Challenges to the Design, Execution, and Analysis of Randomized Controlled Trials for Inflammatory Bowel Disease

Transcription

1 GASTROENTEROLOGY 2012;143: Challenges to the Design, Execution, and Analysis of Randomized Controlled Trials for Inflammatory Bowel Disease GEERT D HAENS,* BRIAN FEAGAN, JEAN FRÉDÉRIC COLOMBEL, WILLIAM J. SBORN, WALTER REINISCH, PAUL RUTGEERTS, # FRANK CARBONNEL,** JEAN YVES MARY, SILVIO DANESE, RICHARD N. FEDORAK, STEVEN HANAUER, and MARC LÉMANN ##, on Behalf of the International Organization for Inflammatory Bowel Diseases (IOIBD) and the Clinical Trial Committee Clincom of the European Crohn s and Colitis Organisation (ECCO) *Academic Medical Center, Amsterdam, The Netherlands, and Robarts Europe; University of Western Ontario and Robarts Clinical Trials, London, Ontario, Canada; CHRU Lille, Lille, France; University of California, San Diego, San Diego, California, and Robarts West; University of Vienna, Vienna, Austria; # Catholic University of Leuven, Leuven, Belgium; **Hôpitaux Universitaires Paris-Sud, Paris, France; INSERM Unité 717, University Paris Diderot, France; IRCCS Istituto Clinico Humanitas, Milan, Italy; University of Alberta, Edmonton, Alberta, Canada; University of Chicago, Chicago, Illinois; and ## Hôpital Saint Louis, Paris, France Treatment of inflammatory bowel disease has greatly improved with the development of targeted, monoclonal antibody based therapies. Tumor necrosis factor antagonists are frequently used to treat patients with Crohn s disease or ulcerative colitis, but they have side effects and their efficacy often decreases with use. New, more effective drugs are therefore needed and in development. However, many agents that appeared to be promising in preclinical studies have failed to show efficacy in clinical trials. We discuss possible reasons for the failures of these reagents in trials, which include the high rate of response to placebo, an inadequate range of doses, inappropriate timing of end point measurements, the changing therapeutic environment, and the competitive trial system. We also review regulatory guidelines for end points and trial design and recommend ways to improve trials. Keywords: Inflammatory Bowel Disease; Crohn s Disease; Ulcerative Colitis; Trial Design; Clinical Trials. Designing clinical trials for complex disorders such as inflammatory bowel disease (IBD) is a challenge because symptoms do not always parallel signs of inflammation, different forms of the disease have different responses to treatment, and therapeutic approaches vary by location. Nonetheless, there is a continuous and intense search for new treatments for Crohn s disease (CD) and ulcerative colitis (UC) (Table 1). Disappointingly, many agents fail in early phases of clinical trials, often because of poor study design, lack of objective measurements, or insufficient sample sizes. Furthermore, regulators propose different approaches in different areas in the world. We attempt to offer guidance for this complicated area of clinical research. The strategies we propose are based on information from lectures presented at meetings of the International Organization for the Study of Inflammatory Bowel Diseases (IOIBD) and the European Crohn s and Colitis Organisation (ECCO) in 2011, our own personal experience in designing investigator-initiated clinical studies of IBD, reports from industrial clinical trials, and our interactions with the US Food and Drug Administration and European Medicines Agency. We also received suggestions from authors who participated in the Clinical Trial Task Force of the IOIBD and the Clinical Committee of the ECCO. Evolution of IBD Trials Until recently, most clinical trials for IBD investigated the ability of a reagent to induce or to maintain a response or remission. Usually, agents found to be effective at inducing a response or remission were then evaluated for maintenance. Patients were given open-label induction treatment with the new agent, and only responders were included in the maintenance phase of the study. A few trials have continued to blind patients to the test article throughout the induction and maintenance phases. However, the distinction between the induction and maintenance phases could gradually become less important as trial end points evolve beyond clinical assessments. Rather, investigators will begin to evaluate long-term benefits such as prevention of structural complications and surgery outcomes that usually cannot be evaluated in short-term trials. 1 Induction Trials Making Comparisons The short-term efficacy of a drug can be compared with either placebo or an active agent. However, it is often unethical to give untreated patients a placebo. 2 The European Medicines Agency has stated that the test reagent should be compared with and be at least as effective and safe as the standard of care, which currently in the majority of cases includes corticosteroids, but this comparison is rarely put into practice. The test reagent should be Deceased. Abbreviations used in this paper: CDEIS, Crohn s Disease Endoscopic Index of Severity; CRP, C-reactive protein; ECCO, European Crohn s and Colitis Organisation; IOIBD, International Organization for the Study of Inflammatory Bowel Diseases; SES-CD, Simple Endoscopic Score for Crohn s Disease; TNF, tumor necrosis factor by the AGA Institute /$

2 1462 D HAENS ET AL GASTROENTEROLOGY Vol. 143, No. 6 Table 1. Molecules With Promising Efficacy Data From Early-Stage Studies (Incomplete List) Name Target Manufacturer Phase Rituximab CD20 Roche 1 2 Visilizumab CD3 PDL BioPharma 1 2 Daclizumab CD25 PDL BioPharma/Roche 1 2 Basiliximab CD25 Novartis 1 2 Fontolizumab IFN- PDL BioPharma 1 2 Etanercept TNF receptor Wyeth 1 2 Onercept TNF receptor Merck Serono 1 2 Abatacept CD28 Bristol-Myers Squibb 3 Briakinumab Interleukin-12/interleukin-23 Abbott Laboratories 3 Sargramostim Granulocyte-macrophage colony stimulating factor Berlex 3 Alicaforsen Intercellular adhesion molecule 1 Isis Pharmaceuticals 3 Somatotropin Growth hormone Elli Lilly 3 safe and effective. However, corticosteroids are only effective for a short time, have many side effects, increase mortality, and offer no durable benefit. 3 Thus, it might not be appropriate for test reagents to be compared with corticosteroids. In contrast, aminosalicylates are safe and effective short-term and long-term therapeutics for patients with UC. It would be reasonable for trials of firstline treatments for mild to moderate UC to compare the test reagent with aminosalicylates. 4 However, these drugs have not been sufficiently effective in trials for CD to be used as the standard of care. 5 In induction trials for CD, placebo is a better standard than aminosalicylate or corticosteroid therapy. In trials of second-line agents (for patients who did not respond to their initial therapy), it is acceptable and even recommended that the test article be compared with placebo. However, combining the investigational agent with treatments such as immunosuppressive agents could increase the risk of adverse events. Thus, in specific situations, it might be appropriate for study participants to discontinue their previous medications. Most randomized controlled trials designed to assess the superiority of an agent over placebo have been powered to detect a minimum, clinically important difference in rate of response or remission (15% 20%), which requires each study group to include approximately 100 to 150 patients. 6 The goal of noninferiority trials is to show that the investigational agent is no worse than the standard therapy. Any observed difference in clinical response must be insignificant (based on a one-sided analysis, with a 95% confidence interval) if the test reagent is to be considered noninferior. Regulatory authorities have proposed that a noninferiority margin as low as 4% should be used to compare 2 different 5-aminosalicylic acid drugs for maintenance of remission in patients with UC. A noninferiority trial with a margin of 7.5% would require randomly assigning more than 1200 patients to 2 groups. In our opinion, a trial of this size is hardly feasible. 7 Recent noninferiority trials of patients with UC have taken more than 2 years to recruit the required number of participants. 8,9 Selection of Study Participants For inclusion in induction trials, patients should have a definitive diagnosis of IBD based on accepted criteria, 10 with disease duration of at least 3 months, objective evidence of active inflammation, and the possibility for improvement from treatment with the test product. Patients with pure fibrostenotic CD without associated inflammation are therefore not good candidates for trials of anti-inflammatory therapy. The symptom scores that have been developed and validated for CD are not reliably associated with the presence of active inflammatory lesions, 11 so it has become common practice to document the presence of active inflammation using markers in serum (C-reactive protein [CRP]) or stool samples (calprotectin) or by direct endoscopic assessment. In some current trials, video-recorded ileocolonoscopies are reviewed by experts to ascertain the presence of active ulceration, a process that has been facilitated by improved technology. Magnetic resonance enterography, a noninvasive, highly accurate diagnostic imaging technology, could eventually replace endoscopy. 12 We believe that objective assessment of disease activity improves trial efficiency by reducing rates of response to placebo. 13 Medications that patients take along with the test agent can affect trial results, so potential drug interactions must be considered. Participants in a trial of the Janus kinase inhibitor tofacitinib had to stop taking thiopurines, because each drug can cause leukopenia. 14 Similarly, it has been proposed that combination therapy with immunosuppressive and anti-integrin agents can increase the risk of progressive multifocal leukoencephalopathy. 15 Although it is necessary to restrict concomitant medications in some situations, this limits our ability to apply findings to the overall population. A recent analysis showed that most patients do not meet the inclusion and exclusion criteria for phase 3 trials of agents for IBD. 16 Some trials have restrictive inclusion criteria based on parameters other than disease activity. For example, studies of changes in progression of CD might require patients with a relatively short history of disease (early-stage CD). 1 Trials of topically delivered agents require patients with disease localized to a specific area. 17,18 Trials investi-

3 December 2012 CHALLENGES OF RCTs FOR IBD 1463 gating fistula healing or prevention of postoperative recurrence require subjects with draining fistulas or recent ileocolonic resection, respectively. 19,20 Symptoms of UC, compared with CD, are more frequently associated with mucosal inflammation. Subjects in trials for UC should receive a baseline endoscopy examination and be reevaluated within 6 to 10 weeks, because mucosal healing (restitution of epithelial integrity) improves prognosis. 21 In general, sigmoidoscopy has been considered sufficient for assessment of mucosal healing, although this procedure might not yield complete information about the proximal extent and inflammatory burden of UC. In terms of severity, participants in most trials of agents for active UC should have, at a minimum, documented friability on endoscopy. 22 Some studies have specific entry criteria. The Study of Biologic and Immunomodulator Naive Patients in Crohn s Disease (SONIC) enrolled patients with active CD who had not received any immunosuppressive or biologic therapies, had not responded to corticosteroid therapy (or been considered for a second course of corticosteroids within 12 months), or had not responded to budesonide or aminosalicylates. 23 The study was designed to identify patients who had not responded to first-line therapies but did respond to immunomodulators. The UC SUCCESS trial (Efficacy and Safety of Infliximab, as Monotherapy or in Combination with Azathioprine, versus Azathioprine Monotherapy in Moderate to Severe Ulcerative Colitis) enrolled patients who had not responded to treatment with aminosalicylates but not received tumor necrosis factor (TNF) antagonists. 24 In SONIC and UC SUCCESS, patients were randomly assigned to groups treated with infliximab, azathioprine, or a combination of both. Trials to test combinations of reagents should be performed after the efficacy of each individual agent has been established. As management strategies become more complex, there will be a greater need to evaluate treatment algorithms. Cluster randomization trials, in which the unit of randomization is a practice or a hospital, are well suited for these types of analyses. 25 Selection and Timing of End Points Most randomized controlled trials of agents for IBD have a primary, several secondary, and then exploratory end points. Historically, symptom-based end points have been used to assess efficacy, particularly for CD. However, times are changing. For example, in the EXTEND (Extend the Safety and Efficacy of Adalimumab through Endoscopic Healing) trial of adalimumab for CD, 26 results of endoscopic analysis were used as the primary end point. The Crohn s Disease Activity Index (CDAI) is used to measure clinical disease activity in almost all trials for CD, although it is does not always identify patients with inflammation. 27 Clinically meaningful improvement has been defined as a reduction in the CDAI by 70 to 100 points, whereas remission is defined as a CDAI less than 150. Thia et al analyzed CDAI data in placebo-controlled induction studies and showed that differences in treatment effects were maximized if the primary end point was defined as a reduction in CDAI of 100 points or a reduction by 70 points at 2 consecutive visits. 28 Several recent trials have used the single-day Harvey Bradshaw Index 29 to measure clinical response; results from the index correlate with the CDAI. 30 However, it is not clear if the Harvey Bradshaw Index is acceptable to regulatory authorities. For induction trials that use the CDAI as a primary end point, assessments are best made at study weeks 4 to 6 if the CDAI is only determined once or at study weeks 8 or 12 if the index is determined twice. Both of these strategies have been successful in limiting the numbers of patients that appear to respond to placebo. In maintenance trials, tapering and withdrawal of corticosteroids can be used as an end point and should be described as corticosteroid-free remission. It is important that corticosteroid tapering regimens are standardized in designing maintenance trials. Although corticosteroid-free remission is an important end point for patients and clinicians, it has not been used or required as a primary end point in registration trials. Reagents are usually tested at predetermined doses. However, the protocol can allow dose adjustments if a dose-response effect is observed without classifying patients with dose adjustments as nonresponders. Better understanding of the relationship between serum drug concentration and clinical efficacy will allow more extensive studies for optimizing drug efficacy. 31 Results from endoscopy analysis are not usually used as secondary end points in induction trials for CD. However, in the EX- TEND trial, mucosal healing was observed by study week Other induction trials have included endoscopic analyses of subsets of patients. 32 It might not be necessary for all patients in a trial to undergo endoscopy, as long as the characteristics of the endoscopy subgroup are comparable to those of the overall study population. Studies with endoscopic end points could considerably reduce sample size requirements. It is therefore important to validate endoscopic features as surrogate markers for clinical efficacy. Actuarial analyses are interesting to include in trials for IBD. Time to response or time to remission could be an important analytical approach to evaluate the rapidity of an effect, which could have clinical relevance. Methods to analyze these times are complicated and not well developed. The Turnbull method should be used instead of Kaplan Meier estimates. 33 Many more tools have been developed to evaluate UC than CD, but few have been validated. 22 The most frequently used tools are indices that combine clinical and endoscopic parameters. In trials of various biologic agents for UC, the Mayo Clinic score has been used consistently. It is a 12-point system that produces 4 subscores (based on stool frequency, rectal bleeding, physician global assessment, and endoscopic findings), each with a range of 0 to 3 points. 34 Response is defined as a decrease from

4 1464 D HAENS ET AL GASTROENTEROLOGY Vol. 143, No. 6 baseline in the total Mayo score of 3 points or 30% and a decrease in the rectal bleeding subscore 1 point or an absolute rectal bleeding subscore of 0 or 1 point; remission is defined as a total Mayo score 2 points with no subscore 1 point. The ideal time to assess the primary end point in UC induction trials is usually 6 to 8 weeks, although an earlier time point (for instance, 7 14 days) might be required for trials that evaluate hospitalized patients with severe UC. Studies of hospitalized patients with severe UC that is refractory to oral treatment (usually corticosteroids) often use clinical improvement or rate of colectomy, rather than remission, as an end point. Trials that have studied this population have often used the Lichtiger score, a derivative of the original modified Truelove and Witts severity classification scheme. 35,36 However, neither the score itself nor the proposed cutoffs have been validated. The score ranges from 0 to 21 and has only patient-reported items and one physician-reported parameter (abdominal tenderness). Recent studies have indicated that short-term clinical improvement based on the Lichtiger score does not parallel clinically meaningful outcomes, such as mucosal healing, clinical remission, corticosteroid withdrawal, and colectomy We therefore do not believe that the Lichtiger score is an ideal instrument of disease activity in patients with UC. Biomarkers are also used as secondary end points in clinical trials. The most widely used biomarker is serum level of CRP. Although the concentration of CRP is normal in up to 20% of patients with active CD and 50% of patients with active UC, 40 reductions in CRP usually correlate with clinical response in patients who have increased levels of CRP at baseline. A post hoc analysis of data from SONIC revealed a better separation between treatment groups of patients that had increased concentrations of CRP at baseline than among those with normal levels of CRP. 23 It is therefore important to determine the presence of active inflammation using objective measures. Other blood biomarkers of inflammation include platelet or leukocyte counts, serum levels of albumin, and erythrocyte sedimentation rate. Fecal biomarkers include calprotectin and lactoferrin. Levels of these markers correlate with clinical symptoms and the presence of mucosal lesions in patients with CD or UC; the magnitude of change usually corresponds with the degree of clinical improvement. 41 Fecal biomarkers are, however, nonspecific, because infections, use of nonsteroidal anti-inflammatory medications, and many other medical conditions can increase levels of calprotectin. 42 When used as a surrogate biomarker for endoscopic severity, levels of fecal calprotectin greater than 250 g/g could identify patients with CD who had large ulcers or significant amounts of inflammation (Mayo score of 2 3) in UC. 43 Tests of fecal biomarkers can be performed rapidly and repeatedly, even at the bedside (a point of care test). Mucosal Healing: A New End Point? Although mucosal healing must be evaluated to assess therapies for UC, it is a relatively new end point for trials of CD. The observations that immunomodulatory and biologic agents could heal the mucosa and that mucosal healing reduced rates of surgery and hospitalization have renewed interest in using endoscopy findings as outcomes. Ileocolonoscopy is the standard for assessing mucosal disease activity. However, studies are underway to determine whether less invasive alternatives, such as capsule endoscopy, computed tomography, magnetic resonance enterocolonography, or ultrasonography are useful alternatives and/or adjuncts to ileocolonoscopy for the objective assessment of disease activity. Rimola et al correlated results from ileocolonoscopy with those from magnetic resonance imaging enterography. 12 Indices that have been validated for ileocolonoscopic assessment of CD include the Crohn s Disease Endoscopic Index of Severity (CDEIS) and the Simple Endoscopic Score for Crohn s Disease (SES-CD). 44,45 The Rutgeerts score (ranging from 0 to 4 points) has been shown to predict recurrence of CD after surgery of the neoterminal ileum. 46 Interestingly, the largest trials that have used mucosal healing as a primary or major secondary end point have used the definition absence of ulcers rather than prespecified cutoff values on the CDEIS or SES-CD. The CDEIS has been validated and found to have appropriate sensitivity to measure changes in the mucosal appearance. On a scale of 0 to 44 points, endoscopic remission (minor or no lesions) is defined as a CDEIS score 6 or 7, and complete endoscopic remission (mucosal healing; ie, no lesions at all or scarred lesions only) is defined as a CDEIS score 3 or 4. An endoscopic response is a decrease from baseline CDEIS score of at least 4 or 5 points. 47 The CDEIS has been used in trials of corticosteroids, azathioprine, and TNF antagonists. The limitations of the CDEIS lie mainly in its complexity, which makes training necessary. No cutoff values have been determined for the SES-CD. Use of the absence of ulcers as an outcome has the advantage of being simple to determine (all or none) but also has important limitations. Specifically, this outcome has not been validated or assessed for reproducibility, and the definition of an ulcer is open to interpretation. Furthermore, it is not always associated with clinical effects, because the persistence of one small ulcer would be classified as an absence of mucosal healing. Interestingly, a recent post hoc analysis of data from SONIC associated a 50% reduction in severity, measured by endoscopy, with a favorable clinical outcome (complete mucosal healing or an absence of ulcers). 48 Further research is needed to determine what degree of improvement, measured by endoscopy, is clinically meaningful. Endoscopic evaluation is an established component of efficacy trials for UC. In 1962, Baron et al designed a scoring system, based on the severity of bleeding (not

5 December 2012 CHALLENGES OF RCTs FOR IBD 1465 ulceration), to assess disease severity using proctoscopes, which are rigid. 49 Although almost all patients are now examined with flexible instruments, the scoring system has progressed only slightly, and the instructions for scoring are unclear. For example, where should disease severity be measured (rectum or sigmoid, most or least inflamed segment)? Which type of bowel cleansing should be used? What extent of the colon should be assessed? Recent efforts have led to the development and validation of a new scoring system for the severity of UC, called the Ulcerative Colitis Endoscopic Index of Severity, 50 which requires further evaluation in clinical trials. It is unclear what definition of mucosal healing should be used in studies of UC. In recent trials of infliximab and adalimumab, mucosal healing was defined by a Mayo subscore of 0 or 1 or a decrease from baseline subscores of 2 or 3. In ACT (Active Ulcerative Colitis Trials), patients with a posttreatment Mayo score of grade 1 were no more likely to undergo a colectomy than those with a score of 0. Mucosal healing might be defined in future trials as an absence of erosions, ulcers, and spontaneous bleeding. There is no consistency in the components included in the scoring systems used for endoscopy analysis. The term minimal or slight friability used in the Mayo scoring system is subjective and leads to inconsistent results. We propose that the presence of any friability should be scored as 2. However, it is difficult to assess friability with centrally viewed recordings. Maintenance Trials Maintenance trials in IBD should last between 6 and 24 months. Although these studies typically evaluate maintenance or durability of response (a change in CDAI 70 or 100 at all time points), some trials have used the more rigorous outcome of maintenance of remission (CDAI 150 at all time points) as an end point or, alternatively, time to loss of response. Trials that take longer than 3 months should enforce withdrawal of corticosteroids, which enables assessment of clinically relevant end points such as complete corticosteroid withdrawal, corticosteroid-free response, and corticosteroidfree remission. Tapering of corticosteroids should be performed according to a defined schedule rather than at the investigator s discretion. Most patients enter maintenance trials following a response or remission to the same agent or to corticosteroids. Maintenance success rates are then presented with the number of patients who entered the trial as the denominator. However, it is important to remember that these patients were initially selected for their ability to respond to the test agent. This trial configuration therefore tends to overestimate the efficacy of therapy. For example, less than 30% of participants in maintenance trials of TNF antagonists maintained remission after 1 year of treatment. A few trials have blinded participants from the induction phase through the maintenance phase. 54 An end point of sustained or durable remission, meaning remission (CDAI 150 or Mayo score 2 with no subscore 1) at all visits of the maintenance trial, is important but difficult to achieve. For CD, the European Medicines Agency recommends the following: 1) that only patients with a CDAI 150 for at least 1 month be included in the trials, 2) that patients with surgically induced remission can be entered directly but should preferably be studied in separate trials, and 3) that trials combining induction and maintenance treatment should preferably only enter patients that have achieved remission with either the trial drug or the comparator. Here, re-randomization should be done. 55 We do not completely agree with these guidelines, which are empiric and not based on evidence. If only patients who enter remission following induction therapy can be included in the maintenance trial, then twice as many patients need to receive induction therapy to ensure a sufficient number of patients for the maintenance trial. This approach is not feasible. Regulators are preoccupied with remission and forget that a response is meaningful to patients. Many responders, given sufficient time, will enter remission. We propose that responders be randomly assigned to the different arms of the maintenance study. The most robust trial design combines the assessment of induction and maintenance, avoiding the problems associated with rerandomization. Nonetheless, randomly assigning responders, rather than remitters, to the study arms has risks. If responders, but not remitters, are future nonremitters, including these patients in the trial could bias the results if the patients are not distributed equally among the study groups. On the contrary, if most responders, but not remitters, turn out to be slow remitters, their inclusion in a maintenance trial is justified. Stratified randomization, based on the presence of remission in addition to response, is another approach. For patients with UC, maintenance of remission now includes maintenance of mucosal healing. This means that patients undergo endoscopic examinations when they enter maintenance trials and when they are suspected of having a relapse (to ascertain the presence of mucosal inflammation) and at the end of the follow-up period. Studies of IBD Progression IBDs are chronic diseases with serious complications. It is important to evaluate patients outcomes beyond the typical 1- or 2-year duration of the clinical trial. To understand the long-term effects of therapies, methodologically rigorous cohort studies are needed to assess the development of structural damage such as strictures, perforations, and need for surgery. Furthermore, new tools are required to assess the relationship between control of inflammation and progression of disease to struc-

6 1466 D HAENS ET AL GASTROENTEROLOGY Vol. 143, No. 6 Table 2. Factors Associated With High Rates of Response to Placebo in IBD Trials Longer time of evaluation Higher number of study visits Lower disease activity at inclusion Parenteral versus oral treatment Longer duration of disease tural complications. One important development is the creation of the Lémann score (IPNIC), which combines data from endoscopic and imaging studies to determine structural damage in patients with CD. 56 The Problem of the Placebo Response Responses to placebo can result from carryover effects from previous treatments, regression toward the mean, benefits of expert care during a clinical trial, and true biologic effects. High rates of response to placebo have been a common phenomenon in IBD trials, mainly for CD. Based on data from meta-analyses, pooled estimates of remission among patients given placebo range from 13% to 24% for the CDAI and from 2% to 12% for the Ulcerative Colitis Disease Activity Index. 57,58 Table 2 presents factors commonly associated with a high rate of response to placebo, including low disease activity at study entry, disease duration, high numbers of study visits, and the complexity of the therapeutic intervention (pills, infusions, apheresis, and so on). Regional variations in the magnitude of the response to placebo have been observed and could result from differences in knowledge, access to health care, and economic factors. One drug development program attempted to reduce the rate of response to placebo by requiring documentation of active disease as an inclusion criterion, documentation of mucosal inflammation by endoscopy (with central review), exclusion of patients with CD who had multiple surgeries, and daily scoring (or telephone registration) of disease activity symptoms to avoid recall bias. Changing Trial Environment Patients with IBD who enter clinical trials today are not comparable to those of a decade ago. Now, most participants have already been treated with immunosuppressive and/or biologic agents. Many trials are at least in part conducted at academic medical centers, where the referral population could be more refractory to treatment. These referral centers might find it difficult to enroll patients in early-phase trials or patients who have not been exposed to second- or third-line agents. The multitude of agents in development creates competition for patients. Site selection is usually performed by contract research organizations that frequently work with the same centers. These factors, in addition to the complex design of trials, the increasing numbers of patients required, and the administrative burdens of Good Clinical Practice, regulatory, and legal requirements, have all slowed trial progression. Efforts should focus on designing feasible trials that answer clinically relevant questions with a minimum of complicated procedures and burden to patients and investigators. Future Directions Many new agents are in late-stage trials and could be approved for treatment of IBD. This could lead to direct comparisons of new drugs or studies of drug combinations. Although these types of studies require extreme vigilance for safety issues, they could answer many tantalizing clinical questions. We hope that international groups will take on these projects and receive support, because pharmaceutical companies do not perform these types of analyses. The ideal management of IBD is likely to require a combination of agents, such as anti-inflammatory and antifibrotic agents. Trials of agents for IBD are complex. The large number of molecules in development, the high cost of drug development programs, and changing regulatory requirements have created challenges for investigators and clinical trial centers. Trials should be designed to include objective outcome parameters, rather than symptom scores, and efforts must be taken to reduce responses to placebo. These changes could lead to feasible sample sizes, which will speed accrual. Validation and standardization of scoring systems for endoscopic and imaging analyses to identify patients with clinically relevant responses and remission are also indispensable. Finally, IBD should be Table 3. Recommendations for IBD Clinical Trials In trials of first-line agents for UC, the test article should be compared with aminosalicylates. In trials for CD, it should be compared with placebo. In trials of second-line agents for IBD, the test article should be compared with placebo. The test agent or placebo should be added to the treatment a patient is already using if that fails to reach symptom/disease control. Patients included in trials of anti-inflammatory agents should have objective evidence of inflammation. Complex interventions (such as with combinations of drugs or different therapeutic strategies) should be studied in cluster randomization trials. The use of endoscopic end points could substantially reduce the sample size of trials; validated endoscopic scoring systems are therefore needed. Maintenance trials should last 6 to 24 months and incorporate a protocol to standardize corticosteroid withdrawal. Maintenance trials for CD should randomly assign patients who responded to induction treatment, rather than remitters, to study groups so that enough patients can be recruited. Maintenance trials for UC should include maintenance of mucosal healing as an end point.

7 December 2012 CHALLENGES OF RCTs FOR IBD 1467 approached as a long-term disease, with an ultimate goal of altering its progression and reducing disability and surgeries (Table 3). References 1. D Haens G, Fedorak R, Lémann M, et al. Endpoints for clinical trials evaluating disease modification and structural damage in adults with Crohn s disease. Inflamm Bowel Dis 2009;15: Guideline on the development of new medicinal products for the treatment of Crohn s disease. GB/document_library/Scientific_guideline/2009/09/WC pdf. 3. Sandborn WJ, Löfberg R, Feagan BG, et al. Budesonide for maintenance of remission in patients with Crohn s disease in medically induced remission: a predetermined pooled analysis of four randomized, double-blind, placebo-controlled trials. Am J Gastroenterol 2005;100: Travis SP, Stange EF, Lémann M, et al; European Crohn s and Colitis Organisation. European evidence-based Consensus on the management of ulcerative colitis: current management. J Crohns Colitis 2008;2: Ford AC, Kane SV, Khan KJ, et al. Efficacy of 5-aminosalicylates in Crohn s disease: systematic review and meta-analysis. Am J Gastroenterol 2011;106: Sandborn WJ, Feagan BG, Hanauer SB, et al. A review of activity indices and efficacy endpoints for clinical trials of medical therapy in adults with Crohn s disease. Gastroenterology 2002;122: Wiens BL. Choosing an equivalence limit for noninferiority or equivalence studies. Control Clin Trials 2002;23: D Haens G, Sandborn WJ, Barrett K, et al. Once-daily MMX mesalamine for endoscopic maintenance of remission of ulcerative colitis. Am J Gastroenterol 2012;107: Sandborn WJ, Korzenik J, Lashner B, et al. Once-daily dosing of delayed-release oral mesalamine (400-mg tablet) is as effective as twice-daily dosing for maintenance of remission of ulcerative colitis. Gastroenterology 2010;138: Van Assche G, Dignass A, Panes J, et al; European Crohn s and Colitis Organisation. The second European evidence-based Consensus on the diagnosis and management of Crohn s disease: definitions and diagnosis. J Crohns Colitis 2010;4: Mary JY, Modigliani R. Development and validation of an endoscopic index of the severity for Crohn s disease: a prospective multicentre study. Gut 1989;30: Rimola J, Rodriguez S, García-Bosch O, et al. Magnetic resonance for assessment of disease activity and severity in ileocolonic Crohn s disease. Gut 2009;58: Feagan BJ, Sandborn WJ, D Haens G, et al. The value of a central image management system (CIMS) in the conduct of randomized controlled trials of therapy for ulcerative colitis (UC) (abstr). Presented at: ACG 2012 Annual Scientific Meeting; October 19 24, 2012; Las Vegas, NV. 14. Sandborn WJ, Ghosh S, Panes J, et al. Phase 2 study of tofacitinib (CP-690,550), an oral Janus kinase inhibitor, in active Crohn s disease. Gastroenterology 2011;140(Suppl 1): Yousry TA, Major EO, Ryschkewitsch C, et al. Evaluation of patients treated with natalizumab for progressive multifocal leukoencephalopathy. N Engl J Med 2006;354: Ha C, Ullman TA, Siegel CA, et al. Patients enrolled in randomized controlled trials do not represent the inflammatory bowel disease patient population. Clin Gastroenterol Hepatol 2012;10: Greenberg GR, Feagan BG, Martin F, et al. Oral budesonide for active Crohn s disease. Canadian Inflammatory Bowel Disease Study Group. N Engl J Med 1994;331: Marshall JK, Thabane M, Steinhart AH, et al. Rectal 5-aminosalicylic acid for induction of remission in ulcerative colitis. Cochrane Database Syst Rev 2010;1:CD Sands BE, Anderson FH, Bernstein CN, et al. Infliximab maintenance therapy for fistulizing Crohn s disease. N Engl J Med 2004; 350: Hanauer SB, Korelitz BI, Rutgeerts P, et al. Postoperative maintenance of Crohn s disease remission with 6-mercaptopurine, mesalamine, or placebo: a 2-year trial. Gastroenterology 2004; 127: Frøslie KF, Jahnsen J, Moum BA, et al. Mucosal healing in inflammatory bowel disease: results from a Norwegian population-based cohort. Gastroenterology 2007;133: D Haens G, Sandborn WJ, Feagan BG, et al. A review of activity indices and efficacy end points for clinical trials of medical therapy in adults with ulcerative colitis. Gastroenterology 2007;132: Colombel JF, Sandborn WJ, Reinisch W, et al. Infliximab, azathioprine, or combination therapy for Crohn s disease. N Engl J Med 2010;362: Panaccione R, Ghosh S, Middleton S, et al. Infliximab, azathioprine, or infliximab azathioprine for treatment of moderate to severe ulcerative colitis: The UC SUCCESS trial. J Crohns Colitis 2012;6 (abstract). 25. Donner A, Birkett N, Buck C. Randomization by cluster. Sample size requirements and analysis. Am J Epidemiol 1981;114: Rutgeerts P, D Haens G, Van Assche G, et al. Adalimumab induces and maintains mucosal healing in patients with moderate to severe ileocolonic Crohn s disease-first results of the EXTEND trial. Gastroenterology 2012;142: Best WR, Becktel JM, Singleton JW, et al. Development of a Crohn s Disease Activity Index. Gastroenterology 1976;70: Thia KT, Sandborn WJ, Lewis JD, et al. Defining the optimal response criteria for the Crohn s disease activity index for induction studies in patients with mildly to moderately active Crohn s disease. Am J Gastroenterol 2008;103: Harvey RF, Bradshaw JM. A simple index of Crohn s-disease activity. Lancet 1980;1: Vermeire S, Schreiber S, Sandborn WJ, et al. Correlation between the Crohn s disease activity and Harvey-Bradshaw indices in assessing Crohn s disease severity. Clin Gastroenterol Hepatol 2010;8: Afif W, Loftus EV Jr, Faubion WA, et al. Clinical utility of measuring infliximab and human anti-chimeric antibody concentrations in patients with inflammatory bowel disease. Am J Gastroenterol 2010;105: Rutgeerts P, Diamond RH, Bala M, et al. Comparison of scheduled and episodic treatment strategies of infliximab in Crohn s disease. Gastroenterology 2004,126: Turnbull B. The empirical distribution function with arbitrarily grouped, censored and truncated data. J R Stat Soc B 1976;38: Schroeder KW, Tremaine WJ, Ilstrup DM. Coated oral aminosalicylic acid therapy for mildly to moderately active ulcerative colitis. N Engl J Med 1987;317: Lichtiger S, Present DH, Kornbluth A, et al. Cyclosporine in severe ulcerative colitis refractory to steroid therapy. N Engl J Med 1994; 330: D Haens G, Lemmens L, Geboes K, et al. Intravenous cyclosporine versus intravenous corticosteroids as single therapy for severe attacks of ulcerative colitis: a double-blind controlled randomised trial. Gastroenterology 2001;120: Laharie D, Bourreille A, Branche J, et al. Cyclosporin versus infliximab in severe acute ulcerative colitis refractory to intravenous steroids: a randomized trial. Lancet 2012 October 9 [Epub ahead of print].

8 1468 D HAENS ET AL GASTROENTEROLOGY Vol. 143, No Sandborn WJ, Colombel JF, Frankel M, et al. Anti-CD3 antibody visilizumab is not effective in patients with intravenous corticosteroid-refractory ulcerative colitis. Gut 2010;59: Baumgart DC, Targan SR, Dignass AU, et al. Prospective randomized open-label multicenter phase I/II dose escalation trial of visilizumab (HuM291) in severe steroid-refractory ulcerative colitis. Inflamm Bowel Dis 2010;16: Langhorst J, Elsenbruch S, Koelzer J, et al. Noninvasive markers in the assessment of intestinal inflammation in inflammatory bowel diseases: performance of fecal lactoferrin, calprotectin, and PMN-elastase, CRP, and clinical indices. Am J Gastroenterol 2008;103: Sipponen T, Savilahti E, Kolho KL, et al. Crohn s disease activity assessed by fecal calprotectin and lactoferrin: correlation with Crohn s disease activity index and endoscopic findings. Inflamm Bowel Dis 2008;14: van Rheenen PF, Van de Vijver E, Fidler V. Faecal calprotectin for screening of patients with suspected inflammatory bowel disease: diagnostic meta-analysis. BMJ 2010;341:c D Haens G, Ferrante M, Vermeire S, et al. Faecal calprotectin is a surrogate marker for endoscopic lesions in inflammatory bowel diseases. Inflamm Bowel Dis 2012 Feb 16 [Epub ahead of print]. 44. Mary JY, Modigliani R, GETAID. Development and validation of an endoscopic index of severity for Crohn s disease: a prospective multicenter study. Gut 1989;30: Daperno M, D Haens GR, Van Assche G, et al. Development and validation of a new, simplified endoscopic activity score for Crohn s disease: the SES-CD. Gastrointest Endosc 2004;60: Rutgeerts P, Geboes K, Vantrappen G, et al. Predictability of the postoperative course of Crohn s disease. Gastroenterology 1990; 99: Hébuterne X, Colombel J, Bouhnik Y, et al. Endoscopic improvement of mucosal lesions in patients with moderate to severe ileocolonic Crohn s disease following treatment with certolizumab pegol. Gut 2012 Jun 19 [Epub ahead of print]. 48. Ferrante M, Colombel JF, Sandborn WJ, et al. Evolution of endoscopic activity scores in patients with Crohn s disease under azathioprine and/or infliximab: a post-hoc analysis of the SONIC data (abstr). Gastroenterology 2011;140:S Baron JH, Connell AM, Kanaghinis TG, et al. Out-patient treatment of ulcerative colitis. Comparison between three doses of oral prednisone. Br Med J 1962;2: Travis SP, Schnell D, Krzeski P, et al. Developing an instrument to assess the endoscopic severity of ulcerative colitis: the Ulcerative Colitis Endoscopic Index of Severity (UCEIS). Gut 2012;61: Hanauer SB, Feagan BG, Lichtenstein GR, et al. Maintenance infliximab for Crohn s disease: the ACCENT I randomized trial. Lancet 2002,359: Schreiber S, Khaliq-Kareemi M, Lawrance IC, et al. Maintenance therapy with certolizumab pegol for Crohn s disease. N Engl J Med 2007;357: Colombel JF, Sandborn WJ, Rutgeerts P, et al. Adalimumab for maintenance of clinical response and remission in patients with Crohn s disease: the CHARM trial. Gastroenterology 2007;132: Sandborn WJ, Feagan BG, Stoinov S, et al. Certolizumab pegol for the treatment of Crohn s disease. N Engl J Med 2007;357: Guideline on the development of new medicinal products for the treatment of Crohn s disease. GB/document_library/Scientific_guideline/2009/09/WC pdf. 56. Pariente B, Cosnes J, Danese S, et al. Development of the Crohn s disease digestive damage score, the Lémann score. Inflamm Bowel Dis. 2011;17: Ilnyckyj A, Shanahan F, Anton PA, et al. Quantification of the placebo response in ulcerative colitis. Gastroenterology 1997; 112: Su C, Lichtenstein GR, Krok K, et al. A meta-analysis of the placebo rates of remission and response in clinical trials of active Crohn s disease. Gastroenterology 2004;126: Reprint requests Address requests for reprints to: Geert D Haens, MD, PhD, Department of Gastroenterology, Academic Medical Center, Meibergdreef 9, PO Box 22700, 1100 DE Amsterdam, The Netherlands. g.dhaens@amc.uva.nl; fax: (31) Acknowledgments Professor Marc Lémann, an expert in IBD trial design, president of the French IBD research group GETAID, member of the IOIBD Clinical Trial Task Force, and founder of the ECCO Clinical Trial Committee, died suddenly on August 26, This paper, which is based on lectures given by Professor Lémann in April 2010, is dedicated to his memory. Marc was a great friend to us all. Conflicts of interest G. D Haens has received consultancy fees from Abbott Laboratories, ActoGeniX, Centocor, Cosmo, engene, Ferring Pharmaceuticals, GlaxoSmithKline, Janssen Biologics, Millennium Pharmaceuticals, MSD, Novo Nordisk, PDL BioPharma, Pfizer, SetPoint, Shire, Takeda, Teva, and UCB; research grants from Abbott Laboratories, Janssen Biologics, Given Imaging, MSD, Falk Pharma, and PhotoPill; and speaking honoraria from Abbott Laboratories, Tillotts Pharma AG, Tramedico, Ferring Pharmaceuticals, MSD, UCB, Norgine, Shire, and Vifor Pharma. B. Feagan has received consultancy fees from Synta Pharmaceuticals, Millennium Pharmaceuticals, Merck, Centocor, Elan/Biogen, Janssen-Ortho, Protein Design Labs, Isis Pharmaceuticals, Teva Pharmaceuticals, Santarus, Schering-Plough, Bristol-Myers Squibb, Celgene, CombinatoRx, UCB Pharma, Napo Pharma, Abbott Laboratories, Procter & Gamble, Osiris, Berlex, AstraZeneca, Gene Logic, Cerimon Pharmaceuticals, Tioga Pharmaceuticals, Serono, Genentech, Tillotts Pharma AG, Unity Pharmaceuticals, Albireo Pharma, Given Imaging, Salix Pharmaceuticals, Ore Pharmaceuticals (previously Gene Logic), Novo Nordisk, GlaxoSmithKline, ActoGeniX, Prometheus Therapeutics and Diagnostics, Athersys, Alba Therapeutics, Axcan, Funxional Therapeutics, Gilead Sciences, Nektar, Pfizer, Shire, Wyeth, and Zealand Pharma; has received research grants from Merck, Otsuka, Millennium Pharmaceuticals, Tillotts Pharma AG, Abbott Laboratories, Protein Design Labs, Boehringer Ingelheim, Novartis, Centocor, Berlex, Synta Pharmaceuticals, Schering Canada, Elan/ Biogen, UCB Pharma, Bristol-Myers Squibb, Procter & Gamble, Osiris, Genentech, CombinatoRx, ActoGeniX, and Wyeth; has received speaking honoraria from UCB, Abbott Laboratories, and J&J/Janssen; and is a member of the scientific advisory board for Protein Design Labs, AstraZeneca, Elan/Biogen, Celltech, Synta Pharmaceuticals, Merck, Celgene, Novartis, Given Imaging, UCB Pharma, Salix Pharmaceuticals, Abbott Laboratories, Centocor, Pfizer, Axcan, Tillotts Pharma AG, and Prometheus Laboratories. J-F Colombel has served as a speaker, a consultant, and/or an advisory board member for Abbott Laboratories, ActoGeniX, Affilogic, Albireo Pharma, AstraZeneca, Bayer Schering Pharma, Biogen Idec, Boehringer Ingelheim, Bristol-Myers Squibb, Cellerix SL, ChemoCentryx, Centocor, Cosmo Technologies, Danone France, Elan Pharmaceuticals, Falk Pharma, Ferring Pharmaceuticals, Genentech, Giuliani SPA, Given Imaging, GlaxoSmithKline, Merck, Millennium Pharmaceuticals, Neovacs SA, Ocera Therapeutics, Otsuka, PDL BioPharma, Pfizer, RiboVacs Biotech, Schering-Plough, Shire Pharmaceuticals, Synta Pharmaceuticals, Takeda, Teva Pharmaceuticals, Therakos, UCB Pharma, and Wyeth; has received research grants from AstraZeneca, Ferring Pharmaceuticals, Schering-Plough, UCB Pharma, Lesaffre, Giuliani SPA, Danisco, Ocera Therapeutics, Danone, Roquette Pharma, Mapi-Naxis, and Disphar; and holds stock/stock options in

9 December 2012 CHALLENGES OF RCTs FOR IBD 1469 Intestinal Biotech Development. W. J. Sandborn has received consulting fees from Abbott Laboratories, ActoGeniX, AGI Therapeutics, Alba Therapeutics, Albireo, Alfa Wasserman, Amgen, AM-Pharma BV, Anaphore, Astellas, Athersys, Atlantic Healthcare, Aptalis, BioBalance, Boehringer Ingelheim, Bristol-Myers Squibb, Celgene, Celek Pharmaceuticals, Cellerix SL, Cerimon Pharmaceuticals, ChemoCentryx, CoMentis, Cosmo Technologies, Coronado Biosciences, Cytokine PharmaSciences, Eagle Pharmaceuticals, engene, Eli Lilly, EnteroMedics, Exagen Diagnostics, Ferring Pharmaceuticals, Flexion Therapeutics, Funxional Therapeutics, Genzyme, Gilead Sciences, Given Imaging, GlaxoSmithKline, Human Genome Sciences, Ironwood Pharmaceuticals, KaloBios Pharmaceuticals, Lexicon Pharmaceuticals, Lycera, Meda Pharmaceuticals, Merck Research Laboratories, Merck Serono, Millennium Pharmaceuticals, Nisshin Kyorin Pharmaceuticals, Novo Nordisk, NPS Pharmaceuticals, Optimer Pharmaceuticals, Orexigen Therapeutics, PDL BioPharma, Pfizer, Procter & Gamble, Prometheus Laboratories, ProtAb, PurGenesis Technologies, Relypsa, Roche, Salient Pharmaceuticals, Salix Pharmaceuticals, Santarus, Schering-Plough, Shire Pharmaceuticals, Sigmoid Pharma, Sirtris Pharmaceuticals, SLA Pharma UK, Targacept, Teva Pharmaceuticals, Therakos, Tillotts Pharma AG, TxCell SA, UCB Pharma, Viamet Pharmaceuticals, Vascular Biogenics, Warner Chilcott UK, and Wyeth; has received research grants from Abbott Laboratories, Bristol-Myers Squibb, Genentech, GlaxoSmithKline, Janssen, Millennium Pharmaceuticals, Novartis, Pfizer, Procter & Gamble, Shire Pharmaceuticals, and UCB Pharma; has received payments for lectures/speaker bureaus from Abbott Laboratories, Bristol-Myers Squibb, and Janssen; and holds stock/stock options in EnteroMedics. W. Reinisch has served as a speaker, a consultant, and/or an advisory board member for Abbott Laboratories, Aesca, Amgen, AstraZeneca, Biogen Idec, Bristol-Myers Squibb, Cellerix, ChemoCentryx, Centocor, Danone Austria, Elan, Ferring Pharmaceuticals, Genentech, Lipid Therapeutics, Millennium Pharmaceuticals, Mitsubishi Tanabe Pharma, MSD, Novartis, Ocera, Otsuka, PDL BioPharma, Pharmacosmos, Pfizer, Procter & Gamble, Prometheus, Schering-Plough, Shire, Takeda, Therakos, TiGenix, UCB, Vifor, Yakult Austria, and 4SC. P. Rutgeerts received consulting fees from Millennium/Takeda, Genentech/Hoffman-La Roche, Neovacs, Merck/Serono, Bristol-Myers Squibb, Robarts, Tillotts Pharma AG, Pfizer, and Falk Pharma and has served as a speaker and a consultant and received payments for lectures/research support from Centocor, Merck, UCB, and Abbott Laboratories. F. Carbonnel discloses no conflicts. J-Y Mary discloses no conflicts. S. Danese has received consulting fees from Schering-Plough, AstraZeneca, Abbott Laboratories, Novo Nordisk, and Millennium/Takeda and lecture fees, including fees for service on speaker bureaus, from UCB Pharma, Ferring Pharmaceuticals, Vifor, and Merck Sharp & Dohme. R. N. Fedorak has served as an advisory board member for Abbott Laboratories, Ferring Pharmaceuticals, Merck, Shire, and VSL#3; is a recipient of clinical/basic research grants from Abbott Laboratories, Alba, Axcan, Bristol-Myers Squibb, Centocor, Ferring Pharmaceuticals, GlaxoSmithKline, Genentech, Merck, Millennium Pharmaceuticals, Novartis, Otsuka, Procter & Gamble, Roche, and VSL#3; and is owner/shareholder of Metabolomic Technologies.