Article Text

Download PDFPDF

Early use of immunosuppressives or TNF antagonists for the treatment of Crohn's disease: time for a change
  1. Ingrid Ordás1,2,
  2. Brian G Feagan3,
  3. William J Sandborn1
  1. 1Division of Gastroenterology, University of California San Diego, La Jolla, California, USA
  2. 2Division of Gastroenterology, Hospital Clinic, CIBER-EHD, IDIBAPS, University of Barcelona, Barcelona, Spain
  3. 3Division of Gastroenterology, Robarts Research Institute, University of Western Ontario, London, Ontario, Canada
  1. Correspondence to Dr William J Sandborn, Division of Gastroenterology, University of California San Diego, 9500 Gilman Drive, Building UC 303, Room 220, La Jolla, CA 92093-0063, USA; wsandborn{at}


Crohn's disease is a chronic inflammatory disorder that follows a progressive and destructive course. Ultimately, uncontrolled inflammation leads to bowel damage from disease-related complications such as strictures, fistulas and abscesses and surgical resection. Conventional ‘step-care’, whereby corticosteroids and immunosuppressives are prescribed sequentially, is an incremental approach to treatment that does not prevent disease progression and conveys an important risk of adverse events from repeated courses of corticosteroids. Although the immunosuppressives azathioprine, 6-mercaptopurine and methotrexate are corticosteroid-sparing, they are not highly effective for inducing mucosal healing or preventing disease progression. Tumour necrosis factor antagonists induce and maintain mucosal healing and reduce surgery and hospitalisation rates. This holds out the possibility that long-term use of these agents may prevent bowel damage. Combination therapy with immunosuppressives and tumour necrosis factor antagonists is likely the best strategy for achieving optimal outcomes in patients at high risk of disease progression. However, accurate prognostic markers must be identified to guide patient selection. Long-term prospective studies with robust outcomes are still needed to establish definitively the efficacy and safety of early combination therapy to prevent bowel damage, loss of gastrointestinal tract function and permanent disability.

  • Crohn's disease
  • immunesuppressives
  • TNF antagonists
  • mucosal healing
  • disability
  • IBD
  • gastrointesinal endoscopy
  • inflammatory bowel disease
  • ulcerative colitis
  • IBD clinical
View Full Text

Statistics from


Crohn's disease (CD) is a chronic gastrointestinal disorder that negatively affects the quality of life of most patients with the disease.1 Data from population-based cohort studies show that the majority of patients experience a progressive course that transitions from pure inflammatory lesions to destructive complications such as stricture, fistula and abscess.2–4 These complications, and the surgical resections required to treat them, result in irreversible bowel damage that in turn leads to loss of gastrointestinal tract function and disability. Given this sequence of events, it is not acceptable to simply treat symptoms of the disease. Instead, we should strive to change the natural history and prevent disease progression, bowel damage and disability.5 To achieve these goals, new treatment paradigms must be considered for patients with early disease who are at risk of disease progression.

The concept of early therapeutic intervention to prevent disease progression comes from rheumatoid arthritis (RA). Early treatment with biological disease-modifying antirheumatic drugs in RA aims to treat the underlying synovial inflammation that results in joint destruction, impaired physical functioning and disability. Current data support the concept that early use of combined immunosuppression based on biological therapies results in better clinical and radiological outcomes than conventional treatment regimens.6–8

At present, two possible treatment strategies exist that could be considered as potentially highly effective in early CD: (1) immunosuppressives in combination with a tapering course of steroids (accelerated step-care); and (2) tumour necrosis factor (TNF) antagonists (either as monotherapy or in combination with immunosuppressives) (figure 1). This article reviews the treatment of early CD with both immunosuppressive-based regimens and those that feature TNF antagonists. An essential axiom of any early treatment paradigm is recognition that overtreatment of low-risk patients will result in a poor therapeutic index. Consequently, it is essential to identify patients who are suitable for early treatment because they are at high risk of disease progression. Benefits must always be carefully weighed against risks.

Figure 1

Proposed algorithm for treatment of early Crohn's disease, defined as disease duration <2 years and no previous use of IMS or TNF antagonists. IMS, immunosuppressive; CS, corticosteroids; TNF, tumour necrosis factor; Dg, diagnosis.

Key messages: Background

  • Crohn's disease is a chronic, progressive and destructive disease.

  • Disease behaviour changes over time leading to development of complications (strictures, fistulas and abscesses).

  • The conventional step-care treatment algorithm does not alter the natural history of the disease.

  • TNF antagonists modify the course of Crohn's disease by reducing surgery and hospitalisation rates.

  • Future treatment goals should include maintenance of normal gastrointestinal function to prevent structural bowel damage and disability.

Early Crohn's disease: an operational definition

An operational definition of early CD is essential to assess whether early intervention algorithms have greater efficacy than conventional step-care. Recently, such a definition has been proposed.9 The essential components of the definition are disease duration <2 years and no previous treatment with immunosuppressives or TNF antagonists. In addition, the absence of existing bowel damage is an important modifier for this definition. In population-based studies, 19–36% of patients newly diagnosed with CD present with bowel damage—that is, disease complications of strictures, fistulas or abscesses. Early surgical resection in such patients is not uncommon.4 10–12 Although these patients clearly have early CD by virtue of their recent diagnosis, the presence of CD-related complications (stenosis, fistula and/or abscess) precludes them from being evaluated in clinical trials where the endpoint is prevention of bowel damage. An instrument to measure the cumulative structural bowel damage in patients with CD is currently under development. This new index, the Lémann score, takes into account the anatomical, severity and extent of bowel damage measured by cross-sectional imaging techniques and history of prior surgical resection.13 Use of this index in clinical trials will allow investigators to evaluate the efficacy of new strategies to prevent bowel damage.

Conventional and evolving treatment strategies

Goals of treatment

The goals of CD treatment should include: (1) steroid-free sustained clinical remission; (2) induction and maintenance of mucosal healing; (3) potential induction and maintenance of radiological healing; (4) prevention of surgery; (5) maintenance of normal gastrointestinal function; and (6) prevention of disability. To achieve these goals, it is essential to use a therapeutic regimen able to treat the underlying inflammation and ultimately to heal the bowel instead of a symptom-based treatment approach.

In the past, surprisingly little attention has been paid to the prevention of disability and the relationship between bowel damage and disability. Even the definition of the term disability has been inconsistent in the CD literature. For the purposes of this article, disability is ‘any restriction or lack (resulting from an impairment) of ability to perform an activity in the manner or within the range considered normal for a human being’, as defined by the WHO.14

Conventional step-care therapy

Current guidelines for the management of moderate to severe CD promote a sequential and additive step-care paradigm that focuses on treatment of acute flares and maintenance of clinical remission thereafter.15–17 Use of corticosteroids is considered first-line treatment, reserving immunosuppressives (azathioprine, 6-mercaptopurine, and methotrexate) for those patients who develop corticosteroid dependency (figure 2).

Figure 2

Conventional and evolving treatment strategies. IMS, immunosuppressive; TNF, tumour necrosis factor.

Although highly effective in the short term, ultimately corticosteroids lead to low sustained remission rates, steroid refractoriness and dependency in a high proportion of patients,18 19 low mucosal healing rates20 and an increased risk of serious infections and mortality.21 Azathioprine (and 6-mercaptopurine) and methotrexate are superior to placebo for induction and maintenance of clinical remission in patients with CD,22 23 although the effect size is modest.24–28

In the conventional step-care, TNF antagonists are reserved for refractory disease or intolerance to conventional therapies. This approach has not been shown to slow or prevent disease progression to bowel damage, although it should be acknowledged that a comparison of conventional step-care to alternative treatment paradigms for the endpoints of disease progression and bowel damage has not been performed.

Evolving approaches: accelerated step-care and early combined immunosuppression

Accelerated step-care

In 1995, Candy and colleagues showed that azathioprine was superior to placebo for the maintenance of remission induced by corticosteroids in combination with azathioprine in patients with CD.22 In this study, 63 patients with active disease were treated with a tapering course of corticosteroids together with azathioprine (2.5 mg/kg/day) for a period of 12 weeks. After 15 months of follow-up, a higher proportion of patients in the azathioprine arm were in remission compared with the placebo group (42% vs 7%; p=0.001). Despite these promising results, it was not until the last decade that the concept of accelerated step-care evaluated in this trial evolved into the earlier introduction of immunosuppressives in patients with newly diagnosed disease.

The first study to assess the effectiveness of this approach in paediatrics was a randomised placebo-controlled trial that compared 6-mercaptopurine in combination with corticosteroids to corticosteroid monotherapy in 53 children with early (disease duration ≤8 weeks) moderate to severely active CD.29 Combination therapy with corticosteroids (induction) and 6-mercarptopurine for maintenance was superior to placebo over 18 months of follow-up (9% of patients relapsed in the 6-mercaptopurine group compared with 47% in the placebo group, p=0.007). Although the absolute risk reduction observed was relatively high (38%), the small sample size of the trial generated a wide CI around the observed difference in relapse rates (38%). A randomised placebo-controlled trial of azathioprine in combination with corticosteroids has recently been reported in an adult population.30 Sans and coworkers evaluated the efficacy of early introduction of azathioprine in patients with recently diagnosed (≤8 weeks) CD. Patients (n=131) were randomised to receive either azathioprine (2.5 mg/kg/day, n=68) or placebo (n=63); the disease activity at baseline as measured by the mean Crohn's Disease Activity Index (CDAI) score was 120 points in the azathioprine group and 90 points in the placebo group. After 18 months of follow-up, corticosteroid-free sustained remission rates did not differ between patients assigned to early initiation of azathioprine and those who were treated with conventional step-care (44.1% vs 38.1%; p=0.52). In summary, the randomised trials that have evaluated the early use of immunosuppressives have been relatively small and have not shown a consistent benefit. Likewise, the data supporting the early use of immunosuppressives to prevent surgery or complications are difficult to interpret. No randomised studies have evaluated this outcome. While some retrospective studies suggest a benefit of early introduction of azathioprine in reducing the rates of surgery,31–33 others have not demonstrated this effect.34 35

Early combined immunosuppression

The early use of combined immunosuppression has been adapted from an emerging therapeutic paradigm in RA. For patients at high risk of disease progression (joint damage), a ‘top-down’ approach has evolved that features early use of methotrexate in combination with a TNF antagonist. The first major concept driving this paradigm was recognition that combined therapy (TNF antagonist plus methotrexate) was more effective than either agent used as monotherapy.36 The second was that early treatment with a TNF antagonist was able to prevent progression of joint destruction which is highly correlated with loss of function.37 38 Finally, the availability of objective predictors of disease progression facilitated selective use of this approach in high-risk patients. While long-term data are still lacking, widespread utilisation of this new paradigm may change the natural history of this debilitating disease.

The rationale for top-down treatment of early CD is based on data from multiple sources.39–42 Subgroup analyses from two placebo controlled trials of adalimumab42 and certolizumab41 suggested that patients with early CD may experience greater efficacy following treatment with TNF antagonists than those with longstanding disease.

To date, four major randomised trials have evaluated the use of combined TNF antagonist and immunosuppressive therapy for induction of remission in patients with active CD.39 40 43 44 The first of these, a 1-year trial conducted by Lémman and colleagues,43 randomly assigned 113 patients with corticosteroid-dependent CD to receive infliximab induction therapy (weeks 0, 2, and 6) or placebo. All patients received concomitant azathioprine treatment. One-half of the patients were receiving azathioprine prior to randomisation. The proportion of patients assigned to the infliximab group in corticosteroid-free remission at weeks 12, 24 and 52 were 75%, 57% and 40%, respectively, compared with 38%, 29% and 22% for those who received placebo. All of these comparisons were statistically significant. Combination therapy was more effective in both subgroups of patients who had previously received azathioprine and those who were naïve to azathioprine. Although this study clearly demonstrated that combination therapy was superior to azathioprine monotherapy in corticosteroid-dependent CD, these data may understate the efficacy of combined immunosuppression since infliximab was only used during the induction component of the trial. This may also help to explain the decrease in therapeutic response over time between both treatment strategies.

The second trial to evaluate combined therapy was performed in patients with relatively new onset disease (average duration 6 months) who had not previously been exposed to TNF antagonists, immunosuppressives or corticosteroids.39 D'Haens and coworkers randomly assigned 133 patients with active CD to a regimen consisting of azathioprine (methotrexate was substituted if intolerance to azathioprine occurred) and infliximab given as a 0, 2, 6 week induction regimen (infliximab was given episodically during the maintenance phase if needed for re-induction) or conventional step-care consisting of corticosteroids followed by azathioprine (or methotrexate) and ultimately, if response was not obtained, infliximab. The co-primary endpoints of the study were corticosteroid-free remission at 6 months and 1 year. Patients assigned to the early combined immunosuppression regimen were more likely to be in remission at both of these times than those assigned to conventional step-care (week 26: 60.0% vs 35.9%, p=0.006; week 52: 61.5% vs 42.2%, p=0.03). Furthermore, the proportion of patients without ulcers on endoscopy at 2 years was substantially higher in the group who received combined therapy (73.1% vs 30.4%, p=0.003). The presence of mucosal healing at year 2 was associated with higher rates of corticosteroid-free remission and lower rates of hospitalisation and surgery during years 3 and 4.45 These results strongly suggest that introduction of combined immunosuppression early in the course of CD is a superior strategy to the traditional step-care approach. However, the design of the trial was criticised because it was not blinded and because of the delay in introducing immunosuppressives in the step-up care group (patients were required to fail two courses of corticosteroids before proceeding to treatment with azathioprine/methotrexate). Although some experts speculated whether similar results could have been obtained had azathioprine been introduced at the same time as corticosteroid therapy in the step-care arm, others emphasised that the findings underestimate the efficacy of the combination because of the intermittent use of infliximab in the experimental arm. No infliximab monotherapy arm was included in the trial so it was not possible to evaluate whether the early introduction of infliximab alone was responsible for the observed differences between the treatment groups.

Many of these issues were addressed in the recently completed SONIC trial.40 Colombel and colleagues randomly assigned 508 patients with active CD, with a mean disease duration of 2 years, to one of three treatment regimens: azathioprine alone, infliximab alone or the combination of the two drugs. Infliximab was given at a dose of 5 mg/kg at weeks 0, 2 and 6 and every 8 weeks thereafter. Patients had no prior history of use of TNF antagonists or immunosuppressives. The primary outcome of the study was corticosteroid-free remission at week 26. At this time, significantly more patients in the combination arm were in remission than in the infliximab monotherapy group (56.8% vs 44.4%, p=0.02) or the azathioprine monotherapy group (30.0%, p<0.001). Similar results were obtained at 1 year. These results convincingly demonstrated the superiority of combination therapy over azathioprine or infliximab monotherapy in patients treated relatively early in the course of the disease who were previously naïve to both drugs.

The final study to consider was conducted by Feagan and colleagues who evaluated the role of combined treatment with corticosteroids, methotrexate and infliximab in patients with established active CD.44 One hundred and twenty-six patients with active CD who had initiated induction therapy with corticosteroids within the preceding 6 weeks were randomised to subcutaneous injections of methotrexate (25 mg per week) or placebo for 50 weeks of treatment. All patients received infliximab therapy, consisting of an induction regimen of three doses of 5 mg/kg followed by 8-weekly maintenance therapy. Prednisone was tapered to discontinuation by week 12. The primary outcome measure was the proportion of patients with treatment success as defined by corticosteroid-free remission at week 14 and maintenance of remission through week 50. The proportion of patients who were in remission at week 14 was 76.2% for those assigned to methotrexate and 77.8% for those who received placebo (p=0.83). Likewise, the proportion of patients with treatment success at week 50 was similar (55.6% vs 57.1%, respectively, p=0.86). In contrast to the strong efficacy signal for combination therapy identified in the SONIC trial, no benefit for combined therapy was observed in this patient population with relatively long-established disease. There are some potential explanations for these findings. First, it is possible that methotrexate is not an effective agent when used in combination with infliximab. This seems unlikely given the data supporting the role of the drug as monotherapy for the treatment of corticosteroid-dependent CD23 and the well-established benefits of methotrexate based on combination therapy in RA.36 Second, disease duration differs between these trials, suggesting that optimum efficacy for combined therapy may be in patients with early disease. A final possibility is that the relatively high prednisone doses received by participants during the induction phase of the trial had a prolonged beneficial interaction that resulted in the high rates of treatment success in both experimental groups. This potential role of adjunctive corticosteroid therapy should be explored in future studies.

Emerging data indicate that a critical aspect to optimise the efficacy of combined therapy regimens is to administer TNF antagonist according to a scheduled regimen rather than episodically.46 Furthermore, therapeutic drug monitoring and determination of anti-drug antibody status has the potential to inform clinical decision-making.47

In summary, the collective results of these trials are consistent with superior efficacy for combined immunosuppression. Although these studies have influenced current practice, long-term studies are necessary to definitively determine if early combined therapy can prevent disease progression.

Early intervention after intestinal resection

Surgery is not curative since the disease tends almost invariably to recur at the anastomotic site.48 In patients with CD the postoperative period prior to disease recurrence can be considered a ‘window of opportunity’ to modify the course of the disease by early intervention. A small (n=24) single-centre study suggested that early treatment with infliximab (within 4 weeks of surgery) might be effective at preventing histological and endoscopic recurrence after ileal resection.49 Based on these promising results, a large placebo-controlled trial (PREVENT) has been initiated to confirm these preliminary observations.50

Predictors of disease progression and disability

Population-based studies have shown that the clinical course of CD is heterogeneous and varies widely over time. The majority of patients (55%) have moderate disease activity within 5 years of diagnosis, whereas 25% experience progressive disease.51 Some clinical features associated with high risk of disease progression that have been identified in population-based cohort studies can potentially be used to identify patients who might be potential candidates for early intervention with highly effective treatment strategies designed to prevent bowel damage. An important part of the logic of this treatment strategy is that the relapse rate within the first year of diagnosis correlates positively with the relapse rate in the subsequent 5 years,52 that the cumulative relapse rate within the first 10 years of diagnosis is as high as 90%11 and that the probability of intestinal resection during the first 10 years after diagnosis ranges from 40% up to 65%.11 51

Both toxicity and cost considerations will likely preclude the use of combined immunosuppression in unselected patient populations. Validated predictors are therefore needed to identify patients who are likely to experience disease progression. Current predictors in CD are primarily clinical, in distinction to the ‘biomarkers’ that have become established as valuable prognosticators in other diseases.53–55 Involvement of the terminal ileum is strongly associated with an increased risk of stricture and internal penetrating behaviour56 57 and an increased risk of surgical resection.58 59

Retrospective studies evaluating disability in patients with CD identified initial requirement for corticosteroid use (OR 3.1, 95% CI 2.2 to 4.4), age <40 years (OR 2.1, 95% CI 1.3 to 3.6) and presence of perianal disease (OR 1.8, 95% CI 1.2 to 2.8) as independent risk factors of complicated disease with rates of disability within 5 years of diagnosis as high as 60% and 85%.60 61 The main limitation in interpreting these results is that no valid definition of disability was used. Recently, a disability index for patients with inflammatory bowel disease has been developed.62 63 This tool might be utilised in future disease-modification trials to objectively measure whether introduction of early effective treatments has the potential to modify the natural history of the disease in terms of reducing the impact that bowel damage has on disability.

In addition to clinical predictors, serological and genetic markers may identify patients at higher risk of developing disease-related complications. In a paediatric population, disease progression from non-complicated CD to more aggressive disease phenotypes (stricturing and/or penetrating) has been significantly associated with the presence and magnitude of serological immune responses to microbial antigens.64 Combining serological immune responses and genetic markers (NOD2 mutations) may also help to identify patients with high risk of disease progression.65

The severity of ulceration on ileocolonoscopy has been shown to be associated with an increased risk of bowel resection. In a retrospective cohort study (n=102) performed in France,66 patients with severe endoscopic lesions (deep ulcers) needed significantly more colonic resections than patients without ulcers (RR 5.43, 95% CI 2.64 to 11.18). The risk of colectomy in patients presenting with colonic ulcerations at 1, 3 and 8 years were 20%, 26% and 42%, respectively. These results need to be confirmed in prospective studies.

Conversely, mucosal healing has been associated with a reduced risk of hospitalisation and surgery67–69 and is a predictor of sustained corticosteroid-free remission in patients with early stage disease.39 45 Based on these data, the presence of deep ulcerations on ileocolonoscopy may prove to be both a valuable factor for risk stratification and a surrogate marker for disease modification.

Risks and costs of treatment issues

The main limitations to the expanded use of early combined immunosuppressive therapy are concerns regarding safety and the high cost of TNF antagonists (figure 3). The most important complications of immunosuppression are serious infection and lymphoma.

Figure 3

Balancing risks and benefits of different treatment strategies. TNF, tumour necrosis factor.

Serious infection

In general, infections occur more frequently in patients with CD than in the general population.70 Thus, the assessment of adverse events related to therapies must be corrected for the baseline risk inherent to the disease itself. Furthermore, patients with CD frequently receive treatment with multiple drugs and have varying degrees of disease severity. This problem of confounding risk factors makes it difficult to be confident that the attributable risk estimated from observational studies is accurate despite the use of appropriate multivariable statistical techniques. Accordingly, data from randomised controlled trials may be particularly useful for safety assessment in this situation despite traditional problems of small sample sizes and limited duration of follow-up. A meta-analysis of placebo controlled trials enrolling 5356 individuals showed that treatment with TNF antagonists did not increase the risk of serious infection or death.71 Nonetheless, these results should be interpreted with caution since the follow-up was limited to 1 year or less. It is worth mentioning that the risk of serious infection in the SONIC trial did not differ among the three treatment groups (5.6%, 4.9% and 3.9% in the azathioprine, infliximab and combination therapy groups, respectively; 95% CI −4.2% to 5.5% for azathioprine vs infliximab, 95% CI −2.9% to 6.2% for azathioprine vs combination therapy, and 95% CI −3.4% to 5.4% for infliximab vs combination therapy).40 Similar results were obtained when the much smaller top-down, COMMIT and Lémman data sets were evaluated. Turning to observational data sources, a case–control study from the Mayo Clinic evaluated the relationship between intensity of immunosuppressive therapy and the risk of opportunistic infections.72 Use of corticosteroids, azathioprine and infliximab were all associated individually with a significantly increased risk of opportunistic infection. Combined treatment with two or three of these drugs yielded an increased risk of opportunistic infection of 14.5 (95% CI 4.9% to 43%). Patients aged >50 years had three times more risk of serious infection than younger patients. However, there are a number of limitations to this study. First, only patients with ‘opportunistic infections’ were included, meaning that those with serious infections that were not opportunistic were not included. Second, opportunistic infections included infections such as oral thrush and vaginal candidiasis which are easily treated and not serious. Third, there were no control patients and no estimate of the absolute risk was provided. It should be noted that evaluation of more than 10 000 patients seen over 5 years at a major referral centre was required to identify 100 opportunistic infections, many of which were not serious.

The TREAT (Crohn's Therapy, Resource, Evaluation, and Assessment Tool Registry) registry was established to prospectively study the long-term safety of infliximab and other therapies in more than 6000 patients with CD, including 16 129 patient-years of exposure to infliximab. The TREAT registry did not find increased risks of mortality or malignancy associated with infliximab treatment. The risk of a serious infection for infliximab alone or in combination with immunosuppressives was similar, conferring a twofold increased risk in both cases.73 74 Notably, the rate of serious infection was markedly increased with the addition of corticosteroids to any treatment regimen.70 72 74 Furthermore, corticosteroids are the only medication associated with an increased risk of mortality in patients with inflammatory bowel disease.21 73 Accordingly, we speculate that a reasonable explanation for the failure of multiple studies to identify an increase in the attributable risk in patients receiving combined immunosuppression compared with patients receiving monotherapy is due to two factors. First, patients receiving more effective combination therapy are less likely to be exposed to prolonged corticosteroid therapy, resulting in a risk offset. Second, more effective therapy results in a reduced likelihood of both severe disease activity and development of disease-related complications, states that are associated with the development of infectious complications. In support of this hypothesis, a recent meta-analysis of population-based studies confirmed a slight but significantly overall increased mortality in patients with CD, with a pooled standardised mortality ratio of 1.39 (95% CI 1.30 to 1.50).75 Interestingly, after excluding deaths related to CD complications, this excess of risk was eliminated (pooled standardised mortality ratio 1.03 (95% CI 0.92 to 1.15)). Most of the studies included in this meta-analysis were performed before the era of biological agents. It is therefore conceivable that the use of less effective treatment regimens results in a higher risk of disease progression and complications that ultimately is related to an increased risk of mortality. Hence, less effective treatment may be inherently more hazardous than highly effective treatment in patients at increased risk of disease progression. Clinicians should be aware that undertreatment of patients with a poor prognosis places them at an increased risk of disease-related complications.


Overall, the risk of lymphoma in patients with CD seems to be similar to or very slightly higher than in the general population.76 The TREAT registry showed a similar incidence of lymphoma in patients treated with infliximab (0.05 per 100 patient-years) as in patients not treated with infliximab (0.05 per 100 patient-years; RR 0.92, 95% CI 0.32 to 2.63).73 These data are supported by a recent meta-analysis.71 Whether the use of TNF antagonists in combination with immunosuppressives is associated with an increased risk of non-Hodgkin's lymphoma in adult patients with CD beyond what would be expected from immunosuppressive therapy alone is unclear. Immunosuppressive therapy with azathioprine and 6-mercaptopurine is clearly associated with an increased risk of non-Hodgkin's lymphoma.77 A meta-analysis78 that incorporated 26 randomised controlled trials, case–control and cohort studies did not show an association between an increased risk of lymphoma and combination therapy with azathioprine and TNF antagonists but, given the low event rates, lack of statistical power must be considered. It is important to keep in mind that the absolute rate of lymphoma is low (6.1 per 10 000 patient-years) and should be weighed against the substantial benefits of treatment.78

Hepatosplenic T cell lymphoma (HSTCL), a rare but usually fatal lymphoproliferative disorder that primarily affects men younger than 35 years, deserves special consideration. To date, 36 cases of HSTCL have been reported in patients with inflammatory bowel disease. All patients had prior or concurrent use of thiopurines for at least 2 years, alone (n=16) or in combination with TNF antagonists (n=20). The absolute risk of HSTCL in patients receiving thiopurines in combination with TNF antagonists is very low (1:21.947).79

These toxicities (serious infection and lymphoma) should be taken into account especially when considering whether to prescribe combined therapy in specific populations such as older and young male patients.

Cost-effectiveness issues

The economic burden of CD is high due to disability, loss of work productivity,80 surgery81 and hospitalisation.82 It is therefore reasonable to speculate that highly effective treatment strategies that are able to prevent these complications and maintain durable remission might be associated with important cost offsets.

A cost-effectiveness analysis comparing conventional step-care with the early use of TNF antagonists for the treatment of RA has recently been published.83 Interestingly, early initiation of TNF antagonist therapy, which delays disease progression, was associated with fewer joint replacements and hospitalisations. In addition, when indirect costs associated with lost productivity were evaluated, the incremental costs of early treatment were further reduced, leading to an overall reduction in cost compared with conventional management. Cost-effectiveness analyses in patients with CD with high risk of disease progression are urgently needed to assess whether early initiation of combined therapy reduces the financial and social burden of the disease.

Conclusions and future directions

CD follows a progressive and destructive course under conventional therapy. Although accelerated step-care regimens are effective for controlling symptoms in corticosteroid-resistant and -dependent patients, the magnitude of the benefit is modest. A considerable body of evidence supports the notion that treatment with TNF antagonists in combination with immunosuppressives is more effective than monotherapy for treating early CD. Furthermore, limited data indicate that TNF-based regimens reduce the risk of surgery and hospitalisation. It therefore seems reasonable to use this approach for newly diagnosed patients in whom the risk of disease progression is high. However, further evidence is necessary to elucidate whether an early introduction of combined therapy has the potential to reduce disease-related bowel damage and disability and whether this approach is cost-effective.

From a safety perspective, accurate selection of patients at high risk of developing disease-related complications is essential to achieve better outcomes and to avoid overtreatment of patients with a benign course. In this regard, development of clinical prediction rules that combine clinical, serological, genetic and imaging markers as predictors of disease progression is a research priority.

Clinicians should recognise that, while the risk of serious infection due to combined immunosuppression exists, other factors must be weighed. Patients with inadequately controlled disease activity, who are exposed to prolonged courses of corticosteroid therapy, are at high risk from both disease-related complications and serious infection. Combined immunosuppression early in the course of the disease provides the patients with the best opportunity of enjoying favourable outcomes.

The use of new instruments that objectively measure both structural and functional bowel damage will be of great value in future clinical trials. Similar to the evolution that has occurred in the treatment of RA, it is time to move to more objective outcomes such as radiological and bowel healing in both clinical trials and clinical practice in patients with CD. Prospective trials to evaluate the benefit of early intervention in patients with CD who are at high risk of disease progression are urgently needed. For patients at low risk, data regarding the best treatment strategy are currently unclear. In this subset of patients, clinicians should employ a strategy of treating to symptomatic remission with careful monitoring.


  • Effective interventions should be initiated before bowel damage becomes irreversible.

  • Clinical parameters can be used to predict an unfavourable disease course and thus promptly identify which patients are at higher risk of disease progression.

  • Patients with disease duration ≤2 years are most likely to benefit from TNF antagonists.

  • Combined treatment with infliximab and azathioprine is more effective than either drug alone in patients with early disease.

  • Benefits of combined treatment need to be balanced against the risk of serious infection and lymphoma.


View Abstract


  • Competing interests WJS has received consulting fees from Abbott Laboratories, ActoGeniX NV, AGI Therapeutics, Alba Therapeutics Corporation, Albireo, Alfa Wasserman, Amgen, AM-Pharma BV, Anaphore, Astellas Pharma, Athersys, Atlantic Healthcare, Axcan Pharma (now Aptalis), BioBalance Corporation, Boehringer-Ingelheim, Bristol Meyers Squibb, Celegene, Celek Pharmaceuticals, Cellerix SL, Cerimon Pharmaceuticals, ChemoCentryx, CoMentis, Cosmo Technologies, Coronado Biosciences, Cytokine Pharmasciences, Eagle Pharmaceuticals, Eisai Medical Research, Elan Pharmaceuticals, EnGene, Eli Lilly, Enteromedics, Exagen Diagnostics, Ferring Pharmaceuticals, Flexion Therapeutics, Funxional Therapeutics, Genzyme Corporation, Genentech (now Roche), Gilead Sciences, Given Imaging, GlaxoSmithKline, Human Genome Sciences, Ironwood Pharmaceuticals (previously Microbia), Janssen (previously Centocor), KaloBios Pharmaceuticals, Lexicon Pharmaceuticals, Lycera Corporation, Meda Pharmaceuticals (previously Alaven Pharmaceuticals), Merck Research Laboratories, MerckSerono, Millennium Pharmaceuticals (subsequently merged with Takeda), Nisshin Kyorin Pharmaceuticals, Novo Nordisk A/S, NPS Pharmaceuticals, Optimer Pharmaceuticals, Orexigen Therapeutics, PDL Biopharma, Pfizer, Procter and Gamble, Prometheus Laboratories, ProtAb, Purgenesis Technologies, Receptos, Relypsa, Salient Pharmaceuticals, Salix Pharmaceuticals, Santarus, Schering Plough Corporation (acquired by Merck), Shire Pharmaceuticals, Sigmoid Pharma, Sirtris Pharmaceuticals (a GSK company), S.L.A. Pharma (UK), Targacept, Teva Pharmaceuticals, Therakos, Tillotts Pharma AG (acquired by Zeria Pharmaceutical), TxCell SA, UCB Pharma, Viamet Pharmaceuticals, Vascular Biogenics Limited (VBL), Warner Chilcott UK Limited, Wyeth (now Pfizer), lecture fees from Abbott Laboratories, Bristol Meyers Squibb and Janssen (previously Centocor) and research support from Abbott Laboratories, Bristol Meyers Squibb, Genentech, Glaxo Smith Kline, Janssen (previously Centocor), Millennium Pharmaceuticals (now Takeda), Novartis, Pfizer, Procter and Gamble Pharmaceuticals, Shire Pharmaceuticals and UCB Pharma. BGF has received research support, consulting and lecture fees from Janssen (previously Centocor), Merck (previously Schering Plough), Abbott Laboratories and UCB Pharma. IO (author name, Ingrid Ordas) has no competing interests.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.