82 Gut 1996; 39: 82-86

Budesonide prolongs time to relapse in ileal and ileocaecal Crohn's disease. A placebo controlled one year study Gut: first published as 10.1136/gut.39.1.82 on 1 July 1996. Downloaded from

R L6fberg, P Rutgeerts, H Malchow, C Lamers, A Danielsson, G Olaison, D Jewell, 0 0stergaard Thomsen, H Lorenz-Meyer, H Goebell, H Hodgson, T Persson, C Seidegard

Abstract are the most efficacious Background and Ains-To evaluate the medical treatment for active Crohn's disease efficacy and safety of the topical cortico- (CD)' with remission rates of 50 to 80 per cent budesonide, given in an oral con- reported after short or intermediate courses of trolled release formulation for treatment.2A Well known adverse effects (for maintenance ofremission in patients with example, acne, moon face, hirsutism, buffalo ileal and ileocaecal Crohn's disease (CD). hump) and systemic side effects (that is, impact Patients and Methods-Out of 176 patients on adrenal gland function, , with active CD who had achieved remis- impaired glucose tolerance) make continued sion (CD activity index score s,S0) after 10 conventional treatment with high weeks' treatment with either budesonide or doses inadvisable. Some patients achieving prednisolone, 90 were randomised to remission with the use of corticosteroids, how- continue with once daily treatment of6 mg ever, may benefit from longterm treatment with budesonide, or 3 mg budesonide or placebo low doses (<7.5 mg/day) of prednisolone or for up to 12 months in a double blind, similar. In the European Cooperative CD multicentre trial. Time to symptomatic study3 a subgroup of patients with active CD relapse was calculated using Kaplan-Meier being brought into remission with 6-methyl- estimates. Morning plasma cortisol was prednisolone fared better during low dose, measured at clinic visits and a corti- longterm maintenance treatment than patients cotropin stimulation test was performed treated with placebo.

after three months oftreatment. The potent and topically acting cortico- http://gut.bmj.com/ Results-Thirty two patients were allo- steroid budesonide has recently been proved to cated to the 6 mg budesonide group, 31 to be efficacious for induction of remission in the 3 mg group, and 27 to the placebo patients with ileal and ileocaecal CD in two group. After three months, 19 per cent of large controlled trials.5 6 Budesonide caused the patients in the 6 mg group had significantly less suppression of adrenal gland relapsed, compared with 45 per cent in the function, and consequently fewer cortico-

3 mg group and 44 per cent in the placebo steroid associated side effects, compared with a on September 23, 2021 by guest. Protected copyright. group (p=0.047). The corresponding standard regimen of prednisolone.5 results after 12 months was 59 per cent in The aim of this study was to determine if the 6 mg budesonide group, 74 per cent in budesonide given in an oral controlled release the 3 mg group, and 63 per cent in the preparation would safely prolong the time placebo group (p=0.44). The median time in remission, as induced by corticosteroids, to relapse or discontinuation was 258 days in CD patients during a one year, placebo in the 6 mg group, 139 days in the 3 mg controlled trial. group, and 92 days in the placebo group (p=0.021). Mean morning plasma cortisol values increased from entry in all three Methods groups with no statistically significant differences at 12 months. All 13 patients Study design remaining in the placebo group after three This was a randomised, double blind, months had a normal corticotropin placebo controlled study with three parallel stimulation response, compared with 18 groups. Patients entering the trial had been of 23 patients in the 6 mg, and 19 of 21 in treated during 10 weeks with either budes- the 3 mg budesonide groups (p=0.14). onide or prednisolone for active CD in a Acne and moon face were slightly more preceding study.5 Eligible patients for this common in the budesonide groups. study had to be in remission, defined as a CD Conclusion-6 mg budesonide once daily activity index (CDAI) score2 of 150 or less, is significantly more efficacious than after 10 weeks of treatment in the above men- Correspondence to: and were then Associate Professor placebo in prolonging time to relapse in tioned short-term study, they R Lofberg, Unit of CD, and causes only minor systemic side offered to continue with the relapse preven- Gastroenterology, Karolinska that started afterwards. Institute, Huddinge effects. tive part directly University Hospital, (Gut 1996; 39: 82-86) Patients were randomly allocated to treat- S-141 86 Huddinge, ment with either budesonide, 3 mg or 6 mg Sweden. once daily, or placebo for up to 12 months. Accepted for publication Keywords: budesonide, corticosteroid, corticotropin 31 January 1996 hormone, cortisol, Crohn's disease. The patients were followed up as outpatients Budesonide for maintenance ofremission in Crohn 's disease 83

with scheduled visits after three, six, nine, Laboratory assessments and 12 months of treatment. At each visit blood samples were taken for clinical chemistry and haematology analyses at each study centre (including: haemoglobin, Patients packed cell volume, white blood cell count

Patients aged 18 years or more, with an with differential, platelets, sodium, potassium, Gut: first published as 10.1136/gut.39.1.82 on 1 July 1996. Downloaded from established diagnosis of CD localised to the calcium, creatinine, glucose, bilirubin, alkaline terminal ileum or ileocaecal region were eligible phosphatase, alanine aminotransferase, aspar- for the trial. Patients gave their written or verbal tate aminotransferase, albumin, erythrocyte informed consent. sedimentation rate, C reactive protein). Patients with septic complications, active At each visit a blood sample was taken inflammation of the rectum, active systemic between 8 and 10 am for analysis of plasma infection or peptic ulcer disease were excluded. cortisol concentrations. A short corticotropin Patients with diabetes mellitus, or a random stimulation test was also performed after three blood glucose exceeding 8 mmol per litre were months in the study. Samples for assessment also excluded, as were patients with signifi- of plasma cortisol values were taken 30 and cant hepatic, renal or cardiovascular disease. 60 minutes after an intravenous bolus injec- Patients were not eligible if they were receiving tion of 0.25 mg synthetic corticotropin total parenteral, enteric or polymeric nutrition, hormone (Synacthen). All plasma samples if they had an ileostomy or previous small were analysed in a blinded manner at Astra bowel resection exceeding 100 cm. Draco, Lund using a high performance liquid chromatography method.8 Study drugs The budesonide gelatin capsules (Entocort Treatment plan capsules, Astra Draco AB, Lund, Sweden) Patients included in the study were randomly contained acid stable microgranules of budes- allocated to one of three treatment groups; 6 onide in ethylcellulose coated with a layer of mg budesonide daily, or 3 mg daily, or placebo acrylic based resin (Eudragit L100-55), which for up to 12 months. The patients were dissolves at a pH above 5-5. The formulation randomised at each centre in blocks of six. was designed to release budesonide during Eleven centres in six European countries passage through the ileum and the ascending participated in the trial. colon.7 Each capsule contained 3 mg ofbudes- Patients were treated until a symptomatic onide. Placebo capsules of identical appear- relapse as judged by an increase of the CDAI ance were manufactured by Astra Draco. The score occurred, either as calculated at a planned http://gut.bmj.com/ drugs were provided in blister packages, which or at an extraordinary visit. Treatment was dis- had to be returned at the next visit to the clinic. continued if surgery for CD became necessary, Unopened blisters were counted to determine if a septic complication occurred, ifthe patient's compliance. Non-compliance was defined as condition significantly deteriorated or if the more than 25 per cent of doses unused. patient became pregnant. No concurrent active medication for CD such as other corticosteroids, sulphasalazine, on September 23, 2021 by guest. Protected copyright. olsalazine or 5-aminosalicylic acid, metronida- Statistical and ethical considerations zole or immunosuppressants was permitted Based on data from the European cooperative during the study. Antidiarrhoeal drugs such as CD study,3 a relapse rate ofat least 50 per cent loperamide, diphenoxylate or other opiates was expected in the placebo group after 12 were permitted. The use ofcholestyramine was months. With 30 patients in each treatment permitted if the dose was kept constant group there would be an 80 per cent probabil- throughout the study. ity of detecting a difference between any active treatment and placebo if the budesonide group only had a relapse rate of 13 per cent. Clinical assessments The analyses were based on all patients The patients' CDAI score was calculated at treated using the last value extended princi- each visit to the clinic. Patients were provided ple, which was applied from the visit at three with diary cards to record their general well months to the final visit at 12 months. The being, number of loose stools, abdominal pain primary variables studied were relapse rate or fever. Relapse was defined as CD activity and time to relapse or treatment discontinua- index score > 150 together with an increase ofat tion. One way analysis ofvariance was used to least 60 units from entry. Relapse in this context compare changes in CDAI and other quanti- also included any patient withdrawn from tative variables between the three groups. the study due to an acute deterioration The x2 test was applied to compare relapse that required other drug treatment or a surgical rates, and survival analysis in the form of procedure, whether or not a CDAI score the generalised Wilcoxon test was used for calculation was available at that time. analyses of time to relapse or discontinua- At each visit signs or symptoms suggestive of tion and time in study. p Values of 5 per corticosteroid associated side effects (moon cent or less were considered statistically face, acne, buffalo hump, striae, hirsutism, significant. ankle swelling) were recorded as were all other The study was approved by all the local adverse events. Pulse and were ethics committees and was performed in measured at each visit. accordance with the Declaration of Helsinki. 84 L/dflerg, Rutgeerts, Malchow, Lamers, Danielsson, Olaison, et al

TABLE I Demography and history of Crohn 's disease patients difference in proportions with corticosteroid associated side effects at entry. Prior treatment Treatment group was not found to have any impact on the 6 mg budesonide 3 mg budesonide Placebo primary outcome variables of the study, Patients (n) 32 31 27 however. Sex (male/female) 15/17 10/21 11/16 Mean age (y) 37 (21-71) 33 (18-69) 35 (22-52) Gut: first published as 10.1136/gut.39.1.82 on 1 July 1996. Downloaded from Mean (SD) CDAI at entry 88 (47) 104 (35) 107 (60) Mean (SD) disease duration (y) 6-7 (5.7) 7-4 (6.0) 8-7 (7.2) Withdrawalfrom treatment Mean (SD) duration of current attack (months) 4-0 (8.7) 8-3 (15-1) 5-5 (10-6) Of the 90 patients randomised a total of 30 Previous bowel resection (n/0/o) 15 (47) 10 (32) 10 (37) patients completed the 12 month study (Table Disease location ileum only (n/%) 23 (72) 22 (71) 21 (78) II). Most patients who stopped treatment did ileocaecal (n/0/o) 9 (28) 9 (29) 5 (19) so because of therapeutic failure. The lowest ascending colon (n/0/o) 0 0 1 (4) Remission induced by frequency of therapeutic failures was seen in prednisolone (n/0/o) 23 (72) 11 (35) 14 (52) the 6 mg budesonide group (47%) followed by budesonide (n/0/o) 9 (28) 20 (65) 13 (48) placebo (63% and 3 mg budesonide (68%). The differences were not statistically different (p=0 22). TABLE II Causes oftreatment withdrawals in the three study groups Two patients in the placebo group were 6 mg budesonide 3 mg budesonide Placebo withdrawn from the study because of adverse events; one patient suffered from constipation Randomised patients 32 31 27 Withdrawals and one became pregnant. Five other patients Adverse events 0 0 2 stopped treatment for various reasons; with- Therapeutic failure 15 21 17 Other reasons 3 1 1 drawal of informed consent (four patients) and Completed study 14 9 7 desire to become pregnant (one patient).

Results Clinical efficacy A total of 90 patients were randomised into After three months, 19 per cent of the patients the study (mean 8-2 patients per centre; range in the 6 mg group had relapsed, compared with 2-22) and all of them received medication. 45 per cent in the 3 mg group and 44 per cent Table I shows the patient characteristics. The in the placebo group (p=0.047). There were treatment groups were similar, apart from no statistically significant differences at six somewhat lower initial CDAI scores in the 6 (p=0.19) or nine (p=0.99) months of treat- mg budesonide group and a larger proportion ment. After 12 months the relapse rate was 59

of women in the 3 mg group (NS). Fifty three per cent in the 6 mg group, 74 per cent in the http://gut.bmj.com/ per cent of the patients had previously been 3 mg group, and 63 per cent in the placebo treated with prednisolone in the preceding group (p=0 44). short-term phase study.5 Despite the re- The median time to relapse was significantly randomisation in this maintenance study, prolonged in the 6 mg budesonide group; 271 there was a skewed allocation into the three days compared with 175 days in the 3 mg treatment groups regarding previous treat- group and 146 days in the placebo group

ment. Seventy two per cent of the patients in (p=0-041). When time to relapse or dis- on September 23, 2021 by guest. Protected copyright. the 6 mg budesonide group had previously continuation was calculated (Fig 1) a similar received prednisolone compared with 35 per result was obtained; 258 days in the 6 mg cent in the 3 mg group and 52 per cent in the group, 139 days in the 3 mg group, and 92 placebo group. This difference showed up as a days in the placebo group (p=0.021). There was also a significant difference with respect to time in study (p=0.004) in favour of the 6 mg 1 budesonide group. A separate analysis was performed with respect to prior treatment with either pred- nisolone or budesonide, but no difference in time to relapse or discontinuation was found 2._ (Fig 2). ._Cu3 6 mg budesonide 0 0. 0.5 [- Safety and tolerability Eu No clinically relevant changes in haematology or clinical chemistry variables were noted E within or between the study groups during the C-) study. Mean pulse rate and blood pressure were unchanged. The mean basal morning plasma cortisol concentrations increased in all three groups during the study. Forty two per I, cent of the patients in the 6 mg budesonide o 0 50 100 150 200 250 300 350 group had a plasma cortisol value below the Time (days) lower reference value (150 nmol per litre) at inclusion, compared with 50 per cent in the Figure 1: Kaplan-Meier estimates of the percentage ofpatients not experiencing a relapse or discontinuing treatment during the 12 month trial (analysis base,don allpatients treated). 3 mg group and 42 per cent in the placebo The difference between the groups was significant (p=0 021). group. By 12 months the corresponding figures Budesonide for maintenance ofremission in Crohn 's disease 85

- n had decreased to 30 per cent in the 6 mg group, 18 per cent in the 3 mg group, and 5 per cent in the placebo group (p=0 10). The response to the short corticotropin hormone stimulation test was considered to be

Cu normal if the basal plasma cortisol concentra- Gut: first published as 10.1136/gut.39.1.82 on 1 July 1996. Downloaded from .0 Prednisolone tion was at least 150 nmol per litre and either the 0 post-stimulation value at 30 or 60 minutes CL a1) 0).5 increased by at least 200 nmol per litre, or had Budesonide increased to more than 400 nmol per litre. With this definition, five of 23 (22 per cent) patients E in the 6 mg group had an abnormal response U compared with two of 21 (10 per cent) in the 3 mg group and none ofthe remaining 13 patients in the placebo group (p=0-14). A direct com- parison between the 6 mg group and the placebo group didnot showa statisticallysignificantdif- o 0 50 100 150 200 250 300 350 ference (p=0 07). Only one patient in each of Time (days) the 6 and 3 mg budesonide groups had a post- Figur e 2: Kaplan-Meier estimates ofthe percentage ofpatients not experiencing a relapse or stimulation increase of plasma cortisol less than disconitinuing treatment during the 12 month trial (analysis based on allpatients treated) 200 nmol per litre (Table III). with rrespect to treatment with either prednisolone or budesonide prior to randomisation Two patients, one in the 6 mg budesonide (NS) group and one in the 3 mg group developed renal colic and were admitted to hospital. Both patients had a previous history of renal calculi. Overall, the incidence of minor adverse events was similar in the three treatment groups.

TABLE III Adrenal gland response to corticotropin stimulation in seven patients with baseline plasma cortisol values (0 minute) below the lower reference limit 150 nmolper litre Corticosteroid associated side effects after 3 months oftreatment of 6 mg or 3 mg budesonide daily Figure 3 shows the proportion of patients Plasma cortisol values experiencing corticosteroid associated side effects. Moon face was the most frequently 0 minute 30 minutes 60 minutes encountered side effect (16 patients) followed

3 mg budesonide acne http://gut.bmj.com/ Patient 4304 39 99 105 by (nine patients). Other corticosteroid Patient 4813* 22 441 527 associated side effects were rare, and only four 6 mg budesonide more two Patient 4107* 83 571 718 patients had than side effects in the 6 Patient 4120* 119 577 1016 mg budesonide group versus three patients in Patient 4912 131 157 NA Patient 5101* 137 417 552 the budesonide 3 mg group and two patients in Patient 5104* 63 445 488 the placebo group. No patient displayed a full Cushingoid appearance. *=Normal post-stimulation value (>400 nmol per litre). NA=no data available. The patient group treated with 6 mg budes- on September 23, 2021 by guest. Protected copyright. onide had the highest incidence of cortico- steroid associated side effects. The differences can be explained by differences in time in study. The number of recorded corticosteroid associated side effects decreased successively in the three groups during the study. 40 El Prednisolone a Budesonide Discussion In the National Cooperative CD Study2 no 30 effect of in maintaining remission could be demonstrated. In the European a) Cooperative CD Study3 a subgroup ofpatients 4C responding to corticosteroids in the short-term a1) 20 phase, had a lower relapse rate during follow a.) up treatment with low doses of 6-methyl CL prednisolone compared with patients treated with placebo. 10 This dedicated relapse prevention study shows that the topically acting corticosteroid budesonide has a beneficial effect in prolong- an stage in patients with ileal n ing asymptomatic 0 3 6 9 12 0 3 6 9 12 0 3 6 9 12 or ileocaecal CD. Budesonide 6 mg given once 6 mg budesonide 3 mg budesonide Placebo daily for up to one year was significantly more Time (months) efficacious than placebo in prolonging the time to a symptomatic relapse of the disease. Figure 3: Percentage ofpatients with corticosteroid associated side effects at study entry, and after three, six, nine, and 12 months in the study. The treatment in the preceding The patients enrolled in this study could short-term study is shown in each bar. be categorised as 'high risk patients'9 for L8fberg, Rutgeerts, Malchow, Lamers, Danielsson, Olatson, et al subsequent relapse, as they had been brought Adverse events during the study were mild into remission with the use of corticosteroids at and did not differ between the budesonide and most 10 weeks before entering this main- placebo treated groups. Systemic influence, tenance trial. In this context the prolongation of such as, for example, hypertension, hypergly- the time to relapse was substantial, and may be caemia, or an increase in the white blood cell

of considerable importance for an individual count, was not seen during the trial. This Gut: first published as 10.1136/gut.39.1.82 on 1 July 1996. Downloaded from patient. In this high risk group ofpatients, how- supports the use ofbudesonide when longterm ever, budesonide did not have any clear impact treatment is considered. on the overall relapse rate after six or 12 We conclude that 6 mg budesonide once months of treatment. The result is in line with daily, given in an oral controlled ileal release the most recent experiences using 5-ASA com- preparation, significantly prolongs the time to pounds for the same purpose as presented in relapse after corticosteroid induced remission preliminary reports.9 10 It is inherently difficult, in CD patients compared with placebo. though, to compare the results of maintenance Budesonide was well tolerated during trials in CD as fundamental factors such as longterm treatment, causing only limited patient cohorts, extent of disease, modes of impact on the adrenal gland function. achieving remission, and duration of remission From the Unit of Gastroenterology, Huddinge University before entering a trial may vary considerably. Hospital, Sweden (RL), the Department of Gastroenterology, Leuven University Hospital, Belgium (PR), Akademisches The plateau of the relapse rate seen in the Lehrkrankenhaus der Univerzitat zu K6ln, Leverkusen, placebo group towards the end of this study is Germany (HM), the Department of Gastroenterology and Hepatology, University Hospital, Leiden, the Netherlands intriguing. As the number ofpatients that were (CL), the Medical Department, University Hospital, Umea, in remission after 12 months was comparably Sweden (AD), the Department of Gastroenterology, University Hospital, Linkoping (GO), the Gastroenterology Unit, small in all the three treatment groups, this Radcliffe Infirmary, Oxford, United Kingdom (DJ), the may make comparisons of the relapse rates at Medical Department C, Herlev University Hospital, Denmark (00T), Stadsiches Krankenhaus Friedrichshafen, Germany this stage unreliable. (HL-M), the Medical Department, Hammersmith Hospital, The mode in which remission had been London, United Kingdom (HH), and the Departments of Biostatistics and Data Processing (TP) and Clinical Research achieved (using budesonide or prednisolone) and Development (CS), Astra Draco, Lund, Sweden. prior to randomisation did not influence the This study was supported by Astra Draco. This study was reported in abstract form at the 1994 AGA efficacy of budesonide in prolonging time to meeting in New Orleans. relapse. Thus, prednisolone did not have carry over properties for longterm relapse prevention 1 Meyers S, Sachar DB. Medical management of Crohn's in this particular group of patients as compared disease. Hepatogastroenterology 1990; 37: 42-55. with budesonide. 2 Summers RW, Switz DM, Sessions JT Jr, Becktel JM, Best WR, Kern F, et al. National Cooperative Crohn's Disease However, corticosteroid associated side Study: Results of drug treatment. Gastroenterology 1979; 77: 847-69. effects were more common in patients pre- http://gut.bmj.com/ 3 Malchow H, Ewe K, Brandes JW, Goebell H, Ehms H, viously treated with prednisolone before entry, Sommer H, et al. European Cooperative Crohn's Disease but the influence diminished gradually during Study (ECCDS): results of drug treatment. Gastroenterology 1984; 86: 249-66. the study period. No serious corticosteroid 4 Munkholm P, Langholz E, Davidsen M, Binder V. associated side effects or adverse events were Frequency ofglucocorticoid resistence and dependency in Crohn's disease. Gut 1994; 35: 360-2. seen. Only two of 44 patients (5 per cent) 5 Rutgeerts P, Lofberg R, Malchow H, Lamers C, Olaison G, treated with budesonide (one on 6 mg and one Jewell D, et al. A comparison of budesonide with pred- nisolone for active Crohn's disease. N Engl J Med 1994; on 3 mg), had an impaired response to corti- 331: 842-5. on September 23, 2021 by guest. Protected copyright. cotropin stimulation after three months of 6 Greenberg G, Feagan B, Martin F, Sutherland LR, Thomson ABR, Williams N, et al. Oral budesonide for the treatment. The adequate response to a short treatment of active Crohn's disease. N Engl Jf Med 1994; corticotropin test does not exclude a distur- 331: 836-41. 7 Edsbacker S, Wollmer P, Nilsson A, Nilsson M. bance of the hypothalamic-pituitary-adreno- Pharmacolinetics and gastrointestinal transit ofbudesonide cortical axis,11 but makes this less probable. controlled ileal release (CIR) capsules. Gastroenterology 1993; 104: A695. As the study was designed with respect to 8 Van den Berg JHM, Mol ChR, Deelder RS, Thijssen JHH. efficacy rather than to side effects, it did A quantitative assay of cortisol in human plasma by high performance liquid chromatography using a selective not have enough statistical power to detect chemically bonded stationary phase. Clin Chim Acta 1977; differences in side effects or changes in corti- 78: 165-72. 9 Modigliani R, Colombel JF, Dupas JL, Dapoigny M, cotropin stimulated response. It is still possible Bouhnik Y, Veyrac M, et al. Mesalamine in Crohn's that budesonide may exert adrenal gland disease patients with prednisolone induced remission: effect on steroid weaning and remission maintenance in suppression in some people when given in a the year after weaning. A placebo-controlled, multicenter dose of 6 mg daily long term, although the study. Gastroenterology 1995; 108: A878. 10 Sutherland LR, Martin F, Bailey RJ, Fedorak R, Dallaire C, impact seems to be limited. Rossman R, et al. 5-aminosalicylic acid (Pentasa) in the Other important issues, such as impact on maintenance of remission of Crohn's disease. Gastroenterology 1995; 108: A924. bone metabolism, were not considered in this 11 Christy NP. Pituitary-adrenal function during cortico- study. However, an eight week study compar- steroid therapy. Learning to live with uncertainty. N Engi _J Med 1992; 326: 266-7. ing budesonide and prednisolone enema treat- 12 Lofberg R, 0stergaard Thomsen 0, Langholz E, Schioler ment in active distal ulcerative colitis, has R, Danielsson A, Suhr 0, et al. A comparative study of budesonide versus prednisolone retention enema in active indicated that budesonide has less effect on the distal ulcerative colitis. A Scandinavian multicenter study. bone specific protein osteocalcin.12 AlimentPharmacol Ther 1994; 8: 623-9.