Microscopic colitis (MC) is an inflammatory disorder of

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2011;9: Prednisolone and Budesonide for Short- and Long-Term Treatment of Microscopic Colitis: Systematic Review and Meta-analysis MICHAEL J. STEWART,* CYNTHIA H. SEOW,*, and MARTIN A. STORR*,, *Department of Medicine and Division of Gastroenterology, University of Calgary, Calgary, Alberta, Canada; and Department of Medicine, University of Munich, Munich, Germany BACKGROUND & AIMS: The incidence of microscopic colitis and its disease burden are increasing, yet there is limited systematic information addressing the use of conventional corticosteroids and budesonide in microscopic colitis. We performed a systematic review and meta-analysis on the short- and long-term efficacy of corticosteroids in treatment of microscopic colitis. METHODS: Randomized controlled trials that met predetermined selection criteria were included. Articles were identified through MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, proceedings of major gastroenterology meetings, and reference lists of trials and review articles. RESULTS: Eight randomized trials were identified. A total of 248 patients were randomized to corticosteroid versus. The intervention was budesonide in 7 trials and prednisolone in 1 trial. Budesonide was significantly more effective than for short-term clinical response (risk ratio [RR], 3.07; 95% confidence interval [CI], ) and long-term clinical response (RR, 3.22; 95% CI, ). Prednisolone was not superior to for short-term clinical response (RR, 2.00; 95% CI, ). Histologic improvement was seen with both short- and long-term budesonide (RR, 3.76; 95% CI, , and RR, 2.50; 95% CI, , respectively). Symptom relapse occurred in 46% 80% of patients within 6 months of treatment cessation. Withdrawal because of adverse effects occurred in 4.4% of patients, with no difference between study groups (P.55). CONCLUSIONS: Both short- and long-term treatment with budesonide is effective and welltolerated for microscopic colitis. However, the rate of symptom relapse once budesonide is discontinued is high. Further studies are needed to determine optimal treatment duration, dose, and withdrawal procedure. Keywords: Microscopic Colitis; Collagenous Colitis; Lymphocytic Colitis; Corticosteroids; Meta-analysis. Microscopic colitis (MC) is an inflammatory disorder of the colon that causes watery, nonbloody diarrhea. Once thought to be rare, MC is now recognized as a common cause of chronic diarrhea, especially in older women. 1 3 Microscopic colitis encompasses 2 entities that are clinically indistinguishable yet histologically distinct, collagenous colitis (CC) and lymphocytic colitis (LC). 4 Patients with MC typically have an endoscopically and radiographically normal colon in the presence of persistent diarrhea with or without other clinical features. Diagnosis is based on colon biopsy demonstrating distinctive histologic findings of either LC or CC. The etiology of MC is unclear, but an association with Hashimoto thyroiditis, celiac disease, and diabetes mellitus suggests an autoimmune diathesis. 5,6 Microscopic colitis also appears to have a stronger association with exogenous factors such as exposure to infectious agents and various medications than the other inflammatory bowel diseases, ulcerative colitis and Crohn s disease The incidence of MC has been increasing in recent years, driven largely by increased diagnosis of LC with relatively stable diagnosis of CC. 11 In 2003, the incidence of MC in a large Canadian center was estimated to be 10.0 cases per 100,000, approaching that of ulcerative colitis and Crohn s disease reported in the same population (11.0 and 16.5 cases per 100,000 population, respectively). 3,12 This degree of disease burden provides some context for understanding the importance of an effective therapeutic approach. However, because of a paucity of high-quality evidence, the management of MC usually focuses on treating associated conditions (ie, celiac disease), avoiding dietary and medication precipitants, and symptomatic treatment of the diarrhea. 2 A number of pharmacologic and alternate therapies (including probiotics) have been used with varying degrees of success, but few randomized controlled trials have been performed. Small studies with budesonide have shown the most promise for treating MC, but data on other corticosteroids have been limited. 13,14 Furthermore, there have been no controlled trials evaluating commonly used immunosuppressive medications such as azathioprine or methotrexate as maintenance therapy in patients with unremitting disease. Despite the morbidity associated with long-term corticosteroid use, steroids remain the cornerstone of therapy for the acute management of many inflammatory and rheumatologic conditions. Budesonide is a potent corticosteroid with extensive first-pass metabolism that results in less systemic absorption and less systemic steroid-related adverse effects. 15 The more favorable side effect profile of budesonide, as well as the relatively low morbidity of MC compared with other inflammatory diseases, might account for the preponderance of budesonide data. However, budesonide is less efficacious than conventional steroids in the treatment of ulcerative colitis and Crohn s disease. 16,17 It is unclear whether budesonide is similarly inferior to other corticosteroids in the treatment of MC. Although the Cochrane Collaboration previously reviewed a variety of therapeutic interventions for LC and CC separately, there are no meta-analyses specifically addressing the use of Abbreviations used in this paper: CC, collagenous colitis; CI, confidence interval; LC, mcocytic colitis; MC, microscopic colitis; QoL, quality of life; RCT, randomized controlled trial; RR, risk ratio by the AGA Institute /$36.00 doi: /j.cgh

2 882 STEWART ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 10 corticosteroids in MC. Recently, 2 randomized controlled trials (RCTs) of oral budesonide for the treatment of LC have been published. 18,19 These are the first published studies of a corticosteroid as a therapeutic intervention for LC. Because of the limited data on therapeutic interventions in MC and increasing demand for therapies to treat these diseases, further evaluation of the efficacy of corticosteroids in MC is warranted. This review uses an expanded search strategy to provide a comprehensive update on the use of corticosteroids for short- and long-term treatment of MC. Objectives The primary objective of this review was to evaluate the clinical efficacy and safety of corticosteroids for the short- and long-term treatment of adult patients with MC. Methods A systematic review and meta-analysis of RCTs were performed on the use of corticosteroid medications for the short- and long-term treatment of MC in adults. This review adheres to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. 20 Types of Studies Randomized controlled trials comparing corticosteroids with for the induction and/or maintenance of clinical response were included. Studies published in abstract form were considered if sufficient data were provided to assess the quality of the study and reported outcomes. Trials published in any language were considered. Types of Participants Microscopic colitis is an umbrella term encompassing 2 clinically indistinguishable but histologically distinct conditions, CC and LC. Both cause nonbloody, watery diarrhea, typically with an insidious onset and a chronic relapsing course. The diagnosis of MC is based on endoscopic biopsies with characteristic histopathologic features. Studies of patients with histologically confirmed MC were considered for this review. Types of Interventions Randomized trials of patients in an active treatment arm receiving a conventional corticosteroid or budesonide compared with a control arm of patients receiving were considered for inclusion. Studies of both short-term and longterm treatment were considered. Short-term treatment was defined as less than 3 months of therapy, and long-term was defined as 3 months or longer. There were no exclusions on the basis of the type, dose, or duration of the corticosteroid intervention. Types of Outcome Measures Primary outcome. The primary outcome was clinical response of MC to corticosteroid therapy. Studies assessing the induction of clinical response and the maintenance of clinical response were analyzed separately. Unlike the other chronic inflammatory bowel diseases, there is no validated tool to measure MC disease activity. Thus, for this study clinical response was defined by a reduction in stool weight by 50%, reduction in bowel movement frequency by 50%, or to less than 4 bowel movements per day after 3 8 weeks of treatment for induction therapy and at least 24 weeks for maintenance therapy. The number of patients achieving a clinical response was expressed as the percentage of patients randomized. All data were analyzed on an intention-to-treat basis. Secondary outcome. The secondary outcomes documented included (1) histologic response to treatment as defined by the authors; (2) improvement in quality of life (QoL) as defined by a validated QoL tool; (3) the proportion of patients who experience clinical relapse of symptoms and the time to relapse; (4) the adverse events reported; and (5) the number of patients who withdrew from the study because of adverse events. Search Strategy A comprehensive literature search was undertaken to identify all relevant citations. The electronic search strategy involved keyword searches of MEDLINE, EMBASE, and the Cochrane Central Register of Controlled Trials. The electronic search was supplemented by manually reviewing the reference lists of included studies as well as relevant review articles and the proceedings of major gastroenterology meetings. Search sources. Electronic database search included MEDLINE ( week 46); EMBASE (1950 November 2010); Cochrane Central Register of Controlled Trials ( ); and ongoing trials that were identified from the registry A manual search was conducted of reference lists. Proceedings from major meetings in gastroenterology were manually searched. This included the annual meetings of the American Gastroenterology Association ( ), the British Society of Gastroenterology ( ), and the United European Gastroenterology Week ( ). Personal contacts and representatives of pharmaceutical companies involved with budesonide marketing were also contacted. Search terms. The electronic database search of MEDLINE and EMBASE used the following search strategy: (1) microscopic colitis.mp. (803 articles); (2) collagenous colitis.mp. (1523 articles); (3) lymphocytic colitis.mp. (741 articles); (4) corticosteroid.mp. (196,047 articles); (5) glucocorticoid.mp. (104,446 articles); (6) steroid.mp. (290,944 articles); (7) hydrocortisone.mp. (147,364 articles); (8) prednisone.mp. (142,246 articles); (9) prednisolone.mp. (113,917 articles); (10) methylprednisolone.mp. (75,745); (11) dexamethasone.mp. (143,854 articles); (12) budesonide.mp. (16,529 articles); (13) 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 (936,914 articles); (14) 1 or 2 or 3 (2247 articles); (15) 13 and 14 (535 articles including duplicates; 380 excluding duplicates). Data Collection and Analysis Study selection. All of the citations identified through the search strategy were reviewed for possible inclusion in this review. The authors (M.J.S. and M.A.S.) independently assessed the full text articles of all potentially relevant trials. Articles that were published in abstract form only were considered if sufficient details were provided to assess the methodology and reported outcomes. Any disagreement was resolved through discussion and consensus. Data collection. Each study that met the predetermined eligibility criteria was reviewed in detail. Study information was first abstracted by one author (M.J.S.) and then checked by a

3 October 2011 CORTICOSTEROID THERAPY IN MICROSCOPIC COLITIS 883 second author (M.A.S.). The data collected included the following: (1) identifying information including the name of the first author, year of publication, and full title; (2) methodological information including details of randomization, allocation concealment, and description of study withdrawals; (3) participant information including description of patient cohort, the number of patients randomized, and the baseline characteristics of participants; (4) inclusion and exclusion criteria; (5) details of the intervention including medication, dose, route of administration, and duration of treatment; (6) details of the control/ arm; (7) primary outcome including description of clinical response and the number of patients in each treatment arm achieving clinical response; (8) secondary outcomes including details of histologic response, improvement in quality of life, and relapse rates; and (9) adverse effects and the number of patients who withdrew because of adverse effects. Assessment of study quality. Quality assessment was performed by using the Jadad Scale, a commonly used tool for assessing the methodological quality of randomized trials. 21 The scale considers randomization, blinding, and description of participant withdrawal and scores studies on a 5-point scale, with scores 3 considered to be high-quality trials. Only studies with Jadad scores 3 were included. The Jadad Scale consists of the following: (1) Was the study described as randomized? (2) Was the method of randomization described and appropriate? (3) Was the study described as double blind? (4) Was the method of double blinding described and appropriate? (5) Was there a description of withdrawals and dropouts? Each affirmative response is awarded 1 point. If either randomization or blinding was described but determined to be inappropriate, then 1 point was lost for each. All studies were independently scored by M.J.S. and M.A.S. Discrepancies in study scores were resolved through discussion and consensus. Statistical analysis. Binary data from the included studies were pooled, and Mantel Haenszel estimates were calculated on an intention-to-treat basis. Studies were weighted by using the DerSimonian and Laird method. Pooled effects were calculated for the primary outcome of clinical response to either short-term induction therapy or maintenance therapy, as well as for the secondary outcome of histologic response. Data were reported as risk ratios (RRs) with 95% confidence intervals (95% CIs). Risk ratios were favored because their interpretation is more intuitive, and they have less tendency than odds ratios to overstate the magnitude of effect size. 22,23 The studies were independently assessed for clinical and methodological heterogeneity. The I 2 measure was calculated to quantify inconsistency and aided in the decision of whether to pool studies. 24 A fixed-effect model for all analyses was used unless significant heterogeneity was identified (I 2 50%), in which case a random-effect model was used. Subgroup analysis was performed for each disease entity (CC and LC), type of corticosteroid intervention, and treatment duration (short- vs long-term treatment). Categorical data were converted into individual 2 2 tables and reported as relative risk and 95% CI. For continuous variables, the results were presented as the weighted mean difference and 95% CI or the standardized mean difference when the outcomes were not comparable. Analyses were performed by using Review Manager (RevMan) Version 5.0 (The Nordic Cochrane Centre, the Cochrane Collaboration, Copenhagen, 2008) Results Thirty-two citations were identified for possible inclusion in the analysis and were reviewed in detail. We subsequently excluded 23 articles for the following reasons: 9 were review articles, 8 were nonrandomized trials, and 6 were case reports. The quantitative analysis included 8 trials. A ninth study reported QoL data from 1 of the 8 included trials and was therefore included for qualitative analysis only. Figure 1 details the flow of study selection. The details of trials included in the meta-analysis are provided in Table 1. All trials were published in English language between 2002 and 2010 and reported similar participant populations as well as inclusion and exclusion criteria. Budesonide was the active intervention in 7 trials. Prednisolone was the active intervention in 1 trial that was designed to be a maintenance trial of immunosuppressive therapy but was discontinued early when many participants failed to achieve remission. 28 This study was reported as an ostensibly negative induction trial; however, post hoc analysis of the results demonstrated a 3-fold increase in the clinical response rate within the treatment arm when the same clinical criteria as other MC induction trials were applied. 18,29,30 The discrepant outcome resulted from the original trial by using a very strict definition of clinical remission rather than the more generally used criteria for clinical response. Six of the 8 included trials were short-term induction trials, with the duration of the active treatment ranging from 3 8 weeks. 18,19,28 31 Two of these studies included an open-label crossover phase on completion of the double-blinded induction phase. Only data from the induction phase were included in the quantitative analysis. A total of 168 patients with active MC were randomized to short-term treatment with a corticosteroid versus. Of these, 74% (n 125) were female, with age ranging from years. The 2 maintenance trials had a 6-week open-label induction phase, followed by a 6-month double-blind maintenance phase. 32,33 Both of these studies evaluated 6 mg of budesonide daily versus in patients with CC. Eighty participants were randomized to long-term treatment with budesonide versus. Of these, 76% (n 61) were female, with age ranging from years. Clinical Response The criteria for clinical response included a reduction in stool weight by 50%, reduction in bowel movement frequency by 50%, or to less than 4 bowel movements per day after 3 8 weeks for induction therapy and at least 24 weeks for maintenance therapy. Six induction trials were suitable for pooled quantitative analysis (Table 2). In total, 82 patients with active MC were randomized to short-term treatment with budesonide, and 9 were randomized to short-term treatment with prednisolone. Seventy-seven patients were randomized to a control group. A significant clinical response was seen in 66 of the 82 (80%) participants treated with budesonide and 19 of the 74 (26%) patients in the corresponding control arm (Figure 2). The number of patients achieving clinical response with shortterm budesonide therapy was significantly greater than with (RR, 3.07; 95% CI, ). The proportion of patients with CC who responded to short-term budesonide appeared to be twice that seen in LC (Figure 3A); however, this difference was not statistically significant (P.2594).

4 884 STEWART ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 10 Figure 1. Selection of articles for inclusion in meta-analysis. In the only prednisolone trial, 9 patients were randomized to short-term prednisolone, and 3 patients were randomized to. Although there was a trend toward increased response in the prednisolone arm, it was not found to be statistically superior to (RR, 2.00; 95% CI, ; Figure 2). Two trials of budesonide versus for maintenance of clinical response in patients with CC were suitable for pooled analysis (Table 2). In total, 40 patients were randomized to each of the budesonide and groups. Long-term treatment with budesonide was demonstrated to be superior to for maintenance of clinical response, with a 3-fold increase in the proportion of patients who maintained clinical response (RR, 3.22; 95% CI, ; Figure 3B). Histologic Response Six of the 7 budesonide trials obtained colon biopsies at baseline and at the end of treatment. Histologic improvement was assessed by evaluating the degree of inflammation, the thickness of the collagen layer (in CC), and the number of intraepithelial lymphocytes (in LC) (Table 2). Three trials of short-term budesonide and 2 trials of long-term budesonide reported binary data that were pooled for quantitative analysis (Figure 3C, D). The histologic response to short-term budesonide therapy was evaluated in a total of 121 patients. The proportion of patients treated with short-term budesonide demonstrating significant histologic improvement (as defined by each trial) was nearly 4-fold that of the group (RR, 3.76; 95% CI, ; Figure 3C). In 2009 Bonderup et al 32 demonstrated significant improvements in both the collagen layer thickness and the degree of inflammation with 6 months of budesonide treatment (P.01 for both), with no significant improvement seen in the arm. In 2008 Miehlke et al 33 demonstrated significant improvement in collagen layer thickness and degree of inflammation within both the budesonide and treatment groups but failed to demonstrate a significant difference in the rate of histologic responders between the budesonide and groups according to study protocol analysis (P.10). The analysis was likely confounded because only half of the participants completed both biopsies. Pooled analysis of these trials confirmed that long-term treatment with budesonide was superior to for maintenance of histologic response (RR, 2.50; 95% CI, ; Figure 3D). Relapse of Symptoms There were insufficient data to calculate the mean time to symptom relapse. Bonderup et al 30 demonstrated a clinical response in all 10 CC patients who were treated with short-term budesonide. On cessation of therapy, 8 of the 10 (80%) experienced recurrence of their symptoms within 8 weeks. Similarly, Baert et al 29 found that 63% patients with CC required further treatment for at least 1 additional month beyond the blinded 8-week budesonide trial because of clinical relapse. Miehlke et al 19 studied short-term budesonide therapy in patients with LC. A total of 26 patients achieved a clinical response with budesonide, 18 during the initial blinded study and 8 treated with open-label budesonide during the crossover portion of the study. All patients who achieved a clinical response were followed for a mean period of 14 months after study completion. Of these 26 patients, 12 (46%) experienced relapse in their symptoms within a mean period of 2 months after cessation of short-term therapy. Maintenance budesonide therapy had similar relapse rates. Bonderup et al 32 followed patients after cessation of long-term treatment and found that 24 weeks of budesonide resulted in sustained clinical response in 13 of 17 patients. However, with

5 Table 1. Characteristics and Quality Assessment of Included RCTs of Corticosteroids to Treat MC Author, year (reference) Jadad score n Dropout, n(%) Female, n (%) Age, y (mean) Induction studies Pardi, 2009 (18) 4 15 LC 0 (0) 12 (80) 59.7 NR Budesonide 9 mg/d vs Miehlke, 2009 (19) Bonderup, 2003 (30) Munck, 2003 (28) Miehlke, 2002 (31) 5 42 LC 4 (10) 28 (67) 61 a Budensonide 9 mg/d vs 5 20 CC 0 (0) 16 (76) 54 a Budesonide 9 mg/d 4 wk, 6 mg/d 2 wk, 3 mg/d 2wk vs 5 12 CC 0 (0) 10 (83) 63 a Prednisolone 50 mg/d 2 wk, 37.5 mg/d 1wk 5 51 CC 6 (12) 39 (76) 60 a Budesonide 9 mg/d vs Baert, 2002 (29) 5 28 CC 3 (11) 20 (71) 56 NR Budesonide 9 mg/d vs Maintenance studies Bonderup, 2009 (32) Miehlke, 2008 (33) 5 34 CC 2 (6) 27 (79) Budesonide 6 mg/d vs 5 46 CC 4 (9) (72 76) Budesonide 6 mg/d vs BM, bowel movement; NR, not reported; NSAID, nonsteroidal anti-inflammatory drug. a Reported only as median age. Age, y (range) Treatment Duration Inclusion criteria Exclusion criteria 8 wk Adults with histologically confirmed LC 6 wk Adults with histologically confirmed LC and 4 BMs/d within7dand negative stool cultures 8 wk Adults with histologically confirmed CC and 5 BMs/d or 200 g/d, and negative stool cultures 3 wk Adults with histologically confirmed MC and 3 mo with diarrhea and a stool volume 350 g/ d, or 200 g/d and a stool frequency 5/d 6 wk Adults (18 80 y) with histologically confirmed CC and 5 BMs/d and negative stool cultures 8 wk Patients with histologically confirmed CC with 2 mo of 3 loose BMs/ d, and negative stool cultures 24 wk Adults with histologically confirmed CC and 4 BMs/d for 3 consecutive d and negative stool cultures 26 wk Adults with histologically confirmed CC and 4 BMs/d on 4 of 7 previous d and at least 4 wk of symptoms NR Treatment with budesonide, salicylates, mesalamine, budesonide, NSAIDs, antibiotics, or systemic steroids within 2 wk; other chronic gastrointestinal disease or colon surgery, malignancy, substance abuse, pregnancy, allergy to budesonide Treatment with mesalamine, corticosteroids, or azathioprine within 3 mo; other chronic gastrointestinal disease Treatment with immunosuppressives, NSAIDs, antiulcer medication, antacids, and antibiotics; infectious cause of diarrhea, celiac disease, or other gastrointestinal diseases or previous gastrointestinal surgery (except cholecystectomy) Treatment with budesonide, salicylates, steroids, prokinetics, antibiotics, ketoconazole, or NSAIDs within 4 wk; pregnancy; other gastrointestinal disease No antidiarrheal medications for 1 wk; no mesalamine for 2 wk; no corticosteroids for 4 wk; no immunosuppressive for 3 mo; other significant gastrointestinal disease Treatment with budesonide, salazopyrine, mesalamine, or systemic corticosteroids within 3 mo; other chronic gastrointestinal disease or ostomy; clinically apparent liver of kidney disease Treatment with budesonide, salicylates, mesalamine, NSAIDs, antibiotics, or systemic corticosteroids within 2 wk; other chronic gastrointestinal disease or colon surgery, malignancy, infectious cause of diarrhea, substance abuse, pregnancy, allergy to budesonide October 2011 CORTICOSTEROID THERAPY IN MICROSCOPIC COLITIS 885

6 Table 2. Descriptions of Primary and Secondary Outcomes With the Number of Patients Achieving Clinical and Histologic Response in the Intervention and Placebo Groups, Intention-to-Treat Values Author, year (reference) Induction studies Pardi, 2009 (18) Miehlke, 2009 (19) Bonderup, 2003 (30) Munck, 2003 (28) Miehlke, 2002 (31) Baert, 2002 (29) Clinical response, description 50% reduction in BM frequency in 3 of last 4 wk of study 3 BMs/d and a reduction by 1 BMs/d 50% reduction in BM frequency or stool weight 50% reduction in BM frequency or stool weight 3 BMs/d during last wk of study 50% reduction in BM frequency or stool weight Intervention, n(%) Control, n(%) Histologic response, description Intervention, n(%) Control, n(%) Secondary end points, description 10/11 (91) 1/4 (25.0) NR NR NR Stool consistency, urgency, abdominal discomfort 18/21 (86) 10/21 (48) Reduction of IEL by 20 or by 50% 13/21 (62) 4/21 (19) Stool consistency and/or reduction in lamina propria inflammation a 10/10 (100) 2/10 (20) Significant reduction in thickness NR NR Stool weight reduction of collagen band and degree of inflammation a 6/9 (66) 1/3 (33) NR NR NR Complete remission defined as a stool weight 200 g/d or stool frequency 2/d 20/26 (77) 3/25 (12) Improvement of 2 of 3 parameters; 14/26 (23) 1/26 (4) Stool consistency inflammation reduced by 1 point, a thickness of collagen band 10 m or 50% reduction, degeneration of surface epithelium 8/14 (57) 3/14 (21) Significant reduction in lamina 13/14 (93) 4/14 (29) Stool consistency, abdominal propria inflammation and/or discomfort, general well-being, time thickness of collagen band to remission Maintenance studies Bonderup, 2009 (32) 3 BMs/d 13/17 (77) 2/17 (12) Significant reduction in thickness of collagen band and degree of inflammation Miehlke, 2008 (33) 3 BMs/d 17/23 (61) 8/23 (30) Significant reduction in thickness of collagen band and/or lamina propria inflammation a BM, bowel movement; IEL, intraepithelial lymphocytes per 100 epithelial cells; NR, not reported. a Based on semiquantitative 4-point scale: no inflammation, mild, moderate, severe. 6/17 (35) 3/17 (18) Time to relapse 14/23 (61) 5/23 (22) NR 886 STEWART ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 10

7 October 2011 CORTICOSTEROID THERAPY IN MICROSCOPIC COLITIS 887 Figure 2. Clinical response in MC patients treated with short-term corticosteroid therapy vs, by corticosteroid. M H, Mantel Haenszel. an additional 24-week follow-up after cessation of treatment, 9 of these 13 responders (69%) had experienced relapse in symptoms. Quality of Life Quality of life data were reported in 2 of the trials included in the quantitative analysis and the 1 additional outcome study. All 3 trials reported improved QoL with budesonide therapy, but because of the varied nature of assessments no pooled quantitative synthesis was possible. 19,33,34 Madisch et al 34 evaluated short-term budesonide treatment by using the Gastrointestinal Quality of Life Index. After 6 weeks of treatment the budesonide group demonstrated significant improvement in mean Gastrointestinal Quality of Life Index scores (P.001), with no significant improvement in the group (P.05). The mean scores for the dimensions of symptoms, emotional functioning, and physical functioning showed significantly greater improvement in the budesonide group compared with the group (P.05). Miehlke et al 33 assessed QoL by using the Short Form-36 Health Survey. Both budesonide and groups demonstrated significant improvements in the domains of physical functioning and bodily pain (P.05); however, the budesonide group had additional improvement in the domain vitality (P.0034). There were no statistical differences within the remaining physical and mental domains or between treatment groups. The effect of short-term budesonide on QoL appeared to persist with long-term maintenance therapy. Miehlke et al 33 assessed QoL by using the Short Inflammatory Bowel Disease Questionnaire and the Short Form-36 and demonstrated improvement on both with induction budesonide therapy, with scores nearing those observed in normal subjects. These scores were maintained throughout 6 months of budesonide treatment (data not provided). No between-group analysis was provided. Study Withdrawals and Adverse Effects Of the 248 participants randomized to corticosteroid or a, 19 failed to complete the trial. The overall withdrawal rate was 7.7% (range, 0% 11.8%). The adverse effect of study medication was cited as the reason for 11 of the 19 study withdrawals. There was no significant difference in withdrawals because of adverse effects between the intervention and control groups (7 of 131 in treatment group and 4 of 117 in group; P.547). Adverse events in each study group were fully reported in 4 trials and are documented in Table 3. There were no serious adverse events reported in the remaining 4 trials. Discussion Microscopic colitis is an inflammatory disorder of the colon that leads to significant morbidity, especially in older women. The incidence of MC is growing, yet the therapeutic options are based largely on uncontrolled studies and focus mainly on symptomatic management of diarrhea, while avoiding dietary and pharmacologic triggers. Lymphocytic colitis and CC are clinically indistinguishable, and the approach to treatment is generally the same. The inflammatory nature of MC presumes that potent anti-inflammatory medications used to treat other inflammatory bowel diseases might have a role in the management of MC. This review identified 8 high-quality RCTs that examined the use of corticosteroids in MC. Nearly all of the trials evaluated budesonide in CC or LC compared with. A total of 236 patients were randomized to budesonide versus, whereas only 12 patients were randomized to prednisolone versus. This meta-analysis confirms that budesonide is both welltolerated and highly effective for short- and long-term treatment of MC. Moreover, the therapeutic effect of short-term budesonide appears equivalent for both LC and CC. Eighty percent of patients with MC achieved a clinical response when treated with a 6- to 8-week course of budesonide, compared with only 26% of patients treated with. This was highly statistically significant, with the budesonide group being 3 times more likely to achieve clinical response compared with the group. The major advantage of budesonide compared with conventional corticosteroids is that its limited systemic absorption leads to improved long-term tolerability. The only studies on the long-term use of budesonide in MC were conducted in patients with CC. Budesonide was demonstrated to be 3 times as effective as for maintaining a clinical response. Long-

8 888 STEWART ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 10 Figure 3. Short- and long-term budesonide vs. (A) Clinical response in MC patients treated with short-term budesonide, by subtype. (B) Clinical response in LC patients treated with long-term budesonide. (C) Histologic response in MC patients treated with short-term budesonide, by subtype. (D) Histologic response in LC patients treated with long-term budesonide. M H, Mantel Haenszel.

9 October 2011 CORTICOSTEROID THERAPY IN MICROSCOPIC COLITIS 889 Table 3. Adverse Effects Reported by Intervention and Control Groups Budesonide side effect Budesonide, n 66 Control, n 65 Arthralgia, myalgia, abdominal pain, 4 7 leg cramps Depression/mood changes 1 1 Diaphoresis 1 1 Dizziness 3 0 Fatigue 0 1 Gastrointestinal (nausea, dyspepsia, 8 6 bloating, flatulence, ulcer) Headache 2 1 Hypertension 1 0 Infection (urinary, respiratory, dental) 2 2 Motor vehicle accident 0 1 Obstipation 1 0 Skin erythema or rash 1 2 Sleep disturbance 1 0 Subarachnoid hemorrhage 1 0 Venous thrombosis 0 1 Bruising 1 1 Weight gain 4 0 Worsening of diabetes 2 0 Total Prednisolone side effect 28 Prednisolone, n 9 Control, n 3 Arthralgia, myalgia, abdominal 3 1 pain, leg cramps Depression/mood changes 4 1 Headache 5 1 Sleep disturbance 8 0 Weight gain 5 1 Total 25 4 term budesonide was well-tolerated, with no difference in adverse effect rates between treatment and groups and no difference in the number of study withdrawals caused by adverse effects. The histologic response to short- and long-term budesonide treatment paralleled the clinical response. Significant histologic improvement, as defined by the authors, was seen in both CC and LC treated with short-term budesonide as well as in LC treated with long-term budesonide. None of the trials evaluated patients for complete histologic remission but instead assessed the biopsies of patients for significant reduction in the degree of colon inflammation. Thus, although it resulted in significant histologic improvement, it is unclear how effective budesonide is at achieving complete histologic remission in MC. The current evidence similarly cannot establish whether the histologic response to budesonide is sustained once therapy is discontinued. Three of the studies included in this review followed patients after budesonide was discontinued. These studies established that although budesonide was highly effective at inducing a clinical response, discontinuing its use resulted in a high rate of symptom relapse. This was true for both CC and LC and was independent of the duration of treatment before budesonide was stopped. The overall rate of relapse remains uncertain; however, between 46% and 80% of patients in the current study had symptom relapse within 6 months of discontinuing budesonide. The focus on budesonide as the corticosteroid of choice for MC is likely due to its significantly less toxic side effect profile and because MC generally remains a nuisance condition rather than a life-threatening disease. It might also be influenced by the 1 prednisolone trial failing to demonstrate a statistically significant effect versus. Unfortunately, this trial was marred by its small size, with only 3 patients randomized to the control arm. Given the clinical and histologic response of MC to budesonide, it is reasonable to presume that prednisolone and other conventional corticosteroids have a similar antiinflammatory effect. The current evidence supports the use of budesonide for treatment of both CC and LC. At a dose of 9 mg daily for 6 8 weeks, budesonide can induce a symptomatic response as well as a significant reduction in colonic inflammation. These improvements are at least sustained by a maintenance dose of budesonide 6 mg daily when assessed at 6 months. Both shortand long-term treatment with budesonide is well-tolerated. Unfortunately, the rate of symptom relapse is high once budesonide is discontinued. Further study is required to determine the optimal duration of budesonide therapy to achieve the maximal sustained effect and the best approach to budesonide withdrawal or taper to minimize symptom relapse. It might also be interesting to see whether alternating courses of budesonide would control symptoms and whether the addition of other medical agents used to treat MC could have an additive effect. References 1. Nyhlin N, Bohr J, Eriksson S, et al. Systematic review: microscopic colitis. Aliment Pharmacol Ther 2006;23: Williams JJ, Beck PL, Andrews CN, et al. Microscopic colitis: a common cause of diarrhoea in older adults. Age Ageing 2010; 39: Williams JJ, Kaplan GG, Makhija S, et al. Microscopic colitisdefining incidence rates and risk factors: a population-based study. Clin Gastroenterol Hepatol 2008;6: Bohr J, Olesen M, Tysk C, et al. Collagenous and lymphocytic colitis: a clinical and histopathological review. Can J Gastroenterol 2000;14: Cindoruk M, Tuncer C, Dursun A, et al. Increased colonic intraepithelial lymphocytes in patients with Hashimoto s thyroiditis. J Clin Gastroenterol 2002;34: Olesen M, Eriksson S, Bohr J, et al. Lymphocytic colitis: a retrospective clinical study of 199 Swedish patients. Gut 2004;53: Ballinger A. Adverse effects of nonsteroidal anti-inflammatory drugs on the colon. Curr Gastroenterol Rep 2008;10: Hilmer SN, Heap TR, Eckstein RP, et al. Microscopic colitis associated with exposure to lansoprazole. Med J Aust 2006;184: Mäkinen M, Niemelä S, Lehtola J, et al. Collagenous colitis and Yersinia enterocolitica infection. Dig Dis Sci 1998;43: Wilcox GM, Mattia AR. Microscopic colitis associated with omeprazole and esomeprazole exposure. J Clin Gastroenterol 2009;43: Pardi DS, Loftus EV Jr, Smyrk TC, et al. The epidemiology of microscopic colitis: a population based study in Olmsted County, Minnesota. Gut 2007;56: Bernstein CN, Wajda A, Svenson LW, et al. The epidemiology of inflammatory bowel disease in Canada: a population-based study. Am J Gastroenterol 2006;101:

10 890 STEWART ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No Chande N, McDonald JW, Macdonald JK. Interventions for treating lymphocytic colitis. Cochrane Database Syst Rev 2008: CD Chande N, McDonald JW, Macdonald JK. Interventions for treating collagenous colitis. Cochrane Database Syst Rev 2008: CD Spencer CM, McTavish D. Budesonide: a review of its pharmacological properties and therapeutic efficacy in inflammatory bowel disease. Drugs 1995;50: Sherlock ME, Seow CH, Steinhart AH, et al. Oral budesonide for induction of remission in ulcerative colitis. Cochrane Database Syst Rev 2010:CD Seow CH, Benchimol EI, Griffiths AM, et al. Budesonide for induction of remission in Crohn s disease. Cochrane Database Syst Rev 2008:CD Pardi DS, Loftus EV, Tremaine WJ, et al. A randomized, doubleblind, -controlled trial of budesonide for the treatment of active lymphocytic colitis. Gastroenterology 2009;136:A519 A Miehlke S, Madisch A, Karimi D, et al. Budesonide is effective in treating lymphocytic colitis: a randomized double-blind controlled study. Gastroenterology 2009;136: Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 2009;151: , W Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17: Altman DG, Deeks JJ, Sackett DL. Odds ratios should be avoided when events are common. BMJ 1998;317: Schwartz LM, Woloshin S, Welch HG. Misunderstandings about the effects of race and sex on physicians referrals for cardiac catheterization. N Engl J Med 1999;341: , discussion Higgins JP, Thompson SG. Quantifying heterogeneity in a metaanalysis. Stat Med 2002;21: Statistical guidelines. Ann Intern Med Attig RB, Clabaugh A. Clinical trials and statistical tribulations. Appl Clin Trials 2008;3: Clinical evidence BMJ group. BMJ Available at: clinicalevidence.bmj.com. 28. Munck LK, Kjeldsen J, Philipsen E, et al. Incomplete remission with short-term prednisolone treatment in collagenous colitis: a randomized study. Scand J Gastroenterol 2003;38: Baert F, Schmit A, D Haens G, et al. Budesonide in collagenous colitis: a double-blind -controlled trial with histologic follow-up. Gastroenterology 2002;122: Bonderup OK, Hansen JB, Birket-Smith L, et al. Budesonide treatment of collagenous colitis: a randomised, double blind, controlled trial with morphometric analysis. Gut 2003; 52: Miehlke S, Heymer P, Bethke B, et al. Budesonide treatment for collagenous colitis: a randomized, double-blind, -controlled, multicenter trial. Gastroenterology 2002;123: Bonderup OK, Hansen JB, Teglbjaerg PS, et al. Long-term budesonide treatment of collagenous colitis: a randomised, doubleblind, -controlled trial. Gut 2009;58: Miehlke S, Madisch A, Bethke B, et al. Oral budesonide for maintenance treatment of collagenous colitis: a randomized, double-blind, -controlled trial. Gastroenterology 2008; 135: Madisch A, Heymer P, Voss C, et al. Oral budesonide therapy improves quality of life in patients with collagenous colitis. Int J Colorectal Dis 2005;20: Reprint requests Address requests for reprints to: Martin Storr, MD, Department of Medicine, University of Munich, Marchioninistr. 15, Munich, Germany. gidoc@gmx.com; tel: Conflicts of interest The authors disclose no conflicts.