Chester Clinic

 

Doi:10.1016/s0140-6736(08)60304-9

Early combined immunosuppression or conventional
management in patients with newly diagnosed Crohn's
disease: an open randomised trial
Geert D'Haens, Filip Baert, Gert van Assche, Philip Caenepeel, Philippe Vergauwe, Hans Tuynman, Martine De Vos, Sander van Deventer, Larry Stitt, Allan Donner, Severine Vermeire, Frank J Van De Mierop, Jean-Charles R Coche, Janneke van der Woude, Thomas Ochsenkühn, Ad A van Bodegraven, Philippe P Van Hootegem, Guy L Lambrecht, Fazia Mana, Paul Rutgeerts, Brian G Feagan, Daniel Hommes, for the Belgian Infl ammatory Bowel Disease Research Group and the North-Holland Gut Club Lancet 2008; 371: 660–67
Background Most patients who have active Crohn's disease are treated initially with corticosteroids. Although this
See Comment page 635
approach usually controls symptoms, many patients become resistant to or dependent on corticosteroids, and long
Imelda Gastrointestinal Clinical
exposure is associated with an increased risk of mortality. We aimed to compare the eff ectiveness of early use of
Research Centre, Bonheiden,
combined immunosuppression with conventional management in patients with active Crohn's disease who had not
Belgium (G D'Haens MD); H Hart
previously received glucocorticoids, antimetabolites, or infl iximab.
Ziekenhuis, Roeselare, Belgium
(F Baert MD); University Hospital
Gasthuisberg, Leuven, Belgium
Methods We did a 2-year open-label randomised trial at 18 centres in Belgium, Holland, and Germany between
(G van Assche MD, S Vermeire MD, May, 2001, and January, 2004. We randomly assigned 133 patients to either early combined immunosuppression or
Prof P Rutgeerts MD); Ziekenhuis
conventional treatment. The 67 patients assigned to combined immunosuppression received three infusions of
Oost-Limburg, Genk, Belgium
infl iximab (5 mg/kg of bodyweight) at weeks 0, 2, and 6, with azathioprine. We gave additional treatment with
(P Caenepeel MD); Groeninge
Ziekenhuis, Kortrijk, Belgium
infl iximab and, if necessary, corticosteroids, to control disease activity. 66 patients assigned to conventional
(P Vergauwe MD); Medisch
management received corticosteroids, followed, in sequence, by azathioprine and infl iximab. The primary outcome
Centrum Alkmaar,
measures were remission without corticosteroids and without bowel resection at weeks 26 and 52. Analysis was by
The Netherlands
modifi ed intention to treat. This trial was registered with ClinicalTrials.gov, number NCT00554710.
(H Tuynman MD); University
Hospital, Gent, Belgium
(M De Vos MD); Academic
Findings Four patients (two in each group) did not receive treatment as per protocol. At week 26, 39 (60·0%) of
Medical Centre, Amsterdam,
65 patients in the combined immunosuppression group were in remission without corticosteroids and without
The Netherlands
surgical resection, compared with 23 (35·9%) of 64 controls, for an absolute diff erence of 24·1% (95% CI 7·3–40·8,
(S van Deventer MD);
Department of Epidemiology
p=0·0062). Corresponding rates at week 52 were 40/65 (61·5%) and 27/64 (42·2%) (absolute diff erence 19·3%,
and Biostatistics, University of
95% CI 2·4–36·3, p=0·0278). 20 of the 65 patients (30·8%) in the early combined immunosuppression group had
Western Ontario, Canada
serious adverse events, compared with 19 of 64 (25·3%) controls (p=1·0).
(L Stitt MSc, Prof A Donner PhD, Prof B G Feagan MD); Robarts
Clinical Trials, Robarts Research
Interpretation Combined immunosuppression was more eff ective than conventional management for induction of
Institute, London, Ontario,
remission and reduction of corticosteroid use in patients who had been recently diagnosed with Crohn's disease.
Canada (A Donner, B G Feagan);
Initiation of more intensive treatment early in the course of the disease could result in better outcomes.
Algemeen Ziekenhuis
St Augustinus, Wilrijk, Belgium
(F J Van De Mierop MD); Clinique
Treatment directed towards tumour-necrosis factor St Pierre, Ottignies, Belgium
Crohn's disease is a chronic infl ammatory disorder of (TNF) has improved the management of refractory (J-C R Coche MD); Erasmus
the gastrointestinal tract. Current practice guidelines Crohn's disease.13–15 TNF antagonists, such as infl iximab, University Medical Centre,
Rotterdam, The Netherlands
recommend that most patients with active disease are conventionally reserved for patients who have failed, (J van der Woude MD);
should be treated initially with corticosteroids.1,2 in sequence, both corticosteroids and antimetabolites. Although this approach is usually eff ective for control In rheumatoid arthritis, however, which has many Universität, Munich, Germany
of symptoms, many patients become resistant to, or patho physiological similarities to Crohn's disease, the (T Ochsenkühn MD);
VU University Medical Centre
dependent on, these drugs.3 Long exposure to early introduction of TNF antagonists in combination Amsterdam, The Netherlands
corticosteroids is also associated with the complications with methotrexate has been shown to treat early disease (A A van Bodegraven MD);
of Cushing's syndrome, and therefore with an increased better than does monotherapy with either agent.16–18 Algemeen Ziekenhuis St Lucas,
risk of mortality.1,2,4–6 For this reason, most clinicians Moreover, one randomised controlled trial has Brugge, Belgium
(P P Van Hootegem MD);
initiate treatment with corti costeroid-sparing drugs suggested that the combination of azathioprine and Algemeen Ziekenhuis Damiaan
such as azathioprine, mercapto purine, or methotrexate infl iximab in corticosteroid-dependent Crohn's disease Campus St Jozef, Oostende,
once corti costeroid-resistance or dependence develops, was more eff ective than azathioprine alone.19 On the Belgium (G L Lambrecht MD);
but initiation of these immuno suppressive drugs earlier basis of these observations, we did a randomised trial of Academisch Ziekenhuis VUB,
Jette, Belgium (F Mana MD);
in the course of the disease is not recommended.7–10 early combined immunosuppression in patients with Leiden University Medical
However, since these antimetabolites are only recently diagnosed Crohn's disease. We aimed to Centre, Leiden, Netherlands
moderately eff ective,1,7,10–12 repeated or long courses of investigate the eff ectiveness of short-term infl iximab corticosteroids are frequently given.
combined with azathioprine or 6-mercaptopurine in www.thelancet.com Vol 371 February 23, 2008
patients with active Crohn's disease who were receiving day 0 onwards. If a patient responded to and tolerated Correspondence to: Geert D'Haens, Department of induction therapy with corticosteroids.
both drugs, azathioprine was continued for the duration Gastroenterology, Imelda of the trial. Patients who were intolerant to azathioprine Gastrointestinal Clinical Research were given subcutaneous methotrexate (Ledertrexate, Centre, Imelda General Hospital, Study design and participants
Wyeth Lederle, Seattle, WA, USA) at an initial dose of Imeldalaan 9, B-2820 Bonheiden, Belgium We did an investigator-initiated trial at 18 centres in 25 mg each week for 12 weeks with the dose reduced to [email protected]
Belgium, Holland, and Germany between May, 2001, 15 mg per week thereafter. We defi ned response and January, 2004. The investigational review board at according to the CDAI score: for patients with an initial each of these centres approved the protocol. All patients score between 200 and 250 points, a 50-point decrement gave written informed consent before random was regarded as a response; corresponding criteria for assignment.
patients with scores between 250 and 350 points and We defi ned eligible patients as those who were aged scores greater than 350 points were 75 and 100 points, 16–75 years; who had been diagnosed with Crohn's respectively. After initial treatment, patients whose disease within the past 4 years; and who had not symp toms worsened (a CDAI increase of greater than previously received corticosteroids, antimetabolites, or 50 points, to give a score greater than 200) were given biological agents. We defi ned active disease as a Crohn's additional infusions of infl iximab. If symptoms disease activity index (CDAI)20 score of greater than persisted, we initiated methyl prednisolone (Medrol, 200 points for a minimum of 2 weeks before Upjohn, Kalamazoo, MI, USA) and continued randomisation. We excluded any patients who had an azathioprine or methotrexate.
immediate need for surgery; symptomatic stenosis or Patients in the conventional management group ileal or colonic strictures with prestenotic dilatation; were treated according to usual clinical practice and cur-signs, symptoms, or laboratory tests that indicated rent guidelines. Patients received induction treat ment severe comorbidity; documented chronic infection; a with either methylprednisolone (Medrol, Upjohn, positive stool culture for pathogens; a positive tuberculin Kalamazoo, MI, USA) or budesonide (Budenofalk, test or a chest radiograph consistent with tuberculosis; FalkPharma, Freiburg, Germany and Entocort or a malignancy. We also excluded any patient who was AstraZeneca, Lund, Sweden). For those who responded allergic to murine proteins, was pregnant, or was a to these treatments, the dose of corticosteroid was substance abuser.
tapered. For methylprednisolone, an initial daily dose 2 weeks before randomisation, eligible patients were given a physical examination; blood tests, including for C-reactive protein; and a skin test for tuberculin. We 140 patients screened obtained a chest radiograph and a stool sample from each patient. Patients were instructed about the use of the infl ammatory bowel disease questionnaire (IBDQ)21 3 had CDAI <200 and of a diary card to score the CDAI. The disease activity index generates a score between 0 and 600, 2 needed immediate where scores of 150 or less defi ne clinical remission. The IBDQ is a disease-specifi c instrument that measures quality of life. Scores range from 32 to 224, 133 patients randomised and higher scores indicate better quality of life.21 Procedures
We randomly assigned patients in blocks of four
67 assigned to early combined 66 assigned to conventional according to a computer-generated schedule. The immunosuppression investigator who generated the randomisation schedule was independent from the rest of the trial. We used a 14 withdrew prematurely 17 withdrew prematurely 6 had bowel resection 8 had bowel resection minimisation procedure to balance diff erences between 4 lost to follow-up treatment groups in prognostic factors (baseline CDAI 1 did not receive 2 did not receive score, cigarette smoking, and disease location). allocated treatment allocated treatment Allocation was not concealed from investigators or Patients assigned to early combined immuno- suppression received three infusions of infl iximab (Remicade, Centocor, Malvern, PA, USA) in doses of 53 analysed after 2-year follow-up 49 analysed after 2-year follow-up 5 mg/kg bodyweight at weeks 0, 2, and 6, in combin-
ation with azathioprine (Imuran, GlaxoSmithKline,
Middlesex, UK) in doses of 2–2·5 mg/kg per day from Figure 1: Trial profi le
www.thelancet.com Vol 371 February 23, 2008
azathioprine and methotrexate also received infl iximab, Early combined
Conventional p value
without antimetabolite treatment. We repeated the infusion on relapse of symptoms in these patients. We assessed patients at a clinic at weeks 2, 6, 10, 14, and 26, and at every 12 weeks thereafter for 104 weeks. At each visit, we calculated a CDAI score, obtained blood for chemical analyses, did a physical examination, Weeks from diagnosis 2·0 (1·0–5·0) 2·5 (1·0–11·0) 0·65† and recorded drug treatments and doses. Patients completed the IBDQ at each visit. At eight of 18 centres, patients underwent ileocolonoscopy at week 104. Lesions on endoscopy were scored with a validated index22 consisting of a four-point scale (in which 0=no ulcers, 1=aphthoid ulcers, 2=large ulcers, and 3=ulcerated sten osis) that assessed fi ve defi ned regions of the bowel, for a composite score between 0 and 15.
Since long exposure to corticosteroids has detrimental eff ects, complete corticosteroid withdrawal is generally seen as an essential element of the clinical defi nition of remission.23 Therefore, we defi outcome as a CDAI score of less than 150 points, absence of corticosteroid treatment, and no intestinal resection. Secondary measures included the time to C-reactive protein relapse after successful induction treatment; mean concentration (mg/L) CDAI and IBDQ scores; median concentrations of Data are number (%), mean (SD), or median (IQR) unless otherwise specifi ed. *χ² test serum C-reactive protein; and mean endoscopic for dichotomous variables. †Student's t test for continuous variables. ‡Fisher's exact severity scores. Other secondary outcomes were the test. §Crohn's Disease Activity Index scores range from 0 to 600; higher scores indicate greater disease activity. ¶Infl ammatory Bowel Disease Questionnaire scores proportion of patients who were in remission (CDAI range from 32 to 224; higher scores indicate better health-related quality of life. <150 and no corticosteroid therapy) at week 14; the proportion given infl iximab, methyl prednisolone, and Table 1: Baseline characteristics
bolites at any time during the study; the proportion without ulcers after 24 months of treatment; of 32 mg was prescribed for 3 weeks, followed by and the daily dose of methylprednisolone.
tapering by 4 mg per week to discontinuation. Patients who received bude sonide were prescribed 9 mg per day for 8 weeks with tapering to discontinuation by 3 mg For the primary analysis, we used Pearson's χ2 test to per week thereafter. Hence treatment with either drug test the hypothesis that the rate of remission was not lasted 10 weeks. If a patient's symptoms worsened diff erent between the two treatment groups. To adjust during the course of corticosteroid tapering, we for comp arisons (of weeks 26 and 52), we prespecifi ed increased the dose to the initial dose and repeated the the alpha errors to declare statistical signifi cance at induction treatment. If their symptoms continued to these times as 0·01 and 0·048, respectively.24 We worsen despite this manoeuvre, we introduced calculated nominal p values for comparisons at other azathioprine (2–2·5 mg/kg per day). time points. The time to relapse was compared by the Patients who relapsed after withdrawal of log-rank test. We compared diff erences in the CDAI corticosteroids were given a second course of and IBDQ scores by use of repeated-measures analyses corticosteroids in combination with azathioprine. For of variance. The use of methylprednisolone was patients who failed 4 weeks of corticosteroid treatment, described by calculating the 95th percentile of the daily we increased the methyl prednisolone dose to 64 mg per dose. We assessed total exposure to corticosteroids by day and added azathioprine. We gave 64 mg methyl- examining the distribution of the average daily dose of prednisolone per day for 2 weeks and then tapered this methylprednisolone, calculated by estimating the dose by 8 mg per week to a daily dose of 32 mg. cumulative dose for each group and dividing this Thereafter, methyl prednisolone was tapered by 4 mg number by the total number of patient days of follow-up. each week. Any patients who remained symptomatic We compared the change in the median serum after 16 weeks of azathioprine treatment received an concentration of C-reactive protein by the Wilcoxon induction course of infl iximab (5 mg/kg bodyweight at Mann-Whitney test; changes in endoscopy scores by weeks 0, 2, and 6) and continued antimetabolite the Wilcoxon test; and the proportion of patients treatment. Patients who relapsed despite the use of without ulceration on endoscopy by the χ2 test. Fisher's methotrexate or those who were intolerant to both exact test was used to compare the rate of adverse www.thelancet.com Vol 371 February 23, 2008
events. We used two-sided tests for signifi cance. We management, an absolute diff erence of 19·4% (95% CI analysed patients who were treated as per protocol in a 2·4–36·3, p=0·0278). After week 52, the proportion of modifi ed intention-to-treat analysis.
patients in remission did not diff er between the two We anticipated that 40% of patients assigned to the groups (fi gure 2).
conventional treatment algorithm would enter clinical The median time to relapse after successful induction remission, and therefore that that we would need a therapy at week 14 was longer for patients assigned to sample size of 130 patients to give 80% power to detect early immunosuppression (329·0 days, IQR 91·0–not an absolute diff erence of 25% between the groups. This reached) than for controls (174·5 days, IQR 78·5–274·0, trial was registered with ClinicalTrials.gov, number p=0·031) (fi gure 3). NCT00554710. Patients assigned to early immunosuppression were exposed to substantially less methylprednisolone than Role of the funding source
were those in the conventional management group Financial support for data monitoring (DRC, Wetteren, (fi gure 4). Budesonide use was minimal in both groups. Belgium) was provided by Centocor BV and Schering At week 52, 2% of patients assigned to early combined Plough, who also provided infl iximab. Robarts Clinical immunosuppression were receiving budesonide, com-Trials analysed the data (Robarts Research Institute, University of Western Ontario, London, Ontario, Canada). All authors had access to the data and jointly Early combined immunosuppression group decided to submit the manuscript.
Conventional management group Results
Figure 1 shows the trial profi le. Of the 133 patients who
were randomly assigned, four did not receive treatment as per protocol. One patient in the early combined of patients in remission (%) intervention group had a gastric carcinoma, and one in the conventional management group had ulcerative colitis. One patient in each group was not willing to accept the treatment to which they had been assigned. 65 patients had combined immunosuppression and 64 had conventional management. Baseline
characteristics of the two groups were similar (table 1), Figure 2: Proportions of patients in remission
although patients assigned to conventional treatment Remission was defi ned as a score of less than 150 on the Crohn's Disease Activity
Index, absence of a bowel resection, and complete withdrawal of corticosteroid had better quality of life scores.
treatment with budesonide or methylprednisolone. Primary endpoints were the Four patients, all in the conventional management proportion of patients in remission at weeks 26 and 52. p values were derived by group, were lost to follow-up. Three patients did not Pearson's χ² test. p values of less than 0·01 and less than 0·048 were considered statistically signifi cant for the week 26 and week 52 comparisons, respectively.
comply with the treatment protocol and three withdrew consent after randomisation. Nine patients withdrew from the group assigned to combined immuno- Early combined immunosuppression suppression, compared with eight controls. Most patients Conventional management withdrew because they had bowel resection for Crohn's disease. One patient in the treatment group withdrew because of severe disease exacerbation and another because of demyelination; one control withdrew because of glucocorticoid intolerance. The baseline characteristics of the patients who underwent ileocolonoscopy were not diff erent from those of the overall study population.
By week 14, a greater proportion of patients in the bined immunosuppression group were in Proportion of patients with no relapse (%) remission than were patients given conventional treatment (p=0·0001; fi gure 2). After 26 weeks, Time after randomisation (weeks) 39/65 (60·0%) of patients given combined Number at risk
Early combined immunosuppression immunosuppression were in remission, com Conventional management with 23/64 (35·9%) controls (p=0·0062), an absolute diff erence of 24·1% (95% CI 7·3–40·8). At 52 weeks, 40/65 (61·5%) in the early combined immuno- Figure 3: Proportion of patients who did not relapse
Kaplan-Meier estimates of the time to relapse after successful induction treatment at week 14. Relapse was defi ned
suppression group were in remission compared by a score of greater than 200 on the Crohn's Disease Activity Index, need for a bowel resection, or the need to add with 27/64 (42·2%) of those assigned to conventional additional treatment according to assigned regimen. The p value was calculated by the log-rank test.
www.thelancet.com Vol 371 February 23, 2008
76·0% of patients in the conventional treatment group Early combined immunosuppression were receiving an antimetabolite.
Conventional management At week 52, the overall proportion of patients receiving azathioprine and methotrexate was 77% and 25% respectively in the early combined immunosuppression group. Corresponding proportions for the conventional management group were 60% and 13%, respectively.
methylprednisolone (%) Proportion of patients given After patients in the early immunosuppression group had completed their induction course of infl iximab, the proportion of patients on infl iximab was similar in the two groups (fi gure 4). 24 (36·9%) patients in the early combined immunosuppression group needed at least one additional dose of infl iximab, compared with 9 (14·0%) patients in the conventional management group.
At week 10, patients assigned to combined immuno- antimetabolites (%) sup pression showed a more rapid drop in CDAI scores; Proportion of patients given the mean reduction was 231 (SD 123) points, compared with 178 (116) points in controls (diff erence 53·3, 95% CI 9·2–97·4, p=0·0184). After week 10, mean scores in both groups were similar and consistently below 150 points. Results from the IBDQs were similar. For patients assigned to early combined immuno- suppression, the mean IBDQ score at week 10 increased by 59·2 (SD 36·6) points from baseline, compared Proportion of patients given with 37·4 (32·8) points in con trols (diff erence Weeks after randomisation 21·8 points, 95% CI 8·7–34·9, p=0·0014).
Patients who were assigned to the combined immuno- Figure 4: Proportion of patients who received methylprednisolone (A),
antimetabolites (B), and infl iximab (C)
suppression strategy had a more rapid reduction in the Proportions were estimated using 4-week windows for use of corticosteroids median serum concentration of C-reactive protein at (either methylprednisolone or budesonide) and antimetabolites and an 8-week week 10 than did con trols (−15·0 mg/L, IQR –52·0 to –2·1 window for infl iximab use.
vs −4·2 mg/L, –25·0 to 1·0, p=0·0244).
At week 104, no ulcers were seen for 19/26 pared with 7% of those in the conventional management (73·1%) patients assigned to the combined immuno- group. The 95th percentile of the daily methyl- suppression group, compared with 7/23 (30·4%) of prednisolone dose was 35 mg for patients assigned to controls (p=0·0028). The corresponding endoscopy conventional management and 0 mg for those assigned scores were 0·7 (SD 1·5) and 3·1 (2·9) (p=<0·001). to early com bined immunosuppression.
Endoscopic healing was not associated with Conversely, fi gure 4 shows that patients assigned to CDAI-defi ned remission.
combined immunosuppression received consistent Table 2 shows adverse events. In the combined treatment with antimetabolites (azathioprine and immuno suppression group, six patients had a bowel metho trexate). Nevertheless, by the end of the trial resection and one had a fi stulotomy. One 25-year old female patient in the combined immunosuppression group developed loss of sensation in her left leg, and Early combined
p value†
MRI showed demyelination in the conus medullaris. immunosuppression (n=65) management (n=64)
Infl iximab was discontinued, after the patient had Serious adverse events
received four infusions. The symptoms resolved and she had no recurrence. Two cases of asymptomatic neutropenia occurred in the combined immuno- suppression group. Eight controls needed surgery for intestinal complications of Crohn's disease and seven were operated on for a perianal abscess or fi stula. All Bowel obstruction eight cases of glucocorticoid-related adverse events Demyelinating disease (four of moon facies, two of acne, one of hyperphagia, Perianal abscess or fi stula and one of mood swings) were in the conventional management group. Nine women were pregnant; four (Continues on next page) of their 10 pregnancies led to miscarriage (two in each group).
www.thelancet.com Vol 371 February 23, 2008
Early combined
p value†
Treatment algorithms in early Crohn's disease and their immunosuppression (n=65) management (n=64)
eff ect on long-term outcomes have not been studied in (Continued from previous page) randomised trials. We have shown that in patients with Crohn's disease who had not previously received cortico- steroids, antimetabolites, or biologicals, use of early combined immunosuppression resulted in remission more quickly than did treatment according to existing Worsening of Crohn's disease consensus guidelines.25–28 Although conventional guide- lines support the use of corticosteroids as fi rst-line treatment, they also recommend that the duration of this treatment should be limited to a short period, usually 3 to 4 months. By 6 months, however (or Elevated liver function tests 26 weeks as in our study) patients should have dis con- steroids to avoid toxic eff ects. Patients assigned to early combined immunosuppression also Upper respiratory tract infection had more rapid normalisation of serum C-reactive protein and consist ently higher rates of remission than Vaginal infections did controls within the fi rst year of treatment.
Urinary tract infections Corticosteroid treatment is a major source of morbidity that is independently associated with an Dermatological disorders increased risk of mortality.5 We showed that remission rates in patients given combined immunosuppression were higher than those reported in other trials in which infl iximab was added to pre-existing treatments for Crohn's disease.13–15 This observation is consistent with the enhanced effi cacy of combination treatment, and Orthopaedic disorders might also refl ect the early initiation of treatment relative to the course of the disease.8 On average, Muscle cramps or myalgia Crohn's disease was diagnosed in study participants less than 4 months before randomisation. In rheumatoid Psychiatric disorders arthritis, early intervention with combined immuno- suppression has been shown to control disease activity more rapidly than conventional treatment, and to prevent progressive joint destruction and improve long term functional outcomes.16–18 Accordingly, patients who received early combined immunosuppression were less likely to have mucosal ulceration after 2 years of treatment. We speculate that signifi cant mucosal ulcer- Data are number (%), unless otherwise specified. *Adverse events that occurred in at least 5% of patients. ation in Crohn's disease is analogous to radio graph- †p values were calculated with Fisher's exact test. ‡All five cases of pancreatitis were associated with the use of azathioprine.
ically defi ned joint lesions in rheumatoid arthritis and that healing of these lesions could change the natural Table 2: Incidence of adverse events*
history of the disease. In support of this concept, the healing of ulcers has been previously associated with a metabolite treatment did not decrease the need for reduction in admissions to hospital and in surgery for surgery.32complications of Crohn's disease.15,29 Indeed, regulatory One controversial aspect of our study was that patients agencies now recommend that clinical trials should use in the combined immunosuppression group received remission with endoscopic healing as an endpoint.30 We intermittent treatment with infl iximab. Although noted diff erences in mucosal healing in the two groups, controlled trials in refractory patients have shown better even though patients assigned to conventional man- clinical and endoscopic results for scheduled treatment agement received more corticosteroids and in most every 8 weeks, our results suggest that intermittent cases were also receiving treatment with antimetabolites infl iximab use, in combination with antimetabolite (77·0% at week 104). These observations are consistent treatment, might be an eff ective strategy for treatment with the results of previous studies in which of newly diagnosed patients, since responsiveness was corticosteroid treatment was shown to be ineff ective for not diminished on repeat infusions, and no patients healing of intestinal ulcers,31 and sequential use of developed clinical manifestations of hypersensitivity to corticosteroids with delayed introduction of anti- the drug. A strategy of maintenance treatment with www.thelancet.com Vol 371 February 23, 2008
infl iximab every 8 weeks could potentially have greater consultant for Centocor. SV has received grant and research support eff ects, but was not yet standard practice when we from UCB, consultancy fees from AstraZeneca and Ferring, and initiated our trial.
speakers fees from Shering Plough, Ferring, and UCB, and has been on the Advisory Committee for Shire, Ferring, and UCB. TO has We identifi ed no important diff erences in the occur- received honoraria, consultancies, and educational grants from rence of adverse events between the two groups. Serious Centocor, Schering Plough, Essex-Germany, UCB, and Abbott, and infection was not more frequent in either group. payments for consultancies from Shire and Boston Scientifi c. PR has However, the number of patients was inadequate to received consulting fees, lecture fees, and grant support from Centocor and Schering Plough, and has served as an expert witness for those address safety diff erences between the strategies.
companies. DH has received consulting fees from Abbott, Centocor, Our study had two main limitations. First, invest- UCB, and Schering Plough; lecture fees from Centocor, AGA and igators and patients were aware of the treatment Schering Plough; and grants from Schering Plough, Abbott and UCB; and is a member of the Initiative on Crohn's and Colitis, Independent assignment, which could have biased their assessment Dutch Academic Non-profi t Organisation for IBD Research. MDV has cacy. However, combined immunosuppres- received consulting fees from Schering Plough; Altana lecture fees sion was also more eff ective for both mucosal healing from Schering Plough and UCB; and grant support from AstraZeneca, and serum C-reactive protein concentration, which Roche, Schering Plough, Novartis Fund for Scientifi c Research Flanders, and Special Research Fund University Ghent. SV has are objective measures of infl ammation. Second, received consulting fees from Shire; lecture fees from Ferring, UCB, although remission was more rapid for patients Abbott, Schering Plough, and Tillotts; and grant support from UCB. assigned to the early combined immunosuppression JVDW has received consulting fees from UCB Schering Plough, Elan, strategy than for those given conventional treat- and Abbott; lecture fees from Schering Plough and Tramedico; and grant support from Initiative on Crohn's and Colitis. AAVB is a ment, simultaneous initiation of antimetabolites and member of the Initiative on Crohn's and Colitis, to which Schering corticosteroids could potentially have produced similar Plough BV and other companies that provide anti-TNF monoclonals results. However, both azathioprine and methotrexate (Abbott BV and UCB Pharma) provide a yearly unrestricted grant. have a slower onset of action than infl iximab.7,11,23 SVD has received consulting fees from Centocor, Elan, Schering Plough, and ISIS and lecture fees from Elan. BGF has received Furthermore, the conventional manage ment regimen research funding from Synta, Millennium, Schering Canada, Celltech, refl ected current clinical practice in that combined Centocor, Elan/Biogen, Berlex, Ortho-Biotech, Protein Design Labs, antimetabolites and corticosteroids are not commonly ISIS, Santarus, Schering Plough, Celgene, UCB Pharma, Napo used as ini tial treatments, and are not recommended Pharma, BMS, Abbott, and Otsuka; and consulting and lecture fees from UCB Pharma, Schering Canada, Proctor and Gamble, by experts.25–28,33,34 Elan/Biogen, Millennium, Protein Design Labs, Berlex, AstraZeneca, Celgene, Abbott, Santarus, GeneLogic, Cerimon Pharmaceuticals, GD'H had the original idea for this trial, designed the protocol with Tioga Pharmaceuticals, BMS, ISIS, Serono, Teva, Genentech, and SVD, served as lead investigator for Belgium, and drafted the manuscript with BGF. SVD designed the protocol with GD'H, served as a lead investigator for the Netherlands, and participated in the We wish to thank Beverley Jasevicius for expert secretarial support.
writing process. LS and AD analysed the data. SV analysed the results of the endoscopic substudy. PR assisted in the design of the trial and the writing process. BGF was a member of the study safety committee, Summers RW, Switz DM, Sessions JT Jr, et al. National coordinated the analysis of the results at Robarts Clinical Trials, and cooperative Crohn's disease study: results of drug treatment.
Gastroenterology 1979; 77: 847–69.
drafted the manuscript. DH was a lead investigator for the Netherlands, and participated in the writing of the manuscript. Malchow H, Ewe K, Brandes JW, et al. European Cooperative Crohn's Disease Study (ECCDS): results of drug treatment. FB, JVDW, GVA, PC, PV, HT, MDV, FVDM, JCC, TO, AAVB, PVH, Gastroenterology 1984; 86: 249–66.
GL, and FM were investigators. All authors participated in the data Munkholm P, Langholz E, Davidsen M, Binder V. Frequency of analysis and reporting stage of this manuscript, and have seen and glucocorticoid resistance and dependency in Crohn's disease. approved the fi nal version.
Gut 1994; 35: 360–62.
Faubion WA Jr, Loftus EV Jr, Harmsen WS, Zinsmeister AR, H Hart Ziekenhuis Roeselare, Belgium (FB); Ziekenhuis Oost- Sandborn WJ. The natural history of corticosteroid therapy for Limburg, Genk, Belgium (PC); Clinique St Pierre, Ottignies, Belgium infl ammatory bowel disease: a population-based study. (JCC); University Hospital Gent, Belgium (MDV); Imelda General Gastroenterology 2001; 121: 255–60.
Hospital, Bonheiden, Belgium (GD'H); Clinique St Joseph, Liège, Lichtenstein GR, Feagan BG, Cohen RD, et al. Serious infections Belgium (F Fontaine); Academic Medical Center, Amsterdam, and mortality in association with the therapies for Crohn's disease: TREAT Registry. Netherlands (DH); AZ Damiaan Campus St Jozef, Oostende, Belgium Clin Gastroenterol Hepatol 2006; (GL); CHU Sart Tilman, Liège, Belgium (E Louis); AZ VUB, Jette, Singleton JW, Law DH, Kelley ML Jr, Mekhjian HS, Belgium (FM); Ludwig-Maximilians Universität, München, Germany Sturdevant RAL. National cooperative Crohn's Disease Study: (TO); AZ St Lucas, Brugge, Belgium (PVH); University Hospital adverse reactions to study drugs. Gastroenterology 1979; Gasthuisberg, Leuven, Belgium (PR); Algemeen Ziekenhuis Jan Palfi jn, Merksem, Belgium (M Simoens); Medisch Centrum Alkmaar, Candy S, Wright J, Gerber M, Adams G, Gerig M, Goodman R. A Netherlands (HT); VU University Medical Centre, Amsterdam, controlled double blind study of azathioprine in the management Netherlands (AAVB); Algemeen Ziekenhuis St Augustinus, Wilrijk, of Crohn's disease. Gut 1995; 37: 674–78.
Belgium (FVDM); Erasmus University Medical Centre, Rotterdam, Markowitz J, Grancher K, Kohn N, Lesser M, Daum F, The Netherlands (JVDW); CAZK Groeninghe Campus St Niklaas, Kortrijk, Pediatric 6MP Collaborative Group. A multicenter trial of 6-mercaptopurine and prednisone in children with newly
diagnosed Crohn's disease. Gastroenterology 2000; 119: 895–902.
Confl ict of interest statement
PC, PV, HT, LS, AD, FVDM, JCC, PVH, GL, and FM reported no
Feagan BG, Rochon J, Fedorak RN, et al. Methotrexate for the treatment of Crohn's disease. The North American Crohn's Study confl icts of interest. GD'H, FB, and GVA have served as consultants Group Investigators. N Engl J Med 1995; 332: 292–97.
and speakers for Centocor and Schering Plough. SVD has acted as a www.thelancet.com Vol 371 February 23, 2008
10 Feagan BG, Fedorak RN, Irvine EJ, et al. A comparison of 22 Daperno M, D'Haens G, van Assche G, et al. Development and methotrexate with placebo for the maintenance of remission in validation of a new, simplifi ed endoscopic activity score for Crohn's disease. North American Crohn's Study Group Crohn's disease: the SES-CD. Gastrointest Endosc 2004; 60: 505–12.
Investigators. N Engl J Med 2000; 342: 1627–32.
23 Sandborn WJ, Tremaine WJ, Wolf DC, et al. Lack of eff ect of 11 Feagan BG, McDonald JWD. Methotrexate for Crohn's disease. intravenous administration on time to respond to azathioprine for N Engl J Med 1995; 333: 600–01.
steroid-treated Crohn's disease. North American Azathioprine 12 Present DH, Korelitz BI, Wisch N, Glass JL, Sachar DB, Study Group. Gastroenterology 1999; 117: 527–35.
Pasternack BS. Treatment of Crohn's disease with 24 Pocock SJ. Monitoring trial progress. In Pocock SJ, ed. Clinical 6-mercaptopurine. a long-term, randomized, double-blind study. trials: a practical approach. Chichester, UK: John Wiley & Sons, N Engl J Med 1980; 302: 981–87.
1983: 142–59.
13 Targan SR, Hanauer SB, van Deventer SJ, et al. A short-term 25 Carter MJ, Lobo AJ, Travis SP; IBD Section, British Society of study of chimeric monoclonal antibody cA2 to tumor necrosis Gastroenterology. Guidelines for the management of factor for Crohn's disease. N Engl J Med 1997; 337: 1029–35.
infl ammatory bowel disease in adults. Gut 2004; 14 Rutgeerts P, D'Haens G, Targan S, et al. Effi cacy and safety of 53 (suppl 5): 1–16.
retreatment with anti-tumor necrosis factor antibody (infl iximab) 26 Travis SP, Stange EF, Lemann M, et al. European evidence based to maintain remission in Crohn's disease. Gastroenterology 1999; consensus on the diagnosis and management of Crohn's disease. Gut 2006; 55: 16–35.
15 Hanauer SB, Feagan BG, Lichtenstein GR, et al. Maintenance 27 Panaccione R, Fedorak RN, Aumais G, et al. Canadian Association infl iximab for Crohn‘s disease: The ACCENT 1 Randomised Trial. of Gastroenterology clinical practice guidelines: the use of Lancet 2002; 359: 1541–49.
infl iximab in Crohn's disease. Can J Gastroenterol 2004; 16 Goekoop-Ruiterman YP, De Vries-Bouwstra JK, Allaart CF, et al. Clinical and radiographic outcomes of four diff erent treatment 28 Schreiber S, Campieri M, Colombel J-F, et al. Use of anti-tumour strategies in patients with early rheumatoid arthritis (the BeSt necrosis factor agents in infl ammatory bowel disease. European study): a randomized, controlled trial. Arthritis Rheum 2005; guidelines for 2001–2003. Int J Colorectal Dis 2001; 16: 1–11.
29 Rutgeerts P, Diamond RH, Bala M, et al. Scheduled maintenance 17 Breedveld FC, Weisman MH, Kavanaugh AF, et al. The PREMIER treatment with infl iximab is superior to episodic treatment for the study: a multicenter, randomized, double-blind clinical trial of healing of mucosal ulceration associated with Crohn's disease. combination therapy with adalimumab plus methotrexate versus Gastrointest Endosc 2006; 63: 433–42.
methotrexate alone or adalimumab alone in patients with early, 30 Sandborn WJ, Feagan BG, Hanauer SB, et al. A review of activity agressive rheumatoid arthritis who had not had previous cacy endpoints for clinical trials of medical therapy methotrexate treatment. Arthritis Rheum 2006; 54: 26–37.
in adults with Crohn's disease. Gastroenterology 2002: 122: 512–30.
18 Smolen JS, van der Heijde DM, St Clair EW, et al. Predictors of 31 Modigliani R, Mary J-Y, Simon JF, et al. Clinical, biological, and joint damage in patients with early rheumatoid arthritis treated endoscopic picture of attacks of Crohn's disease: evolution on with high-dose methotrexate with or without concomitant prednisolone. Gastroenterology 1990; 98: 811–18.
infl iximab: results from the ASPIRE trial. Arthritis Rheum 2006; 32 Cosnes J, Nion-Larmurier I, Beaugerie L, Afchain P, Tiret E, Gendre J-P. Impact of the increasing use of immunosuppressants 19 Lémann M, Mary J-Y, Duclos B, et al. Infl iximab plus in Crohn's disease on the need for intestinal surgery. Gut 2005; azathioprine for steroid-dependent Crohn's disease patients: a randomized placebo-controlled trial. Gastroenterology 2006; 33 Hommes DW, Oldenburg B, van Bodegraven AA, et al. Guidelines for treatment with infl iximab for Crohn's disease. Neth J Med 20 Best WR, Becktel JM, Singleton J, Kern F Jr. Development of a 2006; 64: 219–29.
Crohn's disease activity index. Gastroenterology 1976; 70: 439–44.
34 Sandborn WJ, Hanauer SB. Infl iximab in the treatment of 21 Irvine EJ, Feagan B, Rochon J, et al. Quality of life: a valid and Crohn's disease: a user's guide for clinicians. Am J Gastroenterol reliable measure of therapeutic effi cacy in the treatment of 2002; 97: 2962–72.
infl ammatory bowel disease. Gastroenterology 1994; 106: 287–96.
www.thelancet.com Vol 371 February 23, 2008

Source: http://www.birdgroup.be/upload/files/papers/step+up+top+down.pdf