osteosarcoma cell line MG-63 by Professor, DMSc Julia S. Johansen in collaboration with Professor, PhD Paul A. Price (145).

Mast cell antibody Monoclonal Ab anti-human mast cell tryptase mouse antibody, clone AA1, M 7052 (Dako, Glostrup, Denmark) was used.

23

Macrophage antibody Monoclonal Ab anti-human CD68 mouse antibody, clone KP1, M0814 (Dako, Glostrup, Denmark) was used.

Ethics The study was approved by the Regional Research Ethics Committee (H-KA-20060168). Patients donating fresh tissue gave their informed written consent. The Danish Registry of Human Tissue Utilization was consulted before the use of archival tissue samples. The Danish Data Protection Agency accepted the collection of data.

Methods Primary patient evaluation The Departments of Urology at Copenhagen University Hospitals Gentofte, and since 1976 Herlev have been referral centre for patients with BPS from all regions of Denmark, at the initiative of the late Professor, DMSc Tage Hald (TH). For over 40 years patients with BPS have been examined according to a standard evaluation (Table 1) by the same two senior urologists (Professor, DMSc Jørgen Nordling and TH) (8;82). At the Pathology Department, morphological examination of bladder biopsies has been performed following a standardised protocol by the same two Pathology experts (Professor, DMSc Svend Larsen and Professor, DMSc Thomas Horn). The evaluation procedures are described before (se paragraph Diagnosis) (8;82), and in Papers II and III. Findings from the present patient population were recorded in the database.

Questionnaire study (Paper I) A questionnaire study was conducted between October and December 2007 to report the patients’ evaluation of their bladder disease, severity of symptoms and related problems, and impact on their quality of life. A postal questionnaire in Danish language, consisting of four separate parts, was sent to 278 patients with BPS living in Denmark (Figure 2). Non-respondents received a reminder after three weeks. Response rate was 146 out of 278 (53%). Data were entered onto a computerised database, recoded and checked by a statistical researcher (Ulrik Hesse, UH) and a technical assistant. Finally, 24

data were merged with the pre-existing database for analysis. Further details regarding the study course are described in Paper I. Below the four separate questionnaires are outlined: BPS questionnaire A questionnaire was devised by our group and included 28 items covering demographics; dates for onset of symptoms, diagnosis and last hospital follow-up; pain experience at three different points of time; pain location and character; treatments received for BPS and their efficacy; medication in general; co-morbidity; changes in disease condition and quality of life experienced through past year/since the onset of BPS. Pain was reported on visual analogue scales (VAS) from 1–10 and categorised for analysis (mild pain: VAS score 1-2, moderate: 3-6, severe: 7-10). A more detailed description of the BPS questionnaire and the construction process is given in Paper I. The questionnaire is attached as Appendix B. O’Leary Sant Symptom and Problem Index (OSPI) This validated composite of two subscales assesses the symptom severity associated with BPS, and the problems associated with symptoms, respectively. The ICSI score ranges from 0-20, the ICPI score from 0-16; the higher score, the more severe symptoms and problems. For analysis, ICSI and ICPI scores were categorised into three groups each (ICSI: mild 0-8, moderate 9-13, severe 14-20; ICPI: mild 0-5, moderate 6-12, severe 13-16). The OSPI is validated and used internationally in patients with BPS (21;22;113;119;186-193). The questionnaire has been translated to Danish by Corporate Translations, East Hartford, CT, USA, and validated in studies at the Department of Urology at Herlev Hospital. SF-36 General quality of life is evaluated with this widely used, standardised questionnaire consisting of 36 items in eight health status scales including physical functioning (PF), role limitations due to physical health problems (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), role limitations due to emotional problems (RE) and mental health (MH) (128). The eight scales are scored individually; the maximum for each scale score is 100, and a high score means a good quality of life. The Danish version of SF-36, version 1 was translated and validated in 1993 (128). 24 h voiding diary A 24 h diary evaluated the voiding frequency without measuring urine volume. 25

Methodological evaluation for Paper I A questionnaire consists of a series of questions for the purpose of gathering specific information from respondents. Question construction and wording are of great importance for, how respondents understand and answer the questionnaire, and whether these answers cover the topic the researcher wants to investigate (validity). Some rules are important for constructing items; questions have to be short, clear and unambiguous, user-friendly, fair and not offensive, and respondents can be expected to know the answer (194;195). The answering mode has to be considered; free-text, standardized answers or scales. Standardized answers make it simple to compile data, but have the disadvantage that respondents can not elaborate their answers. The questionnaire items should be structured following some logic order, and informative instructions have to be included. Finally, data collection and processing have to be performed reliably. Apart from two validated questionnaires (SF-36 and OSPI) and a voiding diary used in our study, a BPS questionnaire was constructed in order to collect information on various different aspects of the disease. Following the mentioned construction rules, the questionnaire was revised after to minor pilot tests. Before sending, a statistical researcher (UH) with specific experience in questionnaire surveys was consulted for final item construction, layout, and for data processing after questionnaire collection. Evaluation after analysis revealed that main weaknesses of the BPS questionnaire were found in a few items of great complexity, and the fact that some respondents had problems to recall details from the course of their disease. However, most items revealed valuable information.

Descriptive study (Paper II) A descriptive study was performed to characterize and evaluate a patient population with BPS from Denmark. The formerly mentioned clinical evaluation criteria for BPS (Table 1), introduced by TH, were reported as a clinical severity score (102). A ‘scoring system’ was applied to patients with pain (equivalent to all in the database), giving one point for each variable met and using the following thresholds: nocturnal frequency of twice or more, bladder capacity of ≤500 mL, mucosal glomerulations and/or Hunner’s lesion, detrusor mastocytosis, and detrusor fibrosis, resulting in a final score of 0–5 points. Preliminary evaluation was performed in 2001 by DMSc Svend Mortensen (102).

26

Lacking a definite measurement for clinical outcome, patients were classified into four categories according to the regimen/intensity of treatment they had received through their observation period at the department, using this as an expression for disease severity and prognosis (Table 2). Those placed in one group had implicitly passed through groups of lower ranks previously. Associations between clinical variables, the clinical severity score mentioned above, treatment intensity and BPS types according to the ESSIC classification were investigated. Definitions and procedures are described in details in Paper II.

Methodological evaluation for Paper II Clinical databases are widely used in research and are inevitable for systematic evaluation. The initiation of a database requires expert design to determine which data are necessary and what relationships exist between the different elements, as well as to superimpose a logical structure on these data. There are several sources of bias related to database design and data entry, which may have consequences for analyses and results. In the present thesis, data collected retrospectively from medical files of 386 patients with BPS comprised both quantitative and qualitative data and were used for a descriptive study. The excelbased database was initiated by TH who manually entered approximately 300 patients’ clinical data until 1999; thereafter data were obtained by Benedikte Richter (BR). To minimize entry errors, data were double-checked at registration, and BR cross-checked data from about 150 patients registered by TH, encountering discrepancies in less than 1% of the entered variables. For the present study, the database was found to be sufficient for the purpose of analysing associations between different clinical variables.

Pathology study (Paper III) A Pathology study was conducted in patients with BPS to examine whether the biomarker YKL-40 is present in bladder biopsies and in which specific cells the protein is expressed; to investigate the secretion of YKL-40 in serum and urine; and to evaluate YKL-40’s potential as a non-invasive, prognostic biomarker for bladder fibrosis in BPS.

Immunohistochemistry To verify the presence of YKL-40 immunoreactive cells, tryptase positive mast cells and CD68 positive macrophages, three immunostainings were performed on FFPE, 3 µm thick tissue sections

27

of each bladder biopsy. Detrusor muscle fibrosis was detected as intrafascicular fibrosis using Picro-sirius Red staining (82;196). Briefly, IHC refers to the process of localizing antigens in cells of a tissue section exploiting the principle of antibodies binding specifically to antigens in biological tissues (197). Different cell types express different antigens which can be used for characterization and identification of the specific cells. Visualizing an antibody-antigen interaction can be accomplished by conjugation of an antibody to an enzyme, such as peroxidase, that can catalyze a colour-producing reaction, which was used in the present IHC method for YKL-40.

YKL-40 analysis The IHC method for YKL-40 in frozen and paraffin-embedded tissue was developed by Anne Roslind et al (184), described detailed in a PhD thesis, defended in April 2008 (198). The IHC method evolved after thorough analyses of fixation procedures, epitope retrieval and blocking protein (184). IHC staining of the bladder specimens with mAb YKL-40 201.F9 (3.8 mg/ml, dilution 1:60 in TRIS-buffered saline) (184) was performed manually, applying an indirect method with a peroxidase labelled polymer conjugated to goat antimouse immunoglobulins (Dako EnVision® + System/HRP K4007, Dako, Glostrup, Denmark). As substrate chromogen 3-amino-9ethylcarbazole (AEC), K3469 (Dako, Glostrup, Denmark) was used, polymerizing to an insoluble red polymer at the site of the antigen in the presence of horseradish peroxidase, which easily can be detected in the microscope. The IHC method is described in further detail in Paper III.

Mast cell analysis To localize mast cells in the bladder biopsies, IHC staining with mAb Mast Cell Tryptase (dilution 1:3000) was performed, using a standardized protocol for the Dako Autostainer (Dako, Glostrup, Denmark).

Macrophage analysis To localize macrophages in the bladder biopsies, IHC staining with mAb CD68 (dilution 1:5000) was done following standard procedure for the Dako Autostainer.

Double-labelled immunohistochemistry Double-labelled IHC was performed on six selected biopsies to study YKL-40 expression in mast cells and macrophages in the bladder specimens. This method enables immunostaining of two different antigens on the same slide for co-localization studies. The EnVision® G|2 Doublestain System, Rabbit/Mouse (3,3’diaminobenzidine (DAB+)/Permanent Red) K5361 (Dako, Glostrup, Den28

mark) was used. MAb YKL-40 201.F9 (3.8 mg/ml, 1:60) (184) was labelled with Permanent Red (red label); and tryptase (mAb clone AA1, M 7052; 1:3000) or CD68 (KP1, M0814; 1:10000), respectively, were labelled with DAB+ (brown label).

Quantification of single-labelled immunohistochemistry Single-labelled cells from the 45 consecutive patients with BPS were quantified for correlation with serum and urine YKL-40 concentrations and with clinical parameters. To evaluate a suitable quantification method, 10 biopsies were chosen randomly and evaluated by both morphometric counting and stereology (82;199;200). Briefly, a morphometric method is used for routine detrusor mast cell counting at the Department of Pathology (recommended by the ESSIC (8;82)). A microscope type Leitz, Dialux, Wetzlar, Germany is used at a magnification x25, containing a measuring grid (Leitz periplan GF 10xM ocular containing a standardized grid with 25 squares, each square measuring 0.21 sq.mm)(82). The whole biopsy is evaluated by counting the immunoreactive cells in the detrusor muscle without selection of tissue areas. Detrusor mastocytosis is defined as at least 27 mast cells per sq.mm, using tryptase staining (201). In our study, a median of 10.5 (range 6-25) fields of detrusor muscle were counted in each specimen. Stereological methods, on the other hand, are precise and reproducible for obtaining quantitative information about three-dimensional structures, based mainly on two-dimensional sections, described in detail by Gundersen et al.(200). Calculations are performed by using a counting field and a grid system with a set of regularly spaced points, dividing the number of points hitting the structure with the number of points hitting the total area of the counting field. The total area of detrusor muscle in each biopsy was evaluated by using a grid of 500x500 sq.nm between the points, and at least 5000 points were counted. The counts resulting from the two methods correlated (r=0.72, p=0.02), and since morphometric counting was less time consuming, it was preferred in the final analysis of all specimens. Further details are described in Paper III.

Immunoelectron microscopy analysis The immunoelectron microscopy analysis was performed to investigate the subcellular localization of YKL-40 in bladder tissue (185). Briefly, immunoelectron microscopy is based on the detection principle of antibodies binding to specific antigens in biological tissues parallel to IHC, just at a high resolution by electron microscopy. Fresh tissue samples are immediately fixed in paraformal-

29

dehyde and cryoprotected with polyvinyl pyrrolidone/sucrose, followed by freezing in liquid nitrogen and sectioning at about -115ºC. Ultrathin sections of 110 nm are placed on electron microscopic support grids for subsequent labelling with mAb YKL-40 201.F9. For visualization, the marker Protein A Gold (EM.PAG10, British Biocell International, Cardiff, UK) is applied, which is an immunoglobulin-binding polypeptide conjugated to10 nm gold particles (202). This gold probe detection system reveals YKL-40 through the visibility of the gold nanoparticles in the transmission electron microscope (Morgagni 268D, FEI Company, Eindhoven, The Netherlands), providing a permanent signal with high resolution. The procedure is described in detail in Paper III.

Enzyme-linked immuno sorbent assay analysis ELISA analysis was performed to evaluate YKL-40 concentrations in serum and urine in the consecutive patient population with BPS. Briefly, ELISA is a biochemical technique to detect the presence and concentration of an antigen or an antibody in a sample. In simple terms, an unknown amount of antigen is affixed to a surface, and then a specific antibody is washed over the surface to detect the antigen. This antibody is linked to an enzyme, and in the final step a substance is added and converted to some detectable signal by the enzyme. In our study, a commercial two-site sandwich ELISA (Quidel Corporation, San Diego, CA, USA) was used (149;154). YKL-40 in the patient sample is captured by the Fab fragment of a monoclonal mouse anti-human YKL-40 antibody. This antibody is conjugated to biotin, which binds to the streptavidin-coated microplate wells. A polyclonal rabbit anti-human YKL-40 antibody conjugated to alkaline phosphate binds to the captured YKL-40, after which bound enzyme activity is detected with p-nitrophenyl phosphate as substrate. The analysis is finished within 4 hours and involves three 1-hour incubation steps, is carried out at room temperature and does not require sample dilution (except for samples with very high YKL-40 concentration, i.e. >300 µg/l). Samples were analysed in duplicates. When the visualization step was stopped after one hour, optical densities were determined using a microplate reader at 405 nm. For further details se Paper III.

Methodological evaluation for Paper III Immunohistochemistry Immunohistochemistry is a powerful method for identification of antigens in cells and tissues, increasingly applied for diagnosis, and assessment of biomarkers. Many factors have influence on the final result of immunostaining, including sample obtainment, tissue processing, antigen retrieval

30

and staining methods (203-208). Furthermore, interpretation and quantification of immunostainings, based on subjective evaluation, may introduce measurable variability, requiring standardization of procedures (199). An ideal method aims at high reproducibility, consistency and minimal variability. In the present study, single-labelled and double-labelled IHC was done manually according to a recently developed IHC method for YKL-40 on FFPE tissue (182;184;185). As mentioned before, this method was developed in a process following guidelines for methodological approaches, suggested by an International Consensus Group of Standardization and Quality Control in Immunohistochemistry (199;207). The method was evaluated to be specific for YKL-40, valid and reproducible (184). For staining of sections from FFPE bladder tissue specimens, the method was adjusted requiring analyses with different antibody dilutions to minimize background staining. Incubation of tissue sections with the antibody over night resulted in satisfactory staining results. To ensure the method’s specificity, negative and positive controls were included. To show that labelling was specifically due to the primary antibody, negative control staining was performed by omitting mAb 201F9. The immunostainings were interpreted using a binary positive-negative endpoint without graduation. YKL-40 positive cells appeared with intensely red stained cytoplasmic granules. Quantification of numbers of stained cells per section was done twice by the same person, blinded to clinical data, described in detail in Paper III. The chosen morphometric counting method has been described and recommended by ESSIC, in order to assure comparable results from different studies (8;82). In contrast to this, stereology is largely concerned with the three-dimensional interpretation of planar sections of materials or tissues, providing specific practical techniques for quantification (200). Stereology utilizes random, systematic sampling to provide unbiased and quantitative data. In Paper III, the method recommended by ESSIC was compared to stereological quantification in ten sections, correlating well. Based on the time spent on both quantification methods, the routine ESSIC procedure was preferred. Generally it would be preferable with biopsy evaluation by two independent, blinded examinators to confirm consistency and reproducibility of the quantification, which unfortunately was impossible in the present study. An ideal scenario would be computerized stereological cell quantification in biopsies, which is used in some laboratories (209), since stereology is supposed to provide unbiased reproducible measurements.

Immunoelectron microscopy Immunoelectron microscopy is a powerful technique allowing detection and specific localization of cellular antigens. The development of the methodology for ultrathin frozen sections, which were 31

used in the present study, has made it possible to retain the antigens in a hydrated environment prior to immunolabelling, thereby improving antigen accessibility compared to resin section labelling (210;211). A central issue for immunolocalization at the level of electron microscopy is to preserve the epitope of interest while minimizing the loss of ultrastructural details. As in other histochemical techniques, many factors contribute to the final results of immunoelectron microscopy (212). One important factor relates to fixation conditions, including time to fixation. Since cellular ultrastructures are vulnerable to degradation, immediate fixation of the tissue is important (211). In the present study bladder specimens were obtained at cystoscopy under general anaesthesia and fixed immediately in paraformaldehyde, transported to the Pathology Department within 5 minutes for cryoprotection, followed by freezing in liquid nitrogen and sectioning at about -115ºC, which minimized the loss of ultrastructural details as much as possible. The procedure followed guidelines developed in a recent study on the biomarker YKL-40 in breast cancer (185). The morphology of the bladder tissue showed to be satisfactory for the purpose of investigating the subcellular localization of YKL-40.

ELISA ELISA is a widely used technique to measure even very low protein concentrations in body fluids such as serum and urine. The advantage of this method is the non-invasive assessment of the biomolecule of interest. Since the principle of antigen-antibody binding is the basis for ELISA as for other histochemical methods, the same factors can influence on the analysis. The protein of interest has to be present in detectable concentrations and present with epitopes recognizable by specific antibodies. In the present thesis a commercial ELISA was used to detect YKL-40 concentrations in serum and urine in accordance with the manufacturer’s instructions (149;154). The epitopes of YKL-40 specifically recognized by the antibodies used in the assay are unknown (148). Blood samples taken for analysis have to be handled carefully, since YKL-40 concentration in whole blood increases with time, requiring that the interval between drawing and centrifugation of blood stored at room temperature must be less than three hours for serum (148;213). Otherwise significant and not disease-related elevations of YKL-40 levels are found. If the blood is stored at 4ºC before centrifugation, YKL-40 serum concentration is stable for 24 hours (213). Urine levels of YKL-40 have not been assessed by the used assay before, which is why they were processed within three hours, comparable to blood samples. Previously the median serum YKL-40 level in healthy subjects was determined to be 43 µg/L (range 20-184 µg/L) (149). The wide range seen in healthy subjects shows some overlap with serum concentrations measured in the patients with BPS (22-687 µg/L), showing

32

a test sensitivity far from optimal, which is one of the reasons why YKL-40 is not useful as a routine biomarker.

Statistics Analyses were done using SAS version 9.1., SAS Institute Inc., Cary, NC, USA. P-values of ≤0.05 were considered as statistically significant.

Paper I Associations and differences in categorical variables originating from the questionnaire study, and clinical variables from the patient database, including treatment intensity (Table 2), were evaluated by Chi Squared test, T-Test and Anova.

Paper II A proportional odds model was used to analyse associations between clinical variables (gender, age at diagnosis, nocturnal frequency, bladder capacity under general anaesthesia, mucosal glomerulations, detrusor mastocytosis, detrusor fibrosis) and clinical outcome in terms of treatment intensity (Table 2), with graduation of treatment intensity as the dependent variable. As all patients had pain/pressure/discomfort related to the bladder, this variable was not included in the multivariate analysis. For each of the four treatment groups cumulative logits were modelled by using ordinal logistic regression (214). The cumulative logits were logits of the cumulative probabilities, literally the log odds of group 4 to group 3, 2 and 1, the log odds of group 3 to group 2 and 1 and the log odds of group 2 to 1. The function proc genmod was used (214).

Paper III Correlations between continuous variables were analysed with Spearman’s correlation. General Linear Modelling was performed to analyse age-corrected serum and urine levels of YKL-40 in patients with BPS compared to healthy controls. Wilcoxon Sign Rank, Savage One-Way analysis and Students T-test were used to analyse associations between different groups of clinicopathological parameters.

33

Results and discussion Results from Papers I and II Symptoms, quality of life and clinical characteristics of Danish patients with BPS In Paper I, a questionnaire study was conducted to evaluate symptom severity and quality of life, and associations with clinical variables. The 146 respondents (of 278 patients) had a median age at diagnosis of 50 (range 17-79) years and follow-up at hospital of 5 (range 0-27) years. Reported median duration of symptoms before diagnosis was 4 (interquartile range 1-8, max 66) years; median interval between last follow-up and questionnaire analysis was 5 (range 0-30) years. Clinical characteristics are shown in Table 3. OSPI scores revealed that a third of the respondents had severe symptoms (ICSI≥14 out of 20) and severe, associated problems (ICPI≥13 out of 16). During past month more than a half suffered from moderate to severe pain (VAS 3-10), 25% had no pain (VAS 1) (Figure 3).

Figure 3. Reported average pain at disease onset, one year ago and during past month.

none (VAS=1)

60

mild (VAS=2) moderate (VAS=3-6)

50

severe (VAS=7-10)

40 % of the respondents

missing

30 20 10 0 onset

1 year ago

past month

The bladder and urethra were the most frequently mentioned pain sites (in 67% and 45% of the respondents, respectively) besides others. Pain was most frequently characterized as painful urge to urinate, or as sharp, pressure or burning. Median four treatment cycles using available regimens for BPS were applied, and the condition tended to improve in a third to a half of the treated individuals. Co-existing conditions, such as back pain, arthritis, digestive problems and irritable bowel, fatigue and allergy were frequent. Twenty-one percent had a depression.

34

Table 3. Clinical characteristics of questionnaire respondents (Paper I).

Characteristic Pain N (%) * Nocturnal frequency Median (range) Missing * 24 h voiding frequency Median (range) Missing Bladder capacity at cystoscopy (ml) Median (range) Missing Mucosal glomerulations N (%) No Yes Missing Mast cell count (cells/sq.mm) Median (range) Missing Detrusor fibrosis N (%) No Yes Missing Treatment intensity group N (%) 1 2 3 4 ESSIC types N (%) 1A 1C 2A 2B 2C 3B 3C Missing O'Leary Sant Index median (range) ICSI ICPI

146 (100) 4 (0-10) 50 11 (1-24) 50 520 (100-1500) 11 41 (28) 104 (71) 1 29 (0-170) 2 76 (52) 68 (47) 2 23 (16) 21 (14) 58 (40) 44 (30) 17 (12) 23 (16) 22 (15) 8 (5) 60 (41) 1 12 (8) 3 (2) 11 (0-20) 10 (0-16)

Quality of life scores were considerably lower in all SF-36 domains than in a normative sample of Danish adults (16+ years), with mean score differences between 11 and 42 points (Figure 4) (128). In comparison, score differences found between a normative Danish sample and groups with several

35

self-reported chronic diseases (e.g. diabetes, chronic obstructive pulmonary disease (COPD), back pain, migraines, peptic ulcer, psoriasis) were smaller in all domains than the differences seen in our study, indicating greater decrements in health related quality of life in patients with BPS (128).

Figure 4. SF-36 mean scores in patients with BPS, self-reported COPD, self-reported back pain, and Danish norm data. 100 90 80 70

Score

60 50 40

BPS patients

30

Danish norm data COPD patients

20

Patients with back pain 10 0 PF

RP

BP

GH

VT

SF

RE

MH

Health status scales: physical functioning (PF), role limitations due to physical health problems (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), role limitations due to emotional problems (RE), mental health (MH).

Comparing current overall quality of life with the first year of having BPS, more than half of the respondents felt improvement, 23% of these having moderate to severe average pain during past month; 19% felt worsening (Table 4). Present treatment had improved quality of life in more than a third, whereas it was unchanged in 25% and worse in 9%; 30% did not answer this question. Nevertheless most respondents were unsatisfied with having to live with their bladder disease at present state for the rest of their lives, most of them experiencing moderate to severe average pain during past month.

36

Table 4. Quality of life related to symptoms and treatment, reported by questionnaire respondents (Paper I). Questions included in the self-constructed BPS questionnaire, which was sent together with the validated SF-36 for assessment of quality of life.

Topic

N (%)

A: Did your bladder disease change (better/worse) in the past year? Much better Slightly better Unchanged Slightly worse Much worse Not answered B: Did your bladder disease change (better/worse) since its onset? When you think of your symptoms in the first year of the disease? Much better Slightly better Unchanged Slightly worse Much worse Not answered

27 (19) 19 (13) 47 (32) 22 (15) 16 (11) 15 (10)

55 (37.5) 26 (18) 19 (13) 13 (9) 23 (15.5) 10 (7)

C: How is your present, overall quality of life when you compare it to the first year with your bladder disease? Much better Slightly better Unchanged Slightly worse Much worse Not answered

53 (36) 26 (18) 22 (15) 10 (7) 17 (12) 18 (12)

D: Has your present treatment changed your overall quality of life? Much better Slightly better Unchanged Slightly worse Much worse Not answered

34 (23) 19 (13) 36 (25) 9 (6) 4 (3) 44 (30)

E: How would you feel if you should live with your bladder disease as it is now, for the rest of your life? Satisfied Mostly satisfied Neither satisfied nor unsatisfied Mostly unsatisfied Unsatisfied Not answered

27 (18.5) 22 (15) 22 (15) 27 (18.5) 34 (23) 14 (10)

37

In Paper II, the clinical characteristics of 349 patients with BPS seen at Herlev Hospital were evaluated (Table 5). Follow-up period at the hospital was ≥2 years in 64% and >5 years in 34% of the patients (range 0-38 years). At the primary evaluation before treatment at Herlev Hospital, all had pain, pressure and/or discomfort perceived to be related to the urinary bladder, and almost 75% presented with nocturia ≥2x (range 0–22x). At cystoscopy under general anaesthesia, 42% had a bladder capacity of ≤500 mL (range 50–1500 mL); glomerulations in the urothelium were seen in two-thirds of patients, and a Hunner’s lesion was found in 8%. Detrusor mastocytosis was present in 53%, and detrusor fibrosis in 50%. Patients with pain (equivalent to all in the database) were evaluated by a clinical severity score of 0–5 points described; 61% scored 3-5 points. Furthermore, characterization according to four categories of applied treatment intensity/regimens was performed, 60% receiving multiple treatment regimens (categories 3+4; Table 5).

Associations between quality of life, clinical variables and treatment intensity In Paper I, quality of life worsened with increasing severity of symptoms and problems (ICSI and ICPI), demonstrated by significant differences in all SF-36 domains (p≤0.03). Severe pain perceived during past month was associated with severe overall symptoms, related problems and impaired quality of life (p≤0.01). Other clinical variables (nocturnal frequency, reduced bladder capacity, presence of mucosal glomerulations, detrusor mastocytosis, detrusor fibrosis) did not have significant correlation with OSPI and SF-36 scores. Treatment intensity categories were dichotomised (grouping categories 1+2 and 3+4), and differences were seen between these two groups regarding ICSI (p=0.012), ICPI (p=0.004) and all SF-36 domains (p≤0.028) except mental health (p=0.15), patients undergoing multiple treatment regimens scoring clearly worse. Patients with short (maximum 4 years, N=71) or long hospital follow-up (minimum 5 years, N=75) did not differ regarding symptom severity or quality of life. However, patients with long follow-up were more likely to suffer from bodily pain, but scored better in the domains mental health, role-emotional and role-physical. Retrospective clinical data from 349 patients with BPS were analysed in Paper II in order to identify clinical variables with prognostic value for outcome in terms of treatment intensity (Table 6). Multivariate analysis revealed positive associations between treatment intensity, and nocturnal frequency (p=0.043), detrusor mastocytosis (p500 ml No Yes 0-27 cells pr. mm2 28-50 cells pr. mm2 >50 cells pr. mm2 No Yes 5 yrs (1) Cystoscopy with hydrodistension and deep bladder biopsies (2) as (1) + one other treatment (3) as (1) + several treatment courses (4) High dose steroids, surgery, referral to pain clinic

Percent 22.7 32.1 28.9 16.3 92.5 7.5 25.5 30.1 44.4 13.8 14.3 14.3 57.6 34.4 65.6 47.3 27.8 24.9 50.1 49.9 35.8 30.1 34.1

N 79 112 101 57 323 26 89 105 155 48 50 50 201 120 229 165 97 87 175 174 125 105 119

19.5

68

20.3

71

34.1

119

26.1

91

Detrusor fibrosis and detrusor mastocytosis were interrelated; 58% of the patients with mast cell numbers of 28–50/sq.mm had detrusor fibrosis, having an odds ratio of 3.2, increasing to 9.1 with mastocytosis of >50/sq.mm. Detrusor fibrosis did not correlate with bladder capacity.

39

Utility of the ESSIC classification All patients with BPS included in the present PhD thesis fulfilled the diagnostic ESSIC criteria and were classified into ESSIC types (8). Patients with pathological changes in biopsies (ESSIC type C, 65% of 146 patients) and patients without pathological changes (ESSIC types A and B) were compared in regard to symptom severity and quality of life, showing no differences (Paper I). Patients with pathological changes in biopsies (ESSIC type C, 67% of 349 patients) were treated more intensively with multiple regimens (treatment intensity groups 3+4) compared to patients without pathological changes (ESSIC types A and B) (Chi Squared test p