- Email: [email protected]
Combined budesonide and antibiotic therapy for active Crohn's disease: A randomized controlled trial
GASTROENTEROLOGY 2002;123:33– 40
Combined Budesonide and Antibiotic Therapy for Active Crohn’s Disease: A Randomized Controlled Trial A. HILLARY STEINHART,* BRIAN G. FEAGAN,‡ CINDY J. WONG,‡ MARGARET VANDERVOORT,‡ SHELLEY MIKOLAINIS,* KENNETH CROITORU,§ ERNEST SEIDMAN,㛳 DESMOND J. LEDDIN,¶ ALAIN BITTON,# ERIC DROUIN,㛳 ALBERT COHEN,** and GORDON R. GREENBERG* for the Crohn’s and Colitis Foundation of Canada Inflammatory Bowel Disease Network Investigators *Department of Medicine, Mount Sinai Hospital, Toronto, Ontario; ‡London Clinical Trials Research Group, The John P. Robarts Research Institute, London, Ontario; §Department of Medicine, McMaster University, Hamilton, Ontario; 㛳Gastroenterology Division, St. Justine Hospital, Montreal, Quebec; ¶Department of Medicine, Dalhousie University, Halifax, Nova Scotia; #Royal Victoria Hospital, Montreal, Quebec; and **Jewish General Hospital, Montreal, Quebec, Canada
Background & Aims: Although antibiotics are frequently used to treat Crohn’s disease, this practice is not supported by strong evidence from randomized trials. Methods: We conducted a double-blind multicenter study of patients with active Crohn’s disease of the ileum, right colon, or both. Patients were randomized to receive oral ciprofloxacin and metronidazole, both 500 mg twice daily, or placebo for 8 weeks. All patients received oral budesonide 9 mg once daily. The primary efficacy measure was the proportion of patients in remission at week 8. Results: Of the 134 patients who were randomized, 130 were evaluated for efficacy; 66 received placebo, and 64 received antibiotics. At week 8, 21 patients (33%) assigned to antibiotics were in remission as compared with 25 patients (38%) in the placebo group (P ⴝ 0.55; absolute difference, ⴚ5%; 95% confidence interval, ⴚ21% to 11%). An interaction (P ⴝ 0.025) between treatment allocation and disease location on treatment response was identified. Among patients with disease of the colon, 9 of 17 (53%) were in remission after treatment with antibiotics, compared with 4 of 16 (25%) of those who received placebo (P ⴝ 0.10). Discontinuation of therapy because of adverse events occurred in 13 of 66 (20%) patients treated with antibiotics, compared with 0 of 68 in the group who received placebo (P < 0.001). Conclusions: In patients with active Crohn’s disease of the ileum, the addition of ciprofloxacin and metronidazole to budesonide is an ineffective intervention, but this antibiotic combination may improve outcome when there is involvement of the colon.
rohn’s disease is a chronic inflammatory disorder of unknown origin that commonly affects the ileum and colon. Medical therapy for active Crohn’s disease comprises the use of nonspecific anti-inflammatory drugs, with corticosteroids providing the most consistent benefit.1–5 Recent experimental evidence derived from
C
several genetically susceptible animal models indicates that the normal enteric bacterial flora contribute to chronic pathologic inflammatory responses in the intestine.6 –11 Because these models emphasize the importance of interactions between the mucosal immune system and commensal bacteria,6,7 a sound rationale exists to evaluate treatment regimens that target both of these factors. Thus, the addition of broad-spectrum antibiotics might enhance the efficacy of anti-inflammatory drugs such as corticosteroids. Antibiotic therapy, notably metronidazole and ciprofloxacin, is frequently advocated for Crohn’s disease,12–16 but randomized controlled trials to support this practice are limited. Furthermore, no controlled trials have evaluated the efficacy of antibiotics and corticosteroids in combination. For these reasons, we assessed the use of budesonide, a glucocorticosteroid nearly as effective as prednisolone but without the steroid-associated adverse effects,3 combined with metronidazole and ciprofloxacin for the treatment of active Crohn’s disease of the ileum and of the ileum and proximal colon.
Methods Patients This randomized controlled trial was conducted at 14 university medical centers and 2 community-based gastroenterology practices in Canada. Investigators and patients were unaware of the treatment assignments. The protocol was approved by the institutional review board at each center. All patients gave written informed consent. Abbreviation used in this paper: IBDQ, Inflammatory Bowel Disease Questionnaire. © 2002 by the American Gastroenterological Association 0016-5085/02/$35.00 doi:10.1053/gast.2002.34225
34
STEINHART ET AL.
Eligible patients were 14 years of age or older, with Crohn’s disease involving the ileum, proximal colon (ascending colon to mid transverse colon), or both. Disease location was determined by either colonoscopy or air-contrast barium enema performed within 2 years of entry to the study. Patients had active disease, as defined by a score on the Crohn’s Disease Activity Index (CDAI)17 of 200 – 400 points. This index incorporates the following items: the frequency of soft stools, the severity of abdominal pain and the level of general well-being in the 7 days preceding the assessment, the presence or absence of an abdominal mass or extraintestinal manifestations of the disease, the use of opiates to treat diarrhea, hematocrit, and weight. Scores range from 0 to approximately 600; higher scores indicate greater disease activity. A score of less than 150 indicates inactive disease. Patients receiving treatment with antibiotics (within 1 week of the randomization visit), corticosteroids, parenteral nutrition, enteral feeds (within 2 weeks), or immunosuppressants (within 12 weeks) were not eligible. Patients who had received 5-aminosalicylic acid drugs or sulfasalazine were eligible if the medication was discontinued at least 7 days before the randomization visit. Patients with more than 100 cm of small intestine surgically removed or those with an ileostomy were not included.
Baseline Studies Patients were evaluated for eligibility 1 week before the randomization visit. At this time they were given a diary in which to record the data necessary to calculate the CDAI. At the randomization visit, demographic information was obtained; a physical examination, blood tests, and stool cultures and an assay for Clostridium difficile toxin were performed; and scores on the CDAI and Inflammatory Bowel Disease Questionnaire (IBDQ)18,19 were calculated. Scores on this measure range from 32 to 224; higher scores mean a better quality of life. Scores of greater than 170 are observed in patients whose disease is in remission.
Randomization The patients were randomly assigned, by means of a computer-generated schedule, to receive either oral ciprofloxacin (Cipro; Bayer, Inc., Toronto, Ontario) and metronidazole (Apo-metronidazole; Apotex, Inc., Toronto, Ontario), both 500 mg twice daily, or placebo for 8 weeks. Patients also received oral budesonide2,3 (Entocort; AstraZeneca, Mississauga, Ontario) 9 mg once daily. The randomization schedule was stratified according to center and disease location (ileum or ileum and colon) in permutated blocks of 4 or 6. The placebo tablets were identical in appearance to the ciprofloxacin and metronidazole preparations. Investigators and patients were unaware of the treatment assignment.
Other Treatments for Crohn’s Disease Patients did not use other drugs for the treatment of active Crohn’s disease, including glucocorticoids other than
GASTROENTEROLOGY Vol. 123, No. 1
budesonide, immunosuppressants, and infliximab. Total parenteral nutrition and enteral feeding were not allowed.
Follow-up Patients were assessed in the clinic 2, 4, 6, and 8 weeks after the randomization visit. At each visit, the scores on the CDAI and the IBDQ were calculated, and blood was obtained for chemical analyses. The study medication was discontinued if other treatments for active Crohn’s disease were initiated. At each visit, unused medications were returned, and a pill count was performed to evaluate patients’ adherence with drug therapy.
Measures of Outcome The primary measure of response was remission at week 8. Remission was defined as a score of less than 150 on the CDAI. Secondary measures included the mean changes in scores on the CDAI, quality of life, and adverse drug reactions.
Statistical Analysis The baseline characteristics of the patients were analyzed with descriptive statistics. For the primary analysis, we compared the proportion of patients in the 2 treatment groups who were in remission at week 8 by using Pearson’s 2 test. This analysis included all patients who provided efficacy data from at least 1 postrandomization assessment. A logistic regression analysis evaluated the potential interaction between site of disease and treatment allocation on the primary outcome. The mean scores over time on the CDAI and the IBDQ were analyzed by a repeated-measures analysis of variance. Scores from patients who withdrew from the study prematurely were analyzed by using the last-observation-carriedforward principle. Separate analyses were performed for each stratum of disease location. We used the 2 test or the Fisher exact test, as appropriate, to compare the number of patients with adverse events in each group. Statistical analyses were performed with SAS software.20 A 2-sided P value of 0.05 or less was considered to indicate statistical significance. We estimated that 55% of patients treated with budesonide alone would be in remission at week 8. A minimum clinically important absolute difference in the occurrence of this outcome was considered to be 25%. Thus, a remission rate of 80% in the combination budesonide/antibiotic group was anticipated. Randomization of 52 evaluable patients per group would give a power of 80% to detect this difference at the 0.05 significance level. A total of 134 patients were enrolled to allow for a nonevaluable rate of up to 20%.
Results Randomization Between June 1998 and September 2000, 209 patients were evaluated for eligibility. Thirty-six patients were excluded because of a CDAI score that was not in the specified range. Fourteen patients were not
July 2002
BUDESONIDE AND ANTIBIOTICS FOR CROHN’S DISEASE
35
Table 1. Baseline Characteristics of the Patients Overall and According to the Site of Disease Stratum Overall
Small bowel
Small bowel and colon
Characteristic
Budesonide/ placebo (n ⫽ 68)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 66)
Budesonide/ placebo (n ⫽ 51)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 49)
Budesonide/ placebo (n ⫽ 17)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 17)
Age (yr) Male sex, n (%) Weight (kg) Years since diagnosis Abdominal mass, n (%) Previous surgery for Crohn’s disease, n (%) Cigarette smoker, n (%) CDAI scorec IBDQ scored White blood cell count (⫻ 109/L) Hemoglobin (g/L) Platelet count (⫻ 109/L)
30 ⫾ 1 27 (40) 66 ⫾ 2 6 ⫾ 1.0 8 (12) 23 (34) 33 (49) 278 ⫾ 7 140 ⫾ 3 8 ⫾ 0.3 130 ⫾ 2 340 ⫾ 15
34 ⫾ 2 30 (45) 66 ⫾ 2 5 ⫾ 0.7 10 (15) 25 (38) 25 (38) 273 ⫾ 6 137 ⫾ 3 8 ⫾ 0.3 132 ⫾ 2 330 ⫾ 13
33 ⫾ 1 19 (37) 68 ⫾ 3 7 ⫾ 1.3 7 (14) 19 (37) 30 (59) 280 ⫾ 8 139 ⫾ 3 8 ⫾ 0.3 132 ⫾ 2 320 ⫾ 14
34 ⫾ 1 20 (41) 65 ⫾ 2 6 ⫾ 1.0 7 (14) 20 (41) 19 (39)b 281 ⫾ 8 134 ⫾ 4 8 ⫾ 0.4 133 ⫾ 3 330 ⫾ 15
22 ⫾ 2 8 (47) 59 ⫾ 3 4 ⫾ 0.9 1 (6) 4 (24) 3 (18) 273 ⫾ 15 144 ⫾ 5 9 ⫾ 0.9 125 ⫾ 4 398 ⫾ 41
32 ⫾ 4a 10 (59) 68 ⫾ 4 4 ⫾ 0.7 3 (18) 5 (29) 6 (35) 250 ⫾ 9 145 ⫾ 7 8 ⫾ 0.6 129 ⫾ 3 330 ⫾ 26
NOTE: Plus/minus values are means ⫾ SE, and all other values are numbers of patients followed in parentheses by the percentage of patients in the group. aP ⫽ 0.036 for the comparison between groups. bP ⫽ 0.045 for the comparison between groups; P ⬎ 0.05 for all other comparisons between groups. cHigher scores indicate greater disease activity. Scores of 150 or less indicate that the disease is in remission. dHigher scores indicate better quality of life. Scores greater than 170 are observed in patients in remission.
eligible because of the presence of disease of the distal colon. Other reasons why patients were not entered were the presence of an enteric infection (n ⫽ 4), refusal by the patient or physician (n ⫽ 18), and continued need for contraindicated medications (n ⫽ 3). One hundred thirty-four patients were randomized; of these, 66 were assigned to receive antibiotics and 68 to receive placebo. One hundred patients had disease restricted to the ileum, whereas 34 patients had disease in both sites. The baseline characteristics of the 2 treatment groups were similar (Table 1), with the exceptions that significantly more cigarette smokers were assigned to the placebo group in the ileum-only stratum, and patients in the ileum-and-colon stratum who received antibiotics were older.
by worsening disease activity (n ⫽ 6) and noncompliance with the study protocol (n ⫽ 3). Primary Outcome Four patients, 2 in each treatment group, took at least 1 dose of the study medication but failed to provide efficacy data and were excluded from analysis. The remission rates by week of follow-up are shown in Figure 1. After 8 weeks of treatment, the proportion of patients in remission was 21 of 64 (33%; 95% confidence interval, 21%– 44%) in the antibiotic-treated group and 25 of
Early Termination of Therapy A total of 38 patients (28%) did not complete the 8 weeks of study medication; 22 patients (33%) in the antibiotic group and 16 patients (24%) in the placebo group (P ⫽ 0.21). In patients who received placebo, 12 withdrew from therapy because of worsening disease activity, 2 were lost to follow-up, and 2 were noncompliant; no withdrawal was due to the occurrence of an adverse event. In patients who were treated with antibiotics, the most common reason for withdrawal of therapy was the occurrence of an adverse event (n ⫽ 13), followed
Figure 1. The proportion of patients with Crohn’s disease in remission at each study visit by treatment group, for the patients overall and according to the site of disease. Remission was defined by a score of less than 150 on the CDAI. The P value is based on the comparison of the rate of remission at week 8.
36
STEINHART ET AL.
GASTROENTEROLOGY Vol. 123, No. 1
Table 2. Proportion of the Total Score on the Crohn’s Disease Activity Index Attributable to Each Subcomponent Stratum Overall
Subcomponents (%)a Stool frequency Week 0 Week 8 Abdominal pain Week 0 Week 8 General well-being Week 0 Week 8 Extraintestinal manifestations of disease Week 0 Week 8 Use of opiates to treat diarrhea Week 0 Week 8 Abdominal mass Week 0 Week 8 Hematocrit Week 0 Week 8 Body weight Week 0 Week 8
Small bowel
Small bowel and colon
Budesonide/ placebo (n ⫽ 66)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 64)
Budesonide/ placebo (n ⫽ 50)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 47)
Budesonide/ placebo (n ⫽ 16)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 17)
16 ⫾ 1 18 ⫾ 2
17 ⫾ 1‡ 29 ⫾ 2
17 ⫾ 1 20 ⫾ 2
17 ⫾ 1b 31 ⫾ 3
13 ⫾ 3 11 ⫾ 3
16 ⫾ 2b 23 ⫾ 3
24 ⫾ 1 22 ⫾ 1
23 ⫾ 1 18 ⫾ 1
24 ⫾ 1 22 ⫾ 1
23 ⫾ 1 17 ⫾ 2
24 ⫾ 2 22 ⫾ 3
21 ⫾ 2 20 ⫾ 3
37 ⫾ 1 27 ⫾ 2
35 ⫾ 1 26 ⫾ 2
37 ⫾ 1 28 ⫾ 2
37 ⫾ 2 28 ⫾ 2
34 ⫾ 2 24 ⫾ 4
31 ⫾ 2 20 ⫾ 4
6⫾1 4⫾1
6⫾1 5⫾1
6⫾1 4⫾1
5⫾1 5⫾1
6⫾1 5⫾2
9⫾1 6⫾2
2⫾1 2⫾1
2⫾1 2⫾1
1⫾1 2⫾1
1⫾1 1⫾1
3⫾1 2⫾1
3⫾1 3⫾2
4⫾1 4⫾1
5⫾1 3⫾1
4⫾1 4⫾2
4⫾1 2⫾1
3⫾1 3⫾2
5⫾2 5⫾3
10 ⫾ 1 22 ⫾ 3
11 ⫾ 1 17 ⫾ 2
9⫾1 21 ⫾ 3
10 ⫾ 1 16 ⫾ 2
13 ⫾ 1 28 ⫾ 7
13 ⫾ 2 21 ⫾ 4
2⫾1 1⫾1
1⫾1 1⫾1
1⫾1 0⫾1
1⫾1 1⫾2
4⫾1 5⫾2
2⫾2 2⫾3
NOTE: Data are presented as means ⫾ SE. aThe subcomponent scores are expressed as percentage of the CDAI total score. bP ⬍ 0.001 for the comparison of change from week 0 to week 8 between the treatment groups.
66 (38%; 95% confidence interval, 26%–50%) in those who received placebo (P ⫽ 0.55). For the population as a whole, no difference was noted between the treatment groups in the rate of remission at any time during the follow-up period. However, a significant interaction was found between site of disease and treatment with antibiotics (P ⫽ 0.025). For the 33 patients with disease of both the ileum and colon, 9 of 17 patients (53%) who received antibiotics were in remission at week 8, compared with 4 of 16 (25%) of the patients who received placebo (P ⫽ 0.10). Conversely, for patients with disease of the ileum only, 12 of 47 (26%) patients assigned to antibiotic therapy were in remission, compared with 21 of 50 (42%) patients who received placebo (P ⫽ 0.09; Figure 1). On account of the very low rate of response in antibiotic-treated patients in this group, we conducted exploratory analyses that examined the effects of therapy on the various subcomponents of the CDAI. At the baseline visit, the proportion of the total score on the CDAI attributable to the soft stool frequency component
was similar in patients treated with antibiotics and those who received placebo (Table 2). However, after treatment, patients who received antibiotics had significantly more loose bowel movements (P ⬍ 0.001). At week 8, the stool frequency component of the score accounted for 29% of the total score in patients treated with antibiotics as compared with 18% in the placebo group. No such difference between groups was observed for the other components of the score. Secondary Measures of Response Figure 2A shows the scores on the CDAI. At the baseline visit, the mean scores in both groups were similar and were in keeping with moderately active Crohn’s disease. After treatment, scores decreased in both groups in a similar manner. Analysis of these data according to the disease location strata was again consistent with an interaction between site of disease and treatment allocation (P ⫽ 0.05). However, the mean scores on the CDAI over time were not significantly different between
July 2002
BUDESONIDE AND ANTIBIOTICS FOR CROHN’S DISEASE
37
Figure 2. Disease activity as assessed by the CDAI for the patients overall (A), patients with disease in the small bowel only (B), and patients with disease in both the small bowel and colon (C). Higher scores indicate worse disease activity; scores of less than 150 are found in patients in remission. The P value was derived by repeatedmeasures analysis of variance for the comparison of changes in scores over time between the treatment groups. Values are means ⫾ SE.
the treatment groups in both strata (P ⫽ 0.15 for the ileal-only stratum [Figure 2B] and P ⫽ 0.81 for the ileum-and-colon stratum [Figure 2C]). At 8 weeks, in patients with disease in the ileum and colon who were treated with antibiotics, the decrease in the mean CDAI score was 98.3 ⫾ 16.6 compared with 73.3 ⫾ 24.2 in those who received placebo, whereas in patients with disease of the ileum only, the corresponding decrements were 85.8 ⫾ 13.7 and 113.6 ⫾ 12.7, respectively. Although no overall difference was shown in IBDQ scores between the treatment groups (Figure 3A), patients with disease of the colon who received antibiotics showed a slightly greater improvement in quality of life than those who received placebo (Figure 3C), whereas in patients with disease of only the small bowel (Figure 3B), a slightly greater improvement in quality of life was noted in patients who received placebo. Adherence to the Study Medication Patients assigned to placebo took 92% of the placebo ciprofloxacin tablets and 91% of the placebo metronidazole tablets. Patients who received antibiotics took 93% of the ciprofloxacin tablets and 82% of the metronidazole tablets. Adherence to budesonide therapy was high in both groups; 97% of tablets were taken in the placebo group and 96% in the antibiotic group.
Adverse Events The majority of adverse events were mild or moderate in severity and did not result in withdrawal of the study medication. Patients treated with antibiotics were significantly more likely to experience diarrhea, dizziness, glossitis, taste disturbance, and vaginitis (Table 3). Manifestations of Cushing’s syndrome were uncommon in both treatment groups. In 13 of 66 patients (20%) in the antibiotic group, withdrawal from therapy was due to adverse events, whereas in patients who received placebo, 0 of 68 patients (0%) discontinued treatment for this reason (P ⬍ 0.001). The adverse events that led to withdrawal of therapy were nausea (n ⫽ 4), taste disturbance (n ⫽ 2), fever, weight loss, agitation, fatigue, pancreatitis, skin rash, and peripheral neuropathy (n ⫽ 1 each).
Discussion We found no beneficial effect of the addition of metronidazole and ciprofloxacin to budesonide in this group of patients with active Crohn’s disease of the ileum and proximal colon. The rate of remission after 8 weeks of treatment was similar in both treatment groups. Furthermore, no differences were detected in the analyses of mean changes in scores on the CDAI and the IBDQ. Because broad-spectrum antibiotics are frequently rec-
38
STEINHART ET AL.
GASTROENTEROLOGY Vol. 123, No. 1
Figure 3. The scores on the IBDQ for the patients overall (A), patients with disease in the small bowel only (B), and patients with disease in both the small bowel and colon (C). Higher scores indicate better quality of life; scores of greater than 170 are found in patients in remission. The P values were derived by repeated-measures analysis of variance for the comparison of changes in scores over time between the treatment groups. Values are means ⫾ SE.
Table 3. Adverse Events in the Study Patients According to Groupa
Adverse eventb Diarrhea Dizziness/lightheadedness Glossitis/oral candidiasis Taste disturbance Vaginitis Possible glucocorticoid-related eventsc Acne Arthralgia/myalgia Moon facies Insomnia Purpura Edema Serious adverse events
Budesonide/ placebo (n ⫽ 68)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 66)
4 (6) 4 (6) 0 0 2 (3)
13 (20) 14 (21) 9 (14) 18 (27) 11 (17)
2 (3) 19 (28) 2 (3) 9 (13) 3 (4) 2 (3) 3 (4)
4 (6) 15 (23) 3 (5) 7 (11) 3 (5) 3 (5) 2 (3)
NOTE. Data are presented as n (%). aAll adverse events that were significantly different between the treatment groups are listed along with the possible glucocorticoid-related events. Patients may have had more than 1 adverse event. bAll adverse events were significantly different between the treatment groups at P ⬍ 0.05. cThere were no significant differences between treatment groups in the proportion of patients experiencing possible glucocorticoid-related events.
ommended in combination with glucocorticoids or 5-aminosalicylates as treatment of active Crohn’s disease,12–16 the results of this study are highly relevant to the care of patients. We did not study antibiotics as monotherapy and therefore cannot exclude the possibility that these drugs are beneficial when used on their own. However, previous controlled trials provide limited evidence to support their efficacy.21–25 The largest trial, which compared 2 doses of metronidazole with placebo in 105 patients, showed a dose-dependent significant decrease in the CDAI, but the 3 groups had similarly low remission rates (25%, placebo; 36%, 10 mg/kg metronidazole; 27%, 20 mg/kg metronidazole) after 16 weeks of treatment.23 Controlled trials evaluating ciprofloxacin alone or in combination with metronidazole have inferred a beneficial response with remission rates of 45%–56% reported after 6 –12 weeks of therapy. However, in these studies the absence of a placebo group and an inadequate number of patients are major limitations. The design of our trial required the exclusion of patients with disease involvement distal to the hepatic flexure. This exclusion criterion may have compromised the ability to detect a higher clinical response with the addition of antibiotics, because other studies23,26 have suggested that antibiotics are more effective in patients with Crohn’s disease of the colon either as colitis or
July 2002
ileocolitis; no improvement occurred in patients with isolated disease of the ileum. Indeed, in our study a trend was observed toward greater efficacy in the prespecified subgroup of patients with colonic involvement who received antibiotics; 53% of these patients were in remission at week 8 compared with 25% of those who were assigned to placebo. However, our findings were based on a subgroup analysis and lacked sufficient statistical power to definitively determine whether antibiotics might be effective for patients with Crohn’s disease of the colon. Future trials are required to specifically examine this possibility. The clinical results pointing to a preferential effect by antibiotics on colonic Crohn’s disease are in accord with findings from experimental models in which abrogation of intestinal inflammation after antibiotic therapy is also observed predominantly in the colon.9,10 Although the mechanisms whereby bacteria modulate the T cells responsible for the perpetuation of intestinal inflammation are unknown, the composite findings from experimental models and clinical studies imply that intraluminal concentrations of bacteria, which in the colon are 100 – 1000-fold increased above those in the ileum, may be an important factor.27 Although the response rate to budesonide (38%) was somewhat lower than that seen in other trials (48% to 62%),28,29 our result overlaps the 95% confidence interval of the estimate of response that was based on an aggregation of trials of similar design. Thus, our study lends further support to the utility of oral budesonide for the treatment of patients with active ileocecal Crohn’s disease. However, the low rate of remission (26%) observed in patients with ileal disease who received combined therapy was surprising and warrants comment. The analysis of the subcomponents of the CDAI showed that a significantly greater proportion of the total disease activity was due to loose bowel movements in patients who received antibiotics than in those who were treated with budesonide alone. There is no clear explanation that accounts for this finding, but differences in the activity of the disease or antibiotic-related diarrhea30,31 are both possible. Thus, in patients with colonic involvement, the overall benefit of antibiotics on the inflammatory process may have outweighed any negative effect of an antibiotic-related diarrhea. However, for patients with ileal disease, in which the relative composition and concentration of flora differ compared with the colon, the introduction of antibiotics may have negatively influenced scores on the CDAI, resulting in a lower-thanexpected rate of remission in this group of patients. This latter hypothesis is consistent with the observation that
BUDESONIDE AND ANTIBIOTICS FOR CROHN’S DISEASE
39
more patients who received antibiotics experienced diarrhea as an adverse effect of treatment. The adverse events observed in the study were generally of mild severity and are consistent with the known side effects of the drugs evaluated.32,33 Patients assigned to antibiotics were more likely to experience vaginitis, glossitis, and diarrhea, probably as a consequence of suppression of the normal bacterial flora. The low incidence of glucocorticoid-related adverse effects associated with budesonide therapy in our patients confirms the findings of other trials.2,3,28,29 In conclusion, this study shows that the addition of metronidazole and ciprofloxacin to budesonide is ineffective therapy for patients with active Crohn’s disease of the ileum, but this antibiotic combination may improve outcome in patients with disease involvement of the colon.
Appendix In addition to the authors, the following people at the following institutions participated in the trial (in parentheses, number of patients enrolled): Mount Sinai Hospital, Toronto, Ontario (24): G. Greenberg, A. H. Steinhart, and S. Mikolainis; Victoria General Hospital, Halifax, Nova Scotia (20): D. Leddin, D. MacIntosh, M. Kareemi, J. Love, S. van Zanten, J. Stewart, M. Edmunds, and H. McInnes; London Health Sciences Centre, London, Ontario (16): B. G. Feagan, J. W. D. McDonald, J. Gregor, W. Barnett, T. Ponich, J. Howard, M. Hopkins, H. Harrison, P. Walton-Mennill, and W. McCaw; Royal Victoria Hospital, Montreal, Quebec (14): A. Bitton, H. Vaupshas, and M. Bernier; St. Justine Hospital, Montreal, Quebec (14): E. Drouin, L. Chartrand, and A. Duhaime; Jewish General Hospital, Montreal, Quebec (11): A. Cohen and J. Rivard; McMaster University Medical Centre, Hamilton, Ontario (8): E. J. Irvine, B. J. Salena, K. Croitoru, S. M. Collins, and M. Donnelly; Saint John Regional Hospital, Saint John, New Brunswick (7): A. W. Cockeram and C. Arsenault; Health Sciences Centre, Winnipeg, Manitoba (6): C. Bernstein, B. G. Rosser, and P. Rawsthorne; Hotel Dieu de Quebec, Quebec City, Quebec (4): P. Pare´ and D. Lafrance; University of Alberta Hospitals, Edmonton, Alberta (4): R. Fedorak, J. McKaigney, E. Lalor, and D. Fisher; Ottawa Civic Hospital, Ottawa, Ontario (2): C. Dube and A. Port; Pasqua Hospital, Regina, Saskatchewan (2): J. McHattie and M. Scrivens; Health Sciences Centre, Calgary, Alberta (1): L. Sutherland and V. Todd; and St. Michael’s Hospital, Toronto, Ontario (1): J. Baker, R. Petroniene, N. Sayani, and G. Stewart.
40
STEINHART ET AL.
References 1. Summers RW, Switz DM, Sessions JT Jr., Becktel JM, Best WR, Kern F Jr., Singleton JW. National Cooperative Crohn’s Disease Study: results of drug treatment. Gastroenterology 1979;77: 847– 869. 2. Greenberg GR, Feagan BG, Martin F, Sutherland LR, Thomson ABR, Williams CN, Nilsson LG, Persson T. Oral budesonide as maintenance treatment for Crohn’s disease: a placebo-controlled, dose-ranging study. Gastroenterology 1996;110:45–51. 3. Rutgeerts P, Lofberg R, Malchow H, Lamers C, Olaison G, Jewell D, Danielsson A, Goebell H, Thomsen OO, Lorenz-Meyer H. A comparison of budesonide with prednisolone for active Crohn’s disease. N Engl J Med 1994;331:842– 845. 4. Candy S, Wright J, Gerber M, Adams G, Gerig M, Goodman R. A controlled double blind study of azathioprine in the management of Crohn’s disease. Gut 1995;37:674 – 678. 5. Feagan BG, Rochon J, Fedorak RN, Irvine EJ, Wild G, Sutherland L, Steinhart AH, Greenberg GR, Gillies R, Hopkins M. Methotrexate for the treatment of Crohn’s disease. N Engl J Med 1995; 332:292–297. 6. Elson CO. Commensal bacteria as targets in Crohn’s disease. Gastroenterology 2000;119:254 –257. 7. Duchmann R, May E, Heike M, Knolle P, Neurath M, Meyer zum Buschenfelde KH. T cell specificity and cross reactivity towards enterobacteria, bacteroides, bifidobacterium, and antigens from resident intestinal flora in humans. Gut 1999;44:812– 818. 8. Herfarth HH, Mohanty SP, Rath HC, Tonkonogy S, Sartor RB. Interleukin 10 suppresses experimental chronic, granulomatous inflammation induced by bacterial cell wall polymers. Gut 1996; 39:836 – 845. 9. McKay DM. Intestinal inflammation and the gut microflora. Can J Gastroenterol 1999;13:509 –516. 10. Strober W, Ludviksson BR, Fuss IJ. The pathogenesis of mucosal inflammation in murine models of inflammatory bowel disease and Crohn’s disease. Ann Intern Med 1998;128:848 – 856. 11. Atreya R, Mudter J, Finotto S, Mullberg J, Jostock T, Wirtz S, Schutz M, Bartsch B, Holtmann M, Becker C, Strand D, Czaja J, Schlaak JF, Lehr HA, Autschbach F, Schurmann G, Nishimoto N, Yoshizaki K, Ito H, Kishimoto T, Galle PR, Rose-John S, Neurath MF. Blockade of interleukin 6 trans signaling suppresses T-cell resistance against apoptosis in chronic intestinal inflammation: evidence in Crohn’s disease and experimental colitis in vivo. Nat Med 2000;6:583–588. 12. Bitton A, Peppercorn MA. Medical management of specific clinical presentations. Gastroenterol Clin North Am 1995;24:541– 548. 13. Peppercorn M. Is there a role for antibiotics as primary therapy in Crohn’s ileitis? J Clin Gastroenterol 1993;17:235–237. 14. Present DH. How to do without steroids in inflammatory bowel disease. Inflamm Bowel Dis 2000;6:48 –57. 15. Hanauer SB. Inflammatory bowel disease. N Engl J Med 1996; 334:841– 848. 16. Sartor RB. Antibiotics as therapeutic agents in Crohn’s disease. In: Bayless TM, Hanauer SB, eds. Advanced therapy of inflammatory bowel disease. Hamilton, Ontario: B.C. Decker, Inc., 2001:359 –362. 17. Best WR, Becktel JM, Singleton JW, Kern F Jr. Development of a Crohn’s disease activity index—National Cooperative Crohn’s Disease Study. Gastroenterology 1976;70:439 – 444. 18. Guyatt G, Mitchell A, Irvine EJ, Singer J, Williams N, Goodacre R, Tompkins C. A new measure of health status for clinical trials in inflammatory bowel disease. Gastroenterology 1989;96:804 – 810.
GASTROENTEROLOGY Vol. 123, No. 1
19. Irvine EJ, Feagan B, Rochon J, Archambault A, Fedorak RN, Groll A, Kinnear D, Saibil F, McDonald JW. Quality of life: a valid and reliable measure of therapeutic efficacy in the treatment of inflammatory bowel disease. Gastroenterology 1994;106: 287–296. 20. SAS system software. Version 8. SAS Institute Inc., Cary, NC, 1999. 21. Ursing B, Alm T, Ba´ra`ny F, Bergelin I, Ganrot-Norlin K, Hoevels J, Huitfeldt B, Jarnerot G, Krause U, Krook A, Lindstrom B, Nordle O, Rosen A. A comparative study of metronidazole and sulfasalazine for active Crohn’s disease: the Cooperative Crohn’s Disease Study in Sweden. Gastroenterology 1982;83:550 –562. 22. Ambrose NS, Allan RN, Keighley MRB, Burdon DW, Youngs D, Barnes P, Lennard-Jones JE. Antibiotic therapy for treatment in relapse of intestinal Crohn’s disease. A prospective randomized study. Dis Colon Rectum 1985;28:81– 85. 23. Sutherland L, Singleton J, Sessions J, Hanauer S, Krawitt E, Rankin G, Summers R, Mekhjian H, Greenberger N, Kelly M, Levine J, Thomson A, Alpert E, Prokipchuk E. Double blind, placebo controlled trial of metronidazole in Crohn’s disease. Gut 1991;32:1071–1075. 24. Rutgeerts P, Hiele M, Geboes K, Peeters M, Penninckz F, Aerts R, Kerremans R. Controlled trial of metronidazole treatment for prevention of Crohn’s recurrence after ileal resection. Gastroenterology 1995;108:1617–1621. 25. Prantera C, Zannoni F, Scribano ML, Berto E, Andreoli A, Kohn A, Luzi C. An antibiotic regimen for the treatment of active Crohn’s disease: a randomized, controlled clinical trial of metronidazole plus ciprofloxacin. Am J Gastroenterol 1996;91:328 –332. 26. Greenbloom SL, Steinhart AH, Greenberg GR. Combination ciprofloxacin and metronidazole for active Crohn’s disease. Can J Gastroenterol 1998;12:53–56. 27. Berg RD. The indigenous gastrointestinal microflora. Trends Microbiol 1996;4:430 – 435. 28. Campieri M, Ferguson A, Doe W, Persson T, Nilsson L-G. Global Budesonide Study Group. Oral budesonide is as effective as oral prednisolone in active Crohn’s disease. Gut 1997;41:209 –214. 29. Thomsen OO, Cortot A, Jewell D, Wright JP, Winter T, Veloso FT, Vatn M, Persson T, Pettersson E. A comparison of budesonide and mesalamine for active Crohn’s disease. N Engl J Med 1998; 339:370 –374. 30. McFarland LV. Epidemiology, risk factors and treatments for antibiotic-associated diarrhea. Dig Dis 1998;16:292–307. 31. Chasseny O, Michaux A, Bergmann JF. Drug-induced diarrhea. Drug Saf 2000;22:53–72. 32. Freeman CD, Klutman NE, Lamp KC. Metronidazole. A therapeutic review and update. Drugs 1997;54:679 –708. 33. Brogden RN, McTavish D. Budesonide: an updated review of its pharmacological properties, and therapeutic efficacy in asthma and rhinitis. Drugs 1992;44:375– 407.
Received March 15, 2002. Accepted March 21, 2002. Address requests for reprints to: A. Hillary Steinhart, M.D., Room 445, Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario, M5G 1X5 Canada. e-mail: [email protected]. Supported by the Crohn’s and Colitis Foundation of Canada and by AstraZeneca. The authors thank the patients who participated in the study, to Bayer, Inc., for providing ciprofloxacin and placebo tablets, to Apotex for providing metronidazole and placebo tablets, to AstraZeneca for providing budesonide, and to Beverley Jasevicius for assistance in preparing the manuscript.
Combined Budesonide and Antibiotic Therapy for Active Crohn’s Disease: A Randomized Controlled Trial A. HILLARY STEINHART,* BRIAN G. FEAGAN,‡ CINDY J. WONG,‡ MARGARET VANDERVOORT,‡ SHELLEY MIKOLAINIS,* KENNETH CROITORU,§ ERNEST SEIDMAN,㛳 DESMOND J. LEDDIN,¶ ALAIN BITTON,# ERIC DROUIN,㛳 ALBERT COHEN,** and GORDON R. GREENBERG* for the Crohn’s and Colitis Foundation of Canada Inflammatory Bowel Disease Network Investigators *Department of Medicine, Mount Sinai Hospital, Toronto, Ontario; ‡London Clinical Trials Research Group, The John P. Robarts Research Institute, London, Ontario; §Department of Medicine, McMaster University, Hamilton, Ontario; 㛳Gastroenterology Division, St. Justine Hospital, Montreal, Quebec; ¶Department of Medicine, Dalhousie University, Halifax, Nova Scotia; #Royal Victoria Hospital, Montreal, Quebec; and **Jewish General Hospital, Montreal, Quebec, Canada
Background & Aims: Although antibiotics are frequently used to treat Crohn’s disease, this practice is not supported by strong evidence from randomized trials. Methods: We conducted a double-blind multicenter study of patients with active Crohn’s disease of the ileum, right colon, or both. Patients were randomized to receive oral ciprofloxacin and metronidazole, both 500 mg twice daily, or placebo for 8 weeks. All patients received oral budesonide 9 mg once daily. The primary efficacy measure was the proportion of patients in remission at week 8. Results: Of the 134 patients who were randomized, 130 were evaluated for efficacy; 66 received placebo, and 64 received antibiotics. At week 8, 21 patients (33%) assigned to antibiotics were in remission as compared with 25 patients (38%) in the placebo group (P ⴝ 0.55; absolute difference, ⴚ5%; 95% confidence interval, ⴚ21% to 11%). An interaction (P ⴝ 0.025) between treatment allocation and disease location on treatment response was identified. Among patients with disease of the colon, 9 of 17 (53%) were in remission after treatment with antibiotics, compared with 4 of 16 (25%) of those who received placebo (P ⴝ 0.10). Discontinuation of therapy because of adverse events occurred in 13 of 66 (20%) patients treated with antibiotics, compared with 0 of 68 in the group who received placebo (P < 0.001). Conclusions: In patients with active Crohn’s disease of the ileum, the addition of ciprofloxacin and metronidazole to budesonide is an ineffective intervention, but this antibiotic combination may improve outcome when there is involvement of the colon.
rohn’s disease is a chronic inflammatory disorder of unknown origin that commonly affects the ileum and colon. Medical therapy for active Crohn’s disease comprises the use of nonspecific anti-inflammatory drugs, with corticosteroids providing the most consistent benefit.1–5 Recent experimental evidence derived from
C
several genetically susceptible animal models indicates that the normal enteric bacterial flora contribute to chronic pathologic inflammatory responses in the intestine.6 –11 Because these models emphasize the importance of interactions between the mucosal immune system and commensal bacteria,6,7 a sound rationale exists to evaluate treatment regimens that target both of these factors. Thus, the addition of broad-spectrum antibiotics might enhance the efficacy of anti-inflammatory drugs such as corticosteroids. Antibiotic therapy, notably metronidazole and ciprofloxacin, is frequently advocated for Crohn’s disease,12–16 but randomized controlled trials to support this practice are limited. Furthermore, no controlled trials have evaluated the efficacy of antibiotics and corticosteroids in combination. For these reasons, we assessed the use of budesonide, a glucocorticosteroid nearly as effective as prednisolone but without the steroid-associated adverse effects,3 combined with metronidazole and ciprofloxacin for the treatment of active Crohn’s disease of the ileum and of the ileum and proximal colon.
Methods Patients This randomized controlled trial was conducted at 14 university medical centers and 2 community-based gastroenterology practices in Canada. Investigators and patients were unaware of the treatment assignments. The protocol was approved by the institutional review board at each center. All patients gave written informed consent. Abbreviation used in this paper: IBDQ, Inflammatory Bowel Disease Questionnaire. © 2002 by the American Gastroenterological Association 0016-5085/02/$35.00 doi:10.1053/gast.2002.34225
34
STEINHART ET AL.
Eligible patients were 14 years of age or older, with Crohn’s disease involving the ileum, proximal colon (ascending colon to mid transverse colon), or both. Disease location was determined by either colonoscopy or air-contrast barium enema performed within 2 years of entry to the study. Patients had active disease, as defined by a score on the Crohn’s Disease Activity Index (CDAI)17 of 200 – 400 points. This index incorporates the following items: the frequency of soft stools, the severity of abdominal pain and the level of general well-being in the 7 days preceding the assessment, the presence or absence of an abdominal mass or extraintestinal manifestations of the disease, the use of opiates to treat diarrhea, hematocrit, and weight. Scores range from 0 to approximately 600; higher scores indicate greater disease activity. A score of less than 150 indicates inactive disease. Patients receiving treatment with antibiotics (within 1 week of the randomization visit), corticosteroids, parenteral nutrition, enteral feeds (within 2 weeks), or immunosuppressants (within 12 weeks) were not eligible. Patients who had received 5-aminosalicylic acid drugs or sulfasalazine were eligible if the medication was discontinued at least 7 days before the randomization visit. Patients with more than 100 cm of small intestine surgically removed or those with an ileostomy were not included.
Baseline Studies Patients were evaluated for eligibility 1 week before the randomization visit. At this time they were given a diary in which to record the data necessary to calculate the CDAI. At the randomization visit, demographic information was obtained; a physical examination, blood tests, and stool cultures and an assay for Clostridium difficile toxin were performed; and scores on the CDAI and Inflammatory Bowel Disease Questionnaire (IBDQ)18,19 were calculated. Scores on this measure range from 32 to 224; higher scores mean a better quality of life. Scores of greater than 170 are observed in patients whose disease is in remission.
Randomization The patients were randomly assigned, by means of a computer-generated schedule, to receive either oral ciprofloxacin (Cipro; Bayer, Inc., Toronto, Ontario) and metronidazole (Apo-metronidazole; Apotex, Inc., Toronto, Ontario), both 500 mg twice daily, or placebo for 8 weeks. Patients also received oral budesonide2,3 (Entocort; AstraZeneca, Mississauga, Ontario) 9 mg once daily. The randomization schedule was stratified according to center and disease location (ileum or ileum and colon) in permutated blocks of 4 or 6. The placebo tablets were identical in appearance to the ciprofloxacin and metronidazole preparations. Investigators and patients were unaware of the treatment assignment.
Other Treatments for Crohn’s Disease Patients did not use other drugs for the treatment of active Crohn’s disease, including glucocorticoids other than
GASTROENTEROLOGY Vol. 123, No. 1
budesonide, immunosuppressants, and infliximab. Total parenteral nutrition and enteral feeding were not allowed.
Follow-up Patients were assessed in the clinic 2, 4, 6, and 8 weeks after the randomization visit. At each visit, the scores on the CDAI and the IBDQ were calculated, and blood was obtained for chemical analyses. The study medication was discontinued if other treatments for active Crohn’s disease were initiated. At each visit, unused medications were returned, and a pill count was performed to evaluate patients’ adherence with drug therapy.
Measures of Outcome The primary measure of response was remission at week 8. Remission was defined as a score of less than 150 on the CDAI. Secondary measures included the mean changes in scores on the CDAI, quality of life, and adverse drug reactions.
Statistical Analysis The baseline characteristics of the patients were analyzed with descriptive statistics. For the primary analysis, we compared the proportion of patients in the 2 treatment groups who were in remission at week 8 by using Pearson’s 2 test. This analysis included all patients who provided efficacy data from at least 1 postrandomization assessment. A logistic regression analysis evaluated the potential interaction between site of disease and treatment allocation on the primary outcome. The mean scores over time on the CDAI and the IBDQ were analyzed by a repeated-measures analysis of variance. Scores from patients who withdrew from the study prematurely were analyzed by using the last-observation-carriedforward principle. Separate analyses were performed for each stratum of disease location. We used the 2 test or the Fisher exact test, as appropriate, to compare the number of patients with adverse events in each group. Statistical analyses were performed with SAS software.20 A 2-sided P value of 0.05 or less was considered to indicate statistical significance. We estimated that 55% of patients treated with budesonide alone would be in remission at week 8. A minimum clinically important absolute difference in the occurrence of this outcome was considered to be 25%. Thus, a remission rate of 80% in the combination budesonide/antibiotic group was anticipated. Randomization of 52 evaluable patients per group would give a power of 80% to detect this difference at the 0.05 significance level. A total of 134 patients were enrolled to allow for a nonevaluable rate of up to 20%.
Results Randomization Between June 1998 and September 2000, 209 patients were evaluated for eligibility. Thirty-six patients were excluded because of a CDAI score that was not in the specified range. Fourteen patients were not
July 2002
BUDESONIDE AND ANTIBIOTICS FOR CROHN’S DISEASE
35
Table 1. Baseline Characteristics of the Patients Overall and According to the Site of Disease Stratum Overall
Small bowel
Small bowel and colon
Characteristic
Budesonide/ placebo (n ⫽ 68)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 66)
Budesonide/ placebo (n ⫽ 51)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 49)
Budesonide/ placebo (n ⫽ 17)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 17)
Age (yr) Male sex, n (%) Weight (kg) Years since diagnosis Abdominal mass, n (%) Previous surgery for Crohn’s disease, n (%) Cigarette smoker, n (%) CDAI scorec IBDQ scored White blood cell count (⫻ 109/L) Hemoglobin (g/L) Platelet count (⫻ 109/L)
30 ⫾ 1 27 (40) 66 ⫾ 2 6 ⫾ 1.0 8 (12) 23 (34) 33 (49) 278 ⫾ 7 140 ⫾ 3 8 ⫾ 0.3 130 ⫾ 2 340 ⫾ 15
34 ⫾ 2 30 (45) 66 ⫾ 2 5 ⫾ 0.7 10 (15) 25 (38) 25 (38) 273 ⫾ 6 137 ⫾ 3 8 ⫾ 0.3 132 ⫾ 2 330 ⫾ 13
33 ⫾ 1 19 (37) 68 ⫾ 3 7 ⫾ 1.3 7 (14) 19 (37) 30 (59) 280 ⫾ 8 139 ⫾ 3 8 ⫾ 0.3 132 ⫾ 2 320 ⫾ 14
34 ⫾ 1 20 (41) 65 ⫾ 2 6 ⫾ 1.0 7 (14) 20 (41) 19 (39)b 281 ⫾ 8 134 ⫾ 4 8 ⫾ 0.4 133 ⫾ 3 330 ⫾ 15
22 ⫾ 2 8 (47) 59 ⫾ 3 4 ⫾ 0.9 1 (6) 4 (24) 3 (18) 273 ⫾ 15 144 ⫾ 5 9 ⫾ 0.9 125 ⫾ 4 398 ⫾ 41
32 ⫾ 4a 10 (59) 68 ⫾ 4 4 ⫾ 0.7 3 (18) 5 (29) 6 (35) 250 ⫾ 9 145 ⫾ 7 8 ⫾ 0.6 129 ⫾ 3 330 ⫾ 26
NOTE: Plus/minus values are means ⫾ SE, and all other values are numbers of patients followed in parentheses by the percentage of patients in the group. aP ⫽ 0.036 for the comparison between groups. bP ⫽ 0.045 for the comparison between groups; P ⬎ 0.05 for all other comparisons between groups. cHigher scores indicate greater disease activity. Scores of 150 or less indicate that the disease is in remission. dHigher scores indicate better quality of life. Scores greater than 170 are observed in patients in remission.
eligible because of the presence of disease of the distal colon. Other reasons why patients were not entered were the presence of an enteric infection (n ⫽ 4), refusal by the patient or physician (n ⫽ 18), and continued need for contraindicated medications (n ⫽ 3). One hundred thirty-four patients were randomized; of these, 66 were assigned to receive antibiotics and 68 to receive placebo. One hundred patients had disease restricted to the ileum, whereas 34 patients had disease in both sites. The baseline characteristics of the 2 treatment groups were similar (Table 1), with the exceptions that significantly more cigarette smokers were assigned to the placebo group in the ileum-only stratum, and patients in the ileum-and-colon stratum who received antibiotics were older.
by worsening disease activity (n ⫽ 6) and noncompliance with the study protocol (n ⫽ 3). Primary Outcome Four patients, 2 in each treatment group, took at least 1 dose of the study medication but failed to provide efficacy data and were excluded from analysis. The remission rates by week of follow-up are shown in Figure 1. After 8 weeks of treatment, the proportion of patients in remission was 21 of 64 (33%; 95% confidence interval, 21%– 44%) in the antibiotic-treated group and 25 of
Early Termination of Therapy A total of 38 patients (28%) did not complete the 8 weeks of study medication; 22 patients (33%) in the antibiotic group and 16 patients (24%) in the placebo group (P ⫽ 0.21). In patients who received placebo, 12 withdrew from therapy because of worsening disease activity, 2 were lost to follow-up, and 2 were noncompliant; no withdrawal was due to the occurrence of an adverse event. In patients who were treated with antibiotics, the most common reason for withdrawal of therapy was the occurrence of an adverse event (n ⫽ 13), followed
Figure 1. The proportion of patients with Crohn’s disease in remission at each study visit by treatment group, for the patients overall and according to the site of disease. Remission was defined by a score of less than 150 on the CDAI. The P value is based on the comparison of the rate of remission at week 8.
36
STEINHART ET AL.
GASTROENTEROLOGY Vol. 123, No. 1
Table 2. Proportion of the Total Score on the Crohn’s Disease Activity Index Attributable to Each Subcomponent Stratum Overall
Subcomponents (%)a Stool frequency Week 0 Week 8 Abdominal pain Week 0 Week 8 General well-being Week 0 Week 8 Extraintestinal manifestations of disease Week 0 Week 8 Use of opiates to treat diarrhea Week 0 Week 8 Abdominal mass Week 0 Week 8 Hematocrit Week 0 Week 8 Body weight Week 0 Week 8
Small bowel
Small bowel and colon
Budesonide/ placebo (n ⫽ 66)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 64)
Budesonide/ placebo (n ⫽ 50)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 47)
Budesonide/ placebo (n ⫽ 16)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 17)
16 ⫾ 1 18 ⫾ 2
17 ⫾ 1‡ 29 ⫾ 2
17 ⫾ 1 20 ⫾ 2
17 ⫾ 1b 31 ⫾ 3
13 ⫾ 3 11 ⫾ 3
16 ⫾ 2b 23 ⫾ 3
24 ⫾ 1 22 ⫾ 1
23 ⫾ 1 18 ⫾ 1
24 ⫾ 1 22 ⫾ 1
23 ⫾ 1 17 ⫾ 2
24 ⫾ 2 22 ⫾ 3
21 ⫾ 2 20 ⫾ 3
37 ⫾ 1 27 ⫾ 2
35 ⫾ 1 26 ⫾ 2
37 ⫾ 1 28 ⫾ 2
37 ⫾ 2 28 ⫾ 2
34 ⫾ 2 24 ⫾ 4
31 ⫾ 2 20 ⫾ 4
6⫾1 4⫾1
6⫾1 5⫾1
6⫾1 4⫾1
5⫾1 5⫾1
6⫾1 5⫾2
9⫾1 6⫾2
2⫾1 2⫾1
2⫾1 2⫾1
1⫾1 2⫾1
1⫾1 1⫾1
3⫾1 2⫾1
3⫾1 3⫾2
4⫾1 4⫾1
5⫾1 3⫾1
4⫾1 4⫾2
4⫾1 2⫾1
3⫾1 3⫾2
5⫾2 5⫾3
10 ⫾ 1 22 ⫾ 3
11 ⫾ 1 17 ⫾ 2
9⫾1 21 ⫾ 3
10 ⫾ 1 16 ⫾ 2
13 ⫾ 1 28 ⫾ 7
13 ⫾ 2 21 ⫾ 4
2⫾1 1⫾1
1⫾1 1⫾1
1⫾1 0⫾1
1⫾1 1⫾2
4⫾1 5⫾2
2⫾2 2⫾3
NOTE: Data are presented as means ⫾ SE. aThe subcomponent scores are expressed as percentage of the CDAI total score. bP ⬍ 0.001 for the comparison of change from week 0 to week 8 between the treatment groups.
66 (38%; 95% confidence interval, 26%–50%) in those who received placebo (P ⫽ 0.55). For the population as a whole, no difference was noted between the treatment groups in the rate of remission at any time during the follow-up period. However, a significant interaction was found between site of disease and treatment with antibiotics (P ⫽ 0.025). For the 33 patients with disease of both the ileum and colon, 9 of 17 patients (53%) who received antibiotics were in remission at week 8, compared with 4 of 16 (25%) of the patients who received placebo (P ⫽ 0.10). Conversely, for patients with disease of the ileum only, 12 of 47 (26%) patients assigned to antibiotic therapy were in remission, compared with 21 of 50 (42%) patients who received placebo (P ⫽ 0.09; Figure 1). On account of the very low rate of response in antibiotic-treated patients in this group, we conducted exploratory analyses that examined the effects of therapy on the various subcomponents of the CDAI. At the baseline visit, the proportion of the total score on the CDAI attributable to the soft stool frequency component
was similar in patients treated with antibiotics and those who received placebo (Table 2). However, after treatment, patients who received antibiotics had significantly more loose bowel movements (P ⬍ 0.001). At week 8, the stool frequency component of the score accounted for 29% of the total score in patients treated with antibiotics as compared with 18% in the placebo group. No such difference between groups was observed for the other components of the score. Secondary Measures of Response Figure 2A shows the scores on the CDAI. At the baseline visit, the mean scores in both groups were similar and were in keeping with moderately active Crohn’s disease. After treatment, scores decreased in both groups in a similar manner. Analysis of these data according to the disease location strata was again consistent with an interaction between site of disease and treatment allocation (P ⫽ 0.05). However, the mean scores on the CDAI over time were not significantly different between
July 2002
BUDESONIDE AND ANTIBIOTICS FOR CROHN’S DISEASE
37
Figure 2. Disease activity as assessed by the CDAI for the patients overall (A), patients with disease in the small bowel only (B), and patients with disease in both the small bowel and colon (C). Higher scores indicate worse disease activity; scores of less than 150 are found in patients in remission. The P value was derived by repeatedmeasures analysis of variance for the comparison of changes in scores over time between the treatment groups. Values are means ⫾ SE.
the treatment groups in both strata (P ⫽ 0.15 for the ileal-only stratum [Figure 2B] and P ⫽ 0.81 for the ileum-and-colon stratum [Figure 2C]). At 8 weeks, in patients with disease in the ileum and colon who were treated with antibiotics, the decrease in the mean CDAI score was 98.3 ⫾ 16.6 compared with 73.3 ⫾ 24.2 in those who received placebo, whereas in patients with disease of the ileum only, the corresponding decrements were 85.8 ⫾ 13.7 and 113.6 ⫾ 12.7, respectively. Although no overall difference was shown in IBDQ scores between the treatment groups (Figure 3A), patients with disease of the colon who received antibiotics showed a slightly greater improvement in quality of life than those who received placebo (Figure 3C), whereas in patients with disease of only the small bowel (Figure 3B), a slightly greater improvement in quality of life was noted in patients who received placebo. Adherence to the Study Medication Patients assigned to placebo took 92% of the placebo ciprofloxacin tablets and 91% of the placebo metronidazole tablets. Patients who received antibiotics took 93% of the ciprofloxacin tablets and 82% of the metronidazole tablets. Adherence to budesonide therapy was high in both groups; 97% of tablets were taken in the placebo group and 96% in the antibiotic group.
Adverse Events The majority of adverse events were mild or moderate in severity and did not result in withdrawal of the study medication. Patients treated with antibiotics were significantly more likely to experience diarrhea, dizziness, glossitis, taste disturbance, and vaginitis (Table 3). Manifestations of Cushing’s syndrome were uncommon in both treatment groups. In 13 of 66 patients (20%) in the antibiotic group, withdrawal from therapy was due to adverse events, whereas in patients who received placebo, 0 of 68 patients (0%) discontinued treatment for this reason (P ⬍ 0.001). The adverse events that led to withdrawal of therapy were nausea (n ⫽ 4), taste disturbance (n ⫽ 2), fever, weight loss, agitation, fatigue, pancreatitis, skin rash, and peripheral neuropathy (n ⫽ 1 each).
Discussion We found no beneficial effect of the addition of metronidazole and ciprofloxacin to budesonide in this group of patients with active Crohn’s disease of the ileum and proximal colon. The rate of remission after 8 weeks of treatment was similar in both treatment groups. Furthermore, no differences were detected in the analyses of mean changes in scores on the CDAI and the IBDQ. Because broad-spectrum antibiotics are frequently rec-
38
STEINHART ET AL.
GASTROENTEROLOGY Vol. 123, No. 1
Figure 3. The scores on the IBDQ for the patients overall (A), patients with disease in the small bowel only (B), and patients with disease in both the small bowel and colon (C). Higher scores indicate better quality of life; scores of greater than 170 are found in patients in remission. The P values were derived by repeated-measures analysis of variance for the comparison of changes in scores over time between the treatment groups. Values are means ⫾ SE.
Table 3. Adverse Events in the Study Patients According to Groupa
Adverse eventb Diarrhea Dizziness/lightheadedness Glossitis/oral candidiasis Taste disturbance Vaginitis Possible glucocorticoid-related eventsc Acne Arthralgia/myalgia Moon facies Insomnia Purpura Edema Serious adverse events
Budesonide/ placebo (n ⫽ 68)
Budesonide/ ciprofloxacin/ metronidazole (n ⫽ 66)
4 (6) 4 (6) 0 0 2 (3)
13 (20) 14 (21) 9 (14) 18 (27) 11 (17)
2 (3) 19 (28) 2 (3) 9 (13) 3 (4) 2 (3) 3 (4)
4 (6) 15 (23) 3 (5) 7 (11) 3 (5) 3 (5) 2 (3)
NOTE. Data are presented as n (%). aAll adverse events that were significantly different between the treatment groups are listed along with the possible glucocorticoid-related events. Patients may have had more than 1 adverse event. bAll adverse events were significantly different between the treatment groups at P ⬍ 0.05. cThere were no significant differences between treatment groups in the proportion of patients experiencing possible glucocorticoid-related events.
ommended in combination with glucocorticoids or 5-aminosalicylates as treatment of active Crohn’s disease,12–16 the results of this study are highly relevant to the care of patients. We did not study antibiotics as monotherapy and therefore cannot exclude the possibility that these drugs are beneficial when used on their own. However, previous controlled trials provide limited evidence to support their efficacy.21–25 The largest trial, which compared 2 doses of metronidazole with placebo in 105 patients, showed a dose-dependent significant decrease in the CDAI, but the 3 groups had similarly low remission rates (25%, placebo; 36%, 10 mg/kg metronidazole; 27%, 20 mg/kg metronidazole) after 16 weeks of treatment.23 Controlled trials evaluating ciprofloxacin alone or in combination with metronidazole have inferred a beneficial response with remission rates of 45%–56% reported after 6 –12 weeks of therapy. However, in these studies the absence of a placebo group and an inadequate number of patients are major limitations. The design of our trial required the exclusion of patients with disease involvement distal to the hepatic flexure. This exclusion criterion may have compromised the ability to detect a higher clinical response with the addition of antibiotics, because other studies23,26 have suggested that antibiotics are more effective in patients with Crohn’s disease of the colon either as colitis or
July 2002
ileocolitis; no improvement occurred in patients with isolated disease of the ileum. Indeed, in our study a trend was observed toward greater efficacy in the prespecified subgroup of patients with colonic involvement who received antibiotics; 53% of these patients were in remission at week 8 compared with 25% of those who were assigned to placebo. However, our findings were based on a subgroup analysis and lacked sufficient statistical power to definitively determine whether antibiotics might be effective for patients with Crohn’s disease of the colon. Future trials are required to specifically examine this possibility. The clinical results pointing to a preferential effect by antibiotics on colonic Crohn’s disease are in accord with findings from experimental models in which abrogation of intestinal inflammation after antibiotic therapy is also observed predominantly in the colon.9,10 Although the mechanisms whereby bacteria modulate the T cells responsible for the perpetuation of intestinal inflammation are unknown, the composite findings from experimental models and clinical studies imply that intraluminal concentrations of bacteria, which in the colon are 100 – 1000-fold increased above those in the ileum, may be an important factor.27 Although the response rate to budesonide (38%) was somewhat lower than that seen in other trials (48% to 62%),28,29 our result overlaps the 95% confidence interval of the estimate of response that was based on an aggregation of trials of similar design. Thus, our study lends further support to the utility of oral budesonide for the treatment of patients with active ileocecal Crohn’s disease. However, the low rate of remission (26%) observed in patients with ileal disease who received combined therapy was surprising and warrants comment. The analysis of the subcomponents of the CDAI showed that a significantly greater proportion of the total disease activity was due to loose bowel movements in patients who received antibiotics than in those who were treated with budesonide alone. There is no clear explanation that accounts for this finding, but differences in the activity of the disease or antibiotic-related diarrhea30,31 are both possible. Thus, in patients with colonic involvement, the overall benefit of antibiotics on the inflammatory process may have outweighed any negative effect of an antibiotic-related diarrhea. However, for patients with ileal disease, in which the relative composition and concentration of flora differ compared with the colon, the introduction of antibiotics may have negatively influenced scores on the CDAI, resulting in a lower-thanexpected rate of remission in this group of patients. This latter hypothesis is consistent with the observation that
BUDESONIDE AND ANTIBIOTICS FOR CROHN’S DISEASE
39
more patients who received antibiotics experienced diarrhea as an adverse effect of treatment. The adverse events observed in the study were generally of mild severity and are consistent with the known side effects of the drugs evaluated.32,33 Patients assigned to antibiotics were more likely to experience vaginitis, glossitis, and diarrhea, probably as a consequence of suppression of the normal bacterial flora. The low incidence of glucocorticoid-related adverse effects associated with budesonide therapy in our patients confirms the findings of other trials.2,3,28,29 In conclusion, this study shows that the addition of metronidazole and ciprofloxacin to budesonide is ineffective therapy for patients with active Crohn’s disease of the ileum, but this antibiotic combination may improve outcome in patients with disease involvement of the colon.
Appendix In addition to the authors, the following people at the following institutions participated in the trial (in parentheses, number of patients enrolled): Mount Sinai Hospital, Toronto, Ontario (24): G. Greenberg, A. H. Steinhart, and S. Mikolainis; Victoria General Hospital, Halifax, Nova Scotia (20): D. Leddin, D. MacIntosh, M. Kareemi, J. Love, S. van Zanten, J. Stewart, M. Edmunds, and H. McInnes; London Health Sciences Centre, London, Ontario (16): B. G. Feagan, J. W. D. McDonald, J. Gregor, W. Barnett, T. Ponich, J. Howard, M. Hopkins, H. Harrison, P. Walton-Mennill, and W. McCaw; Royal Victoria Hospital, Montreal, Quebec (14): A. Bitton, H. Vaupshas, and M. Bernier; St. Justine Hospital, Montreal, Quebec (14): E. Drouin, L. Chartrand, and A. Duhaime; Jewish General Hospital, Montreal, Quebec (11): A. Cohen and J. Rivard; McMaster University Medical Centre, Hamilton, Ontario (8): E. J. Irvine, B. J. Salena, K. Croitoru, S. M. Collins, and M. Donnelly; Saint John Regional Hospital, Saint John, New Brunswick (7): A. W. Cockeram and C. Arsenault; Health Sciences Centre, Winnipeg, Manitoba (6): C. Bernstein, B. G. Rosser, and P. Rawsthorne; Hotel Dieu de Quebec, Quebec City, Quebec (4): P. Pare´ and D. Lafrance; University of Alberta Hospitals, Edmonton, Alberta (4): R. Fedorak, J. McKaigney, E. Lalor, and D. Fisher; Ottawa Civic Hospital, Ottawa, Ontario (2): C. Dube and A. Port; Pasqua Hospital, Regina, Saskatchewan (2): J. McHattie and M. Scrivens; Health Sciences Centre, Calgary, Alberta (1): L. Sutherland and V. Todd; and St. Michael’s Hospital, Toronto, Ontario (1): J. Baker, R. Petroniene, N. Sayani, and G. Stewart.
40
STEINHART ET AL.
References 1. Summers RW, Switz DM, Sessions JT Jr., Becktel JM, Best WR, Kern F Jr., Singleton JW. National Cooperative Crohn’s Disease Study: results of drug treatment. Gastroenterology 1979;77: 847– 869. 2. Greenberg GR, Feagan BG, Martin F, Sutherland LR, Thomson ABR, Williams CN, Nilsson LG, Persson T. Oral budesonide as maintenance treatment for Crohn’s disease: a placebo-controlled, dose-ranging study. Gastroenterology 1996;110:45–51. 3. Rutgeerts P, Lofberg R, Malchow H, Lamers C, Olaison G, Jewell D, Danielsson A, Goebell H, Thomsen OO, Lorenz-Meyer H. A comparison of budesonide with prednisolone for active Crohn’s disease. N Engl J Med 1994;331:842– 845. 4. Candy S, Wright J, Gerber M, Adams G, Gerig M, Goodman R. A controlled double blind study of azathioprine in the management of Crohn’s disease. Gut 1995;37:674 – 678. 5. Feagan BG, Rochon J, Fedorak RN, Irvine EJ, Wild G, Sutherland L, Steinhart AH, Greenberg GR, Gillies R, Hopkins M. Methotrexate for the treatment of Crohn’s disease. N Engl J Med 1995; 332:292–297. 6. Elson CO. Commensal bacteria as targets in Crohn’s disease. Gastroenterology 2000;119:254 –257. 7. Duchmann R, May E, Heike M, Knolle P, Neurath M, Meyer zum Buschenfelde KH. T cell specificity and cross reactivity towards enterobacteria, bacteroides, bifidobacterium, and antigens from resident intestinal flora in humans. Gut 1999;44:812– 818. 8. Herfarth HH, Mohanty SP, Rath HC, Tonkonogy S, Sartor RB. Interleukin 10 suppresses experimental chronic, granulomatous inflammation induced by bacterial cell wall polymers. Gut 1996; 39:836 – 845. 9. McKay DM. Intestinal inflammation and the gut microflora. Can J Gastroenterol 1999;13:509 –516. 10. Strober W, Ludviksson BR, Fuss IJ. The pathogenesis of mucosal inflammation in murine models of inflammatory bowel disease and Crohn’s disease. Ann Intern Med 1998;128:848 – 856. 11. Atreya R, Mudter J, Finotto S, Mullberg J, Jostock T, Wirtz S, Schutz M, Bartsch B, Holtmann M, Becker C, Strand D, Czaja J, Schlaak JF, Lehr HA, Autschbach F, Schurmann G, Nishimoto N, Yoshizaki K, Ito H, Kishimoto T, Galle PR, Rose-John S, Neurath MF. Blockade of interleukin 6 trans signaling suppresses T-cell resistance against apoptosis in chronic intestinal inflammation: evidence in Crohn’s disease and experimental colitis in vivo. Nat Med 2000;6:583–588. 12. Bitton A, Peppercorn MA. Medical management of specific clinical presentations. Gastroenterol Clin North Am 1995;24:541– 548. 13. Peppercorn M. Is there a role for antibiotics as primary therapy in Crohn’s ileitis? J Clin Gastroenterol 1993;17:235–237. 14. Present DH. How to do without steroids in inflammatory bowel disease. Inflamm Bowel Dis 2000;6:48 –57. 15. Hanauer SB. Inflammatory bowel disease. N Engl J Med 1996; 334:841– 848. 16. Sartor RB. Antibiotics as therapeutic agents in Crohn’s disease. In: Bayless TM, Hanauer SB, eds. Advanced therapy of inflammatory bowel disease. Hamilton, Ontario: B.C. Decker, Inc., 2001:359 –362. 17. Best WR, Becktel JM, Singleton JW, Kern F Jr. Development of a Crohn’s disease activity index—National Cooperative Crohn’s Disease Study. Gastroenterology 1976;70:439 – 444. 18. Guyatt G, Mitchell A, Irvine EJ, Singer J, Williams N, Goodacre R, Tompkins C. A new measure of health status for clinical trials in inflammatory bowel disease. Gastroenterology 1989;96:804 – 810.
GASTROENTEROLOGY Vol. 123, No. 1
19. Irvine EJ, Feagan B, Rochon J, Archambault A, Fedorak RN, Groll A, Kinnear D, Saibil F, McDonald JW. Quality of life: a valid and reliable measure of therapeutic efficacy in the treatment of inflammatory bowel disease. Gastroenterology 1994;106: 287–296. 20. SAS system software. Version 8. SAS Institute Inc., Cary, NC, 1999. 21. Ursing B, Alm T, Ba´ra`ny F, Bergelin I, Ganrot-Norlin K, Hoevels J, Huitfeldt B, Jarnerot G, Krause U, Krook A, Lindstrom B, Nordle O, Rosen A. A comparative study of metronidazole and sulfasalazine for active Crohn’s disease: the Cooperative Crohn’s Disease Study in Sweden. Gastroenterology 1982;83:550 –562. 22. Ambrose NS, Allan RN, Keighley MRB, Burdon DW, Youngs D, Barnes P, Lennard-Jones JE. Antibiotic therapy for treatment in relapse of intestinal Crohn’s disease. A prospective randomized study. Dis Colon Rectum 1985;28:81– 85. 23. Sutherland L, Singleton J, Sessions J, Hanauer S, Krawitt E, Rankin G, Summers R, Mekhjian H, Greenberger N, Kelly M, Levine J, Thomson A, Alpert E, Prokipchuk E. Double blind, placebo controlled trial of metronidazole in Crohn’s disease. Gut 1991;32:1071–1075. 24. Rutgeerts P, Hiele M, Geboes K, Peeters M, Penninckz F, Aerts R, Kerremans R. Controlled trial of metronidazole treatment for prevention of Crohn’s recurrence after ileal resection. Gastroenterology 1995;108:1617–1621. 25. Prantera C, Zannoni F, Scribano ML, Berto E, Andreoli A, Kohn A, Luzi C. An antibiotic regimen for the treatment of active Crohn’s disease: a randomized, controlled clinical trial of metronidazole plus ciprofloxacin. Am J Gastroenterol 1996;91:328 –332. 26. Greenbloom SL, Steinhart AH, Greenberg GR. Combination ciprofloxacin and metronidazole for active Crohn’s disease. Can J Gastroenterol 1998;12:53–56. 27. Berg RD. The indigenous gastrointestinal microflora. Trends Microbiol 1996;4:430 – 435. 28. Campieri M, Ferguson A, Doe W, Persson T, Nilsson L-G. Global Budesonide Study Group. Oral budesonide is as effective as oral prednisolone in active Crohn’s disease. Gut 1997;41:209 –214. 29. Thomsen OO, Cortot A, Jewell D, Wright JP, Winter T, Veloso FT, Vatn M, Persson T, Pettersson E. A comparison of budesonide and mesalamine for active Crohn’s disease. N Engl J Med 1998; 339:370 –374. 30. McFarland LV. Epidemiology, risk factors and treatments for antibiotic-associated diarrhea. Dig Dis 1998;16:292–307. 31. Chasseny O, Michaux A, Bergmann JF. Drug-induced diarrhea. Drug Saf 2000;22:53–72. 32. Freeman CD, Klutman NE, Lamp KC. Metronidazole. A therapeutic review and update. Drugs 1997;54:679 –708. 33. Brogden RN, McTavish D. Budesonide: an updated review of its pharmacological properties, and therapeutic efficacy in asthma and rhinitis. Drugs 1992;44:375– 407.
Received March 15, 2002. Accepted March 21, 2002. Address requests for reprints to: A. Hillary Steinhart, M.D., Room 445, Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario, M5G 1X5 Canada. e-mail: [email protected]. Supported by the Crohn’s and Colitis Foundation of Canada and by AstraZeneca. The authors thank the patients who participated in the study, to Bayer, Inc., for providing ciprofloxacin and placebo tablets, to Apotex for providing metronidazole and placebo tablets, to AstraZeneca for providing budesonide, and to Beverley Jasevicius for assistance in preparing the manuscript.