|Home | About | Journals | Submit | Contact Us | Français|
In March 2008, a roundtable discussion was convened by the inflammatory bowel disease (IBD) specialist panel the BRIDGe (Building Resources and Research in IBD Globally) group, which consists of junior faculty gastroenterologists who have undergone advanced fellowship training at IBD referral centers in the United States, Canada, the United Kingdom, and Australia. An agenda was formulated to discuss three current controversies in Crohn's disease management: the role of 5-aminosalicylates, the use of biologic combination therapy versus monotherapy, and the use of step-up therapy versus top-down therapy for Crohn's disease. The aim of the meeting was three-fold: to review the data pertaining to each topic; to collect opinions from the participants as to their analysis of the literature and their current practice; and, where possible, to formulate recommendations of current best practice given the available evidence. This manuscript summarizes the discussions on these three areas of controversy in the current management of Crohn's disease.
Treatment paradigms in Crohn's disease are rapidly evolving, advancing from the current standard of inducing and maintaining corticosteroid-free remission to achieving mucosal healing and thereby potentially altering the natural history of disease. With these advances, the efficacies of older treatments are being questioned, and newer treatment strategies are unearthing controversies and issues of their own. In March 2008, a roundtable meeting was organized to discuss three areas of current controversy in the management of Crohn's disease: the role of 5-aminosalicylates (5-ASAs), the use of biologic combination therapy versus monotherapy, and the use of step-up versus top-down treatment of Crohn's disease. A literature review was performed, and a set of key publications was provided to serve as background for the discussion (Table 1). This article reviews these data and summarizes the group's discussions pertaining to these controversies and, where possible, given the existing evidence, provides recommendations as to the current best practice in each clinical scenario.
After decades of widespread acceptance, the role of 5-ASAs in the management of Crohn's disease is now being challenged. Advocates of 5-ASAs have referenced their safety and the length of experience with their use, whereas skeptics have questioned their efficacy and concluded that their use simply postpones the initiation of effective therapy. The evidence supporting and refuting their use was reviewed by the group.
The National Cooperative Crohn's Disease Study, published in 1979, was a large two-part, randomized, placebo-controlled trial comparing sulfasalazine, predni-sone, and azathioprine for the induction and maintenance of remission in 569 patients with Crohn's disease. In the 17-week induction study, 43% of patients receiving sulfasalazine (1 g/15 kg daily, mean dose of 4.7 g daily) entered remission compared to 30% of patients receiving placebo. Although this difference was not statistically significant (P=.08), subgroup analysis revealed that sulfasalazine was significantly more effective than placebo for patients with colonic (P=.006) or ileocolonic (P=.027) disease. However, sulfasalazine was not effective in patients already exposed to steroids, likely indicating that the need for corticosteroids signifies a group of patients with more severe disease. In the 24-month maintenance study of 274 patients with medically or surgically induced quiescent disease, sulfasalazine (0.5 g/kg daily, mean dose 2.5 g daily) was no more effective than placebo at maintaining remission, irrespective of disease location (P=.5).1
This early study identified two important principles of 5-ASA usage in Crohn's disease that have been subsequently validated. First, 5-ASAs are most effective in patients with colonic disease, and second, they are ineffective in patients who required corticosteroids in the past, essentially confining their role to patients with mild disease, particularly those who have been recently diagnosed.
A meta-analysis of controlled-release mesalamine (Pentasa, Shire) in mild-to-moderate Crohn's disease explored the use of 5-ASAs as induction agents. In this analysis, three 16-week randomized, placebo-controlled trials comparing 4 g of controlled-release mesalamine to placebo were included. One study, which had previously been published, was positive,2 whereas the other two studies were negative (P=.7 and P=.5),3,4 causing some concern for the potential of publication bias. Nevertheless, when the three trials were combined, the intention-to-treat meta-analysis showed a 63-point decrease in Crohn's Disease Activity Index (CDAI) in the treatment group compared to a 45-point placebo reduction in CDAI (P=.04).5 The clinical significance of this small 18-point CDAI difference is debatable.
A Cochrane meta-analysis was performed to evaluate the role of 5-ASAs in maintaining medically induced remission in Crohn's disease. This analysis consisted of six randomized controlled trials of at least 6 months' duration comparing mesalamine to placebo for the maintenance of medically induced remission and revealed an odds ratio (OR) of 1.0 for relapse with mesalamine (95% confidence interval [CI], 0.80–1.24). It was concluded that given the conclusively negative result, further randomized trials examining 5-ASAs for this indication are unjustified.6 An earlier meta-analysis of 15 randomized controlled trials revealed that 5-ASAs were only effective at reducing symptomatic relapse after surgically, not medically, induced remission, though the number needed to treat to prevent 1 relapse was 16.7
In contrast to the paucity of evidence supporting their use in Crohn's disease, it is widely acknowledged that 5-ASAs remain a common therapy in clinical practice. When asked to answer affirmatively or negatively, only 4 of the 12 meeting attendees continue to initiate 5-ASA therapy in Crohn's disease patients, though no participants proactively discontinue them in patients who are tolerating them and doing well. It was recognized that in clinical practice the safety profile of 5-ASAs could make them a reasonable first-line agent, though they should be reserved for patients with mild colonic disease of recent onset. The meeting participants agreed that patients initiated on 5-ASAs require close clinical follow-up, as early as 4–8 weeks after the start of therapy, which should identify treatment failures, allowing timely escalation of therapy when necessary.
There is an increasing, albeit inconclusive, body of evidence supporting the chemopreventive properties of 5-ASAs in ulcerative colitis.8 This beneficial effect appears to be in addition to the anti-inflammatory effects of these medications and requires doses of at least 1.2 g daily of 5-ASAs.9 Intuitively, this chemopreventive potential should also be applicable to Crohn's colitis patients, though as of yet there is insufficient evidence to support 5-ASA usage for this indication alone.10 Use of 5-ASAs for this speculative indication may be appropriate in patients with additional carcinogenesis risk factors such as a family history of colorectal cancer or primary sclerosing cholangitis
After a decade of experience, the optimal approach to using infliximab (Remicade, Centocor) to induce and maintain remission in Crohn's disease remains to be determined. In the era of episodically administered biologic therapies, co-administration of immunomodulators, either thiopurines or methotrexate, was strongly encouraged as a means of reducing immunogenicity and therefore increasing the duration of response to these agents. As the standard of care has appropriately moved onto the use of scheduled maintenance biologic therapies, the need for concomitant immunomodulator therapy has been questioned. The advantages and disadvantages of combination therapy versus anti-tumor necrosis factor (TNF) monotherapy and the relevant clinical trial data pertaining to this issue were discussed by the panel.
Concomitant administration of immunomodulators and biologic therapies confers the advantage of reduced immunogenicity to the biologic agent, in turn maximizing the tolerability and duration of response to these therapies. The main reason for the recent heightened debate of this issue has involved concerns regarding the safety of combination therapy, in particular the risk of infections or malignancies. These discussions have been triggered by the reporting of a uniformly fatal form of T-cell lymphoma in inflammatory bowel disease (IBD) patients receiving dual immunomodulation. As of May 2008, postmarketing surveillance has identified 15 cases of hepatosplenic T-cell lymphoma (HSTCL) in IBD patients receiving both anti-TNF agents (13 cases on infliximab and 2 on adalimumab [Humira, Abbott] who had previously received infliximab) and thiopurines (personal communication, Centocor, Malvern, PA). All but 1 case have occurred in males, and most have been young patients with Crohn's disease. No cases of HSTCL have been reported in patients treated with methotrexate or infliximab monotherapy, although 9 cases have been reported in patients on thiopurines alone. In addition, combination therapy increases the risk for opportunistic infections. A case-control study of 100 IBD patients with known opportunistic infections demonstrated, on univariate analysis, that both thiopurines (OR, 3.1; 95% CI, 1.7–5.5) and infliximab (OR, 4.4; 95% CI, 1.2–17.1) were individually associated with an increased infection risk. Multivariate analysis revealed that the use of combinations of either 2 or 3 of corticosteroids, thiopurines, or infliximab yielded an OR of 14.5 (95% CI, 4.9–43) for opportunistic infection, with the elderly at the greatest risk.11
Clinical trial data pertaining to combination therapy versus monotherapy are best reviewed by considering whether infliximab was administered episodically or as scheduled maintenance therapy. Episodic infliximab is immunogenic, and the presence of antibodies to infliximab (ATI) is associated with a reduced duration of response and an increased risk of acute infusion reactions to subsequent infliximab doses.12 Analysis of a cohort of 174 Crohn's disease patients treated with episodic infliximab and followed prospectively revealed that concomitant immunomodulator use decreased the incidence of ATI formation from 73% to 46% (P<.01). This reduction was equally evident in patients treated with azathioprine (48%) or methotrexate (44%). Consistent with data from maintenance trials, patients taking immunomodulators experienced higher infliximab levels 4 weeks after infusions compared to patients taking infliximab monotherapy (median, 6.45 µg/mL vs 2.42 µg/mL; P=.065).13
The ACCENT I trial design included groups receiving both episodic and scheduled maintenance infliximab therapy. In both the 5 mg/kg and 10 mg/kg scheduled maintenance groups, there was no difference in the prevalence of ATI between patients receiving and those not receiving concomitant immunomodulators. Conversely, in the episodically treated patients, 38% of patients receiving infliximab alone developed ATI compared to 16% of patients treated with immunomodulators. In the entire cohort, infusion reactions were more common in ATI-positive patients than ATI-negative patients (36% vs 24%; P=.026). However, similar proportions of ATI-positive patients and ATI-negative patients achieved clinical response (64% and 62%, respectively; P=.35) and clinical remission (41% and 39%, respectively; P=.76) at 1 year.14 Nevertheless, in view of the effects upon ATI formation, it can be reasonably concluded that if infliximab is administered episodically, concomitant immunomodulator ther-apy is recommended.
Scheduled maintenance trials using infliximab, adali-mumab, and certolizumab pegol (Cimzia, UCB) demonstrate similar 1-year clinical response and remission rates in patients receiving and those not receiving concomitant immunomodulators.15–17 However, fewer than half of the patients in these trials received immunomodulators, and clinical outcomes beyond 12 months remain unknown. Additionally, only clinical response and remission rates for each treatment strategy have been determined, and mucosal healing rates with monotherapy versus combination therapy are unknown.
Another issue is whether to initiate anti-TNF and immunomodulator therapy concurrently in previously unexposed patients. Recently, the equivalence of combination therapy and monotherapy was again demonstrated in the COMMIT trial, which examined 126 patients randomized to induction and maintenance infliximab therapy, with or without concomitant methotrexate, in addition to a tapering schedule of corticosteroids. There was no difference in remission rates between the infliximab monotherapy group and the combination therapy group at either 14 weeks (77.8% vs 76.2%, respectively; P=.83) or 50 weeks (57.1% vs 55.6%, respectively; P=.86), and adverse event frequencies were similar in both groups. Data regarding immunogenicity and infliximab concentrations are yet to be published.18
The panel reached a consensus that patients receiving episodic infliximab should be treated with concomitant immunomodulators to ensure that the patients are “optimized, not immunized” to biologic therapies. Although episodic dosing is inferior to scheduled main-tenance therapy, pharmacoeconomic restrictions limit the availability of infliximab to episodic dosing in some healthcare systems worldwide; in this scenario, thiopur-ines or methotrexate should be used at their usual therapeutic doses. When infliximab is administered as scheduled maintenance therapy, the question of whether concomitant immunomodulators should be used remains unanswered. However, the trend appears to be moving toward the use of monotherapy; when asked to answer affirmatively or negatively, all 12 meeting attendees favored monotherapy over combination therapy, though they acknowledged that this decision should be determined on a case-by-case basis.
Given the incomplete evidence base, formal guide-lines regarding the use of combination therapy or mono-therapy do not yet exist. However, a variety of factors should be taken into account. Loss of response to inflix-imab is of less consequence if alternative anti-TNF agents such as adalimumab or certolizumab pegol are available. In this scenario, patients can be switched to one of these agents with a reasonable likelihood of ongoing response. Therefore, if alternatives are available, biologic monotherapy may be a plausible option. In contrast, if only infliximab is available, then loss of response carries more significant consequences, in which case continuing immunomodulators may be considered. A reasonable compromise may be to discontinue immunomodulators following a defined period of combination therapy, for example, after 6 months, and then to continue monotherapy thereafter. Another option is to switch the immunomodulator to low-dose oral methotrexate, 7.5 mg weekly, which is known to reduce immunogenicity and has been suggested in rheumatologic studies to act synergistically with infliximab, possibly by reducing its renal clearance.19
To date, the decision of whether to discontinue immunomodulator therapy has applied to patients already on combination therapy. However, this question now also has relevance for patients initiating induction infliximab therapy. Although COMMIT trial data suggest that infliximab monotherapy is adequate from the time of initiation, longer-term outcome data of this strategy are unknown.18 It is hoped that data from the ongoing early intervention SONIC trial, which compares combination therapy to monotherapy with either infliximab or azathioprine, will help to further clarify this dilemma.
Conventional management of Crohn's disease has traditionally involved a “step-up” treatment strategy, in which the least potent, but safest, medications are used first and then stronger, but potentially more harmful, agents are sequentially added, if required, until remission is achieved. This approach has typically involved the initial use of 5-ASAs or antibiotics, followed as needed by corticosteroids, immunomodulators, and biologic agents, in that order. Recently, this strategy has been challenged due to its ineffectiveness in altering the natural history of disease. In turn, a “top-down” treatment strategy has been proposed, in which the most potent agents, biologic therapies, and immunomodulators are utilized soon after diagnosis in an effort to rapidly induce a steroid-free remission and thereby prevent progression toward fibrostenosing and penetrating disease complications.
Recent treatment trends in Crohn's disease have appropriately moved toward more judicious use of cor-ticosteroids as well as the earlier introduction of immuno-modulators, but this has not translated to evidence that the natural history of disease is being altered. The requirement for corticosteroids is now recognized as a marker of more severe disease and a “tipping point” at which alternate steroid-sparing strategies must be introduced. A population-based study of an inception cohort of newly diagnosed Crohn's disease patients showed that only 43% of patients required steroids within 1 year of diagnosis, though of those patients who received these agents, only 32% remained in a steroid-free remission without surgery at 1 year.20 Increased caution with corticosteroid usage has also arisen from registry data that independently associate their usage with increased mortality.21
Accordingly, the use of immunomodulators has progressively increased in the last two decades, as demonstrated by a retrospective analysis of a large referral center cohort, in which the probability of receiving an immunomodulator during a 5-year period increased from 0% in patients examined between 1978 and 1982 to 56% in patients treated between 1998 and 2002 (P<.001). Despite this finding, rates of surgical resection during this time remained unchanged (from 35% to 34% at 5 years; P=.81), as did the frequency of fibrostenosing or penetrating complications, suggesting that immunomodulators are ineffective at altering the natural history of disease.22 These data may underestimate the effect of these agents, as 43% of patients underwent surgery within the first 3 months of diagnosis, and of these surgery patients, only 9% had been receiving immunomodulators for at least 3 months.
Skeptics of top-down therapy deem that its widespread usage would unnecessarily expose large numbers of patients to potentially harmful therapies that are otherwise needed in only a minority of patients. Population-based studies have demonstrated that fewer than half of patients require corticosteroids, which is consistent with this theory.20 Most concerning is the increased incidence of infections and malignancies, in particular lymphoma, that are associated with combined immunosuppression. Infection rates are significantly increased with combination therapy, and single-center studies have reported increased mortality rates in patients receiving infliximab, though this has not been reproduced in registry data. The elderly in particular appear to be susceptible to infectious complications.11,21,23
The risk of lymphoma in combination therapy has recently been dominated by the reporting of HSTCL cases, though it is unclear whether infliximab (or any other anti-TNF agent) alone increases the risk of lymphoma. A recent meta-analysis appears to suggest that infliximab exposure does increase the risk of non-Hodgkin lymphoma (NHL). Standardized incidence ratios (SIRs) of NHL in infliximab-treated patients were increased when compared to the expected rates in the general population (SIR, 3.23; 95% CI, 1.5–6.9). When compared to patients treated with immunomodulators alone, the risk was also increased; however, this finding was not statistically significant (SIR, 1.70; 95% CI, 0.5–7.1).24,25 Although the absolute rates of adverse events with combined immunosuppression are low, the potential consequences are significant, and these risks must be discussed with patients when selecting a treatment strategy.
The recent publication of results from a large multicenter European early intervention trial conducted between 2001 and 2004 has brought the concept of top-down therapy to the forefront. One hundred and thirty-three patients with newly diagnosed Crohn's disease or active disease of less than 4 years of duration (CDAI >200 points) were randomized to receive open-label early combined immunosuppression (top-down) or conventional (step-up) treatment and were followed for 2 years. All patients were naive to steroids, immunomodulators, and biologics. Top-down treatment was comprised of infliximab 5 mg/kg at 0, 2, and 6 weeks, and azathioprine (2–2.5 mg/kg daily) or meth-otrexate (25 mg subcutaneously weekly for 12 weeks, followed by 15 mg weekly) if azathioprine was not tolerated. Further disease flares were treated with episodic infliximab and finally corticosteroids, if required. Patients in the step-up group received corticosteroids, and upon flaring, an additional course of steroids was administered before patients received immunomodulators and finally a 3-dose induction course of infliximab, if required. Coprimary endpoints were defined as remission without steroids or surgery at 26 and 52 weeks, and secondary endpoints included safety and mucosal healing as assessed by an endoscopic substudy.
Both coprimary endpoints were reached at 26 weeks; 60% of the top-down group were in a steroid-free remission without surgery in comparison to 35.9% of the step-up group (P<.01). Corresponding rates at Week 52 were 61.5% versus 42.2%, respectively (P=.03). After 52 weeks, there was no difference in the proportion of patients in remission in each group. Cumulative corticosteroid exposure was significantly less in the top-down treatment group (mean days of steroid usage in 2 years, 5.6 vs 79.7; P<.01), though this finding is not entirely unexpected, given the trial design. Forty-nine patients entered the endoscopic substudy; although appearances improved in both groups, 73% of top-down patients demonstrated complete disappearance of mucosal ulceration compared to 30% of the step-up group (P<.01). Endoscopic healing did not correlate with CDAI-defined remission. The frequency of adverse events was similar between groups (30.8% compared to 25.3% in the top-down and step-up groups, respectively; P=1.0), though 1 patient receiving early combined immunosuppression developed demyelination.26
Although these impressive results deserve the attention they have received, several obvious flaws in the design of the trial must be taken into consideration. Initially, the trial, though randomized, was open-label and therefore susceptible to bias. Two courses of steroids were administered prior to the introduction of immunomodulators in the step-up group, possibly reflecting practice patterns during the trial recruitment period several years earlier, whereas most experts currently recommend introducing immunomodulators following the failure to taper off the first course of steroids, if not earlier in several cases. Finally, and most importantly, episodic rather than scheduled maintenance infliximab therapy was administered after the initial induction course in top-down patients, which may have contributed to the lack of difference between the groups in the second year of the trial. Despite these design flaws, both predefined coprimary endpoints were achieved, and the results pertaining to mucosal healing, in particular, are remarkable, though entry details regarding the patients in the endoscopic substudy have not been provided.
Mucosal healing intuitively equates to altering the natural history of disease, and post-hoc analysis of the ACCENT I trial revealed a trend toward patients with healed mucosa having fewer Crohn's disease–related hos-pitalizations.27 In addition, prospective studies in Crohn's disease have shown that patients who achieve mucosal healing will experience decreased ongoing inflammation and reduced steroid requirements in the future.28,29 However, its true significance in clinical practice remains unknown, and interestingly, in this study mucosal healing rates did not correlate with CDAI-defined remis-sion.26 Which agent contributed to mucosal healing cannot be ascertained from this trial, as it could have been infliximab, azathioprine, the combination of both, or the reduction in corticosteroid exposure that allowed for the restoration of mucosal integrity. Although mucosal healing is a justifiable goal, whether treatment should be escalated in clinically well patients in order to achieve this goal is debatable. Dose escalation of existing therapies such as 5-ASAs or immunomodulators in order to achieve mucosal healing would appear to be appropriate, but the risks of initiating biologic agents solely for this indication may outweigh the benefits and at present this cannot be recommended.
Top-down therapy inverts the therapeutic pyramid in the management of Crohn's disease. However, this theoretically intriguing concept cannot currently be recommended for all patients with moderate-to-severe Crohn's disease. When asked to answer affirmatively or negatively, all 12 meeting attendees agreed that the current treatment algorithms for moderate-to-severe Crohn's disease should not be altered to include top-down therapy at present, though all participants concurred that the trend toward the earlier use of immunomodulators is appropriate. Additional data are required from ongoing and future early intervention trials, preferably using scheduled maintenance anti-TNF agents, before the place of top-down therapy in current clinical practice can be better defined.
Identifying predictors of aggressive disease will help us stratify patients who are most appropriate for top-down therapy. Predictors of aggressive disease in an individual patient may be clinical or based upon biomarkers, including serologic and genetic profiles. Clinical predictors of complicated disease include young age of onset, the requirement for corticosteroids at diagnosis, fibrostenosing or penetrating disease at diagnosis, perianal disease, active smoking, and extensive intestinal involvement.30,31 The role of serologic markers in predicting complicated disease behavior is evolving and is not currently understood completely. However, recent studies have demonstrated a qualitative and quantitative relationship between the presence and titers of the antibodies ASCA, anti-OmpC, anti-I2, and anti-Cbir1 and fibrostenosing and penetrating small bowel Crohn's disease and the need for earlier surgery.32,33 The genotype of a patient represents another method of quantifying their risk of disease progression, though genetic advances to date have not influenced clinical decision-making. Most attention has focused on the NOD2 (CARD15) gene on chromosome 16q, which encodes a protein in innate immune cells that acts as an intracellular receptor for bacterial muramyl dipeptide. Mutations in the leucine rich repeat region of NOD2 are evident in 40% of white Crohn's disease patients, and when present, the mutations have been associated with an earlier onset of disease, fibrostenosing ileal disease, and an increased requirement for surgery.34
At present, our knowledge of clinical, serologic, and genotypic markers of complicated disease is insufficient to validate their use alone as screening tools for the identification of patients who should receive top-down therapy. However, it is hoped that in the near future, these and other biomarkers will reach clinical practice and provide a method for triaging Crohn's disease patients into different prognostic subtypes and “personalizing” treatment plans accordingly.
The excitement of new approaches for the treatment of Crohn's disease also brings uncertainty as new questions and challenges arise. Based upon the data currently available, the BRIDGe group agreed that 5-ASAs are not warranted in most patients with Crohn's disease and that treatment trends are moving toward the use of anti-TNF monotherapy. Finally, the group attendees agreed that the possibility that top-down therapy has the ability to alter the natural history of Crohn's disease is enticing and the resultant attention is well justified. However, before current Crohn's disease management algorithms can be reversed, clarification is needed as to which patients will benefit most from this treatment strategy in order to justify its not insignificant risks. The answers to these issues and the subsequent improvements in clinical care are eagerly awaited by IBD practitioners and, most importantly, their patients.
The authors would like to thank Procter & Gamble Pharmaceuticals Inc. for the unrestricted educational grant given to the BRIDGe group to support this meeting and manuscript production.