Treatment of Crohn’s Disease: The “Long” of It
Article Outline
In this issue of Gastroenterology Lemann et al report on the long-term remissions of Crohn’s disease maintained by azathioprine in a randomized, double-blind, controlled withdrawal trial performed by the Groupe d’Etude Therapeutique des Affections Inflammatoires du tube Digestive (GETAID).1 The investigators premised their trial on a previous, retrospective series of patients, reported by the same group, in which voluntary withdrawal from azathioprine after 4 years of successful remission appeared to provide equal protection from relapse as maintenance of azathioprine therapy.2 In the current trial, a similar cohort of patients, those in clinical remission Crohn’s Disease Activity Index (CDAI) <150 receiving continuous treatment with azathioprine for at least 42 months without “flare-ups” (on <10 mg prednisone daily) and no other “maintenance” medications within 6 months of entry, were randomized in a double-blind process to continue azathioprine at the same dose (mean 1.7 mg/kg body weight) or placebo for an additional 18 months. The investigators used a “non-inferiority” design based upon the assumption that the control treatment (continuing azathioprine) is an effective therapy with “withdrawal of azathioprine” as a “non-inferior” comparator.
Azathioprine and its metabolite, 6-mercaptopurine, have been accepted as maintenance therapies after steroid-induced remissions in Crohn’s disease3 although recent controversy exists regarding optimal dosing and potential benefits of therapeutic drug monitoring of metabolites.4, 5 Similar issues have arisen regarding the role for these agents in the prevention or delay of relapse after surgical resections.6, 7, 8 It is of interest that, in the current trial, the mean baseline dose of azathioprine (1.7 mg/kg per day) that was maintaining remissions in this French cohort of Crohn’s disease patients was substantially lower than the dose used in clinical trials (2.5 mg/kg per day). Another controversy, related to practice differences between North American and European practice, is that patients were allowed to take less than 10 mg/day of prednisone whereas complete steroid-withdrawal is a more rigid remission criterion.
The investigators report the 18-month relapse rate of 8% for the patients maintained on azathioprine compared with 21% for patients randomized to placebo. Additional follow-up data presented at Digestive Disease Week 2004 demonstrated additional relapses in up to 60% of patients at 54 months after discontinuing azathioprine.9 The only toxicity associated with continuation of azathioprine was worrisome; fatal myelodysplasia with bone marrow abnormalities in chromosome 7 in a patient who had received azathioprine for 68 months. The absence of other toxicities, as described by the authors, is likely due to the selection of patients already tolerant of azathioprine for 3.5 years. In studies initiating azathioprine or 6-mercaptopurine approximately 20% of patients are unable to tolerate the drug due to nonspecific symptoms, hypersensitivity, (including pancreatitis) or bone marrow suppression.
The ability to select which patients are most likely to benefit from long-term therapy with azathioprine or 6-mercaptopurine is compromised by use of the clinical, but not biological, end-point of the CDAI, as many patients “in remission” at entry had either biologic (elevated C-reactive protein, hemoglobin <12 g/dL) or endoscopic evidence of persisting inflammatory disease. While it is not surprising that biologic evidence of inflammation was a predictor of clinical (CDAI) relapse,10 neither the presence of residual endoscopic lesions nor continuation of cigarette smoking were associated with a worse clinical prognosis. The lack of association between residual endoscopic lesions and long-term prognosis is consistent with the previous observations of GETAID regarding the impact of corticosteroids on mucosal healing,11 but contrasts with the group’s observations regarding the harmful impact of smoking on the clinical activity of Crohn’s disease.12 The other significant confounding variable in the trial was use (or time off) corticosteroids such that continuation of low-dose steroids was protective of a clinical relapse according to the CDAI.
While readers of Gastroenterology are quite comfortable with the standard comparison of treatments in a randomized clinical trial investigating potential superiority of 1 of 2 treatments, the analysis of a non-inferiority trial may be less familiar. The non-inferiority design is most commonly used to determine whether the efficacy of a generic pharmaceutical is the same as its proprietary counterpart. Proving that 2 treatments have the same average effect is virtually impossible, since one cannot distinguish minor (but real) differences from sampling variability even with enormous sample sizes. Instead, one argues that, if the 2 treatments had substantially equivalent effectiveness, then an adequately powered head-to-head comparison would be able to rule out a major difference in effect. The investigators for the GETAID study suggest that ruling out a 20% increase in 18-month relapse rate would be enough to conclude that azathioprine withdrawal is not (substantially) worse than continued azathioprine treatment. Conversely, if an adequately powered trial cannot rule out that large a difference, it must be because withdrawal is less effective than continuation. It is the second result that the authors report.
In a standard superiority trial, the interpretation of a statistically significant difference is unaffected by the study’s statistical power, while non-significant findings may occur because of inadequate power. A similar statement is true for non-inferiority studies: interpreting findings of non-inferiority does not involve power, but interpreting results that are consistent with inferiority do. In the GETAID study, we are in the latter situation, so the conclusion that azathioprine withdrawal is inferior to continuation depends upon the study having adequate power. The power calculation spelled out in the paper show that the sample size was adequate to detect a 20 percentage point difference with 80% power, assuming a baseline recurrence rate of 10% in the azathioprine continuation group, a rate very similar to the 8% actually observed.
The standard 2-sided 95% confidence interval for the difference ranges from 28% to −1% greater placebo recurrence rates. In a standard (superiority) trial, this would almost rule out equivalence. (If one accepts the authors’ arguments for testing differences in only one direction, the difference is statistically significant [P = 0.04].) Indeed, the difference in recurrence rates is striking, about 13%. Thus, only 8 patients need be continued on therapy to avoid 1 additional recurrence.
Should the authors have conducted the study (and the analysis) as a standard superiority study instead of using the non-inferiority approach? From the ethical standpoint going into the study, there was reason to believe that if withdrawal was worse than continuation, it would not be by much. In effect, the authors argue that if a 20% difference could be ruled out, it would then make sense for us to act as if withdrawal were not inferior to continuation. Powering a study to be able to rule out even smaller differences would not subject patients to much greater risk, but it would increase both inconvenience to more study patients and expense for the study sponsors. By contrast, if the investigators believed not only that withdrawal would be substantially (at least 20 percentage points) worse than continuation, but also that would be important to demonstrate this difference conclusively, then a conventional superiority trial with the sample sizes used here would be appropriate. But the ethical foundations of such a study would be shaky, because the premise underpinning the study would have been based on introducing a harmful regimen. The evidence going in to GETAID from retrospective studies clearly supported the non-inferiority design, even as the results of the study emphasize the continuing need for and benefits from prospective randomized trials.
These data, along with expanding data for the maintenance benefits of other immune modifiers (ie, methotrexate, infliximab) emphasize that, despite inductive therapies and prolonged immune modulation, the initiating and amplifying events in the etiopathogenesis of Crohn’s disease are persisting and require continuous down-regulation. As with all other agents, there will be some “cost” as pertains to potential adverse events, including rare cases of infections and neoplasia that are likely related to the level of immune suppression and unique attributes of the target cells or molecules (eg, tumor necrosis factor). When the overall risks and benefits of prolonged maintenance therapy with azathioprine are balanced, it is likely that most clinicians and patients will accept the small, as yet unquantified, risk of a lymphoid malignancy (or myelodysplasia) and the small risk of opportunistic infections to prevent the ongoing morbidity and impact on quality of life related to the chronic symptomatic activity of Crohn’s disease. In addition to clinical activity, future long-term trials will eventually include endoscopic, biologic, pharmacoeconomic end-points, and evidence for modification of the inflammatory sequela of Crohn’s disease.
References
- . A randomized double-blind controlled withdrawal trial in Crohn’s disease patients in long-term remission on azathioprine . Gastroenterology . 2005;128:1812–1818
- . Long-term follow-up of patients with Crohn’s disease treated with azathioprine or 6-mercaptopurine [see comments] . Lancet . 1996;347:215–219
- . Azathioprine for maintaining remission of Crohn’s disease . Cochrane Database Syst Rev . 2000;CD000067
- . Review article (monitoring of immunomodulators in inflammatory bowel disease) . Aliment Pharmacol Ther . 2005;21:307–319
- . Use of laboratory testing to guide 6-mercaptopurine/azathioprine therapy . Gastroenterology . 2004;127:1558–1564
- . Postoperative maintenance of Crohn’s disease remission with 6-mercaptopurine, mesalamine, or placebo (a 2-year trial) . Gastroenterology . 2004;127:723–729
- . Azathioprine and mesalamine for prevention of relapse after conservative surgery for Crohn’s disease . Gastroenterology . 2004;127:730–740
- . The efficacy of azathioprine and 6-mercaptopurine for the prevention of postoperative recurrence in patients with Crohn’s disease remains uncertain . Gastroenterology . 2004;127:990–993
- . Azathioprine withdrawal in patients with Crohn’s disease maintained on prolonged remission under treatment is associated with a high risk of relapse (abstract) . Gastroenterology . 2004;126(Suppl 2):A-113
- . C-reactive protein as a marker for inflammatory bowel disease . Inflamm Bowel Dis . 2004;10:661–665
- . Clinical, biological, and endoscopic picture of attacks of Crohn’s disease. Evolution on prednisolone. Groupe d’Etude Therapeutique des Affections Inflammatoires Digestives . Gastroenterology . 1990;98:811–818
- . Smoking cessation and the course of Crohn’s disease (an intervention study) . Gastroenterology . 2001;120:1093–1099
PII: S0016-5085(05)00702-X
doi:10.1053/j.gastro.2005.04.018
© 2005 American Gastroenterological Association. Published by Elsevier Inc. All rights reserved.
Refers to article:
- A Randomized, Double-Blind, Controlled Withdrawal Trial in Crohn’s Disease Patients in Long-term Remission on Azathioprine
