Results of the 2nd scientific workshop of the ECCO (IV): Therapeutic strategies to enhance intestinal healing in inflammatory bowel disease

Abstract

Evidence supporting the importance of assessment of mucosal healing in inflammatory bowel disease has increased in the last years. Mucosal healing has been integrated in the assessment of treatment efficacy in ulcerative colitis, but in Crohn's disease this thought has arised after biological agents have been evaluated in clinical trials. Although a validated definition of mucosal healing still does not exist, its use is also assuming an increasingly important role in the follow-up of individual patients in clinical practice. Corticosteroids induce mucosal healing in a small proportion of patients with Crohn's disease and are of no benefit to maintain it. By contrast, mucosal healing in Crohn's disease can be achieved and maintained, with varying degrees of evidence and success, with thiopurines and biological agents. In ulcerative colitis, the ability of corticosteroids to induce mucosal healing is well recognized. 5-aminosalicylates, thiopurines and biological agents are also able to induce mucosal healing and, additionally, to maintain it. Mucosal healing assessment should be considered in clinical practice when symptoms persist despite therapy or when treatment discontinuation is being considered. Conversely, in patients whose clinical remission is not associated with mucosal healing, intensification of treatment is not currently recommended because of lack of evidence.

Crohn's disease, Ulcerative colitis, Inflammatory bowel disease, Mucosa healing, Treatment

1 Introduction

Crohn's disease and ulcerative colitis, the two major forms of idiopathic inflammatory bowel disease, are characterized by the presence of inflammatory changes in the gut associated with ulcerations of various sizes and depths. Crohn's disease is frequently associated with complications, and surgery is necessary in up to 70% of patients. Postoperative recurrence is common, however, and repeated cycles of surgery-relapse can lead to progressive and disabling loss of intestinal function.¹ As for ulcerative colitis, total proctocolectomy is necessary in some patients with poorly controlled disease.

The management of inflammatory bowel disease has traditionally been aimed at improving the patient's symptoms, but relatively little attention has been focused on the longer-term impact of this endpoint or on its relation to endoscopically ascertained resolution of the mucosal lesions.

Thus, the scientific committee of ECCO has launched a scientific workshop in 2010 that focused on this significant clinical research question. The overall objective of this workshop was to better understand and explore the importance of mucosal healing in inflammatory bowel disease. The outcome of this workshop is presented into four parts: Mechanisms of Intestinal Healing (Basic science), Measures and Markers of Prediction to achieve, detect, and monitor Intestinal Healing, Impact of Intestinal Healing on the Course of IBD (Natural history), and Therapeutic Strategies to enhance Intestinal Healing (Therapy). This manuscript summarizes current knowledge about therapeutic strategies to enhance Intestinal Healing and highlights several key issues that need to be addressed in future studies.

A host of indirect evidence indicates that the presence or absence of mucosal healing is prognostically relevant. In a Norwegian population-based cohort, mucosal healing was found to have a significant influence on the outcome of inflammatory bowel disease.² In this study, 458 patients with ulcerative colitis or Crohn's disease were assessed clinically and endoscopically at baseline/the time of diagnosis and 1 year and 5 years after diagnosis/later. The finding of mucosal healing at the 1 year examination was associated with reduced rates of surgery during long-term follow-up. Results were confirmed during the 10-year follow-up of the same cohort, thus highlight the potential benefits of mucosal healing on surgical intervention rates.³ In Crohn's disease, mucosal healing has been associated with longer-lasting corticosteroid-free clinical remission,⁴ longer times to clinical relapse,⁵ and reduced need for surgical intervention and hospitalization.^6,7 The concept of “deep remission” is a recently introduced end point, which includes corticosteroid-free clinical remission and mucosal healing. It has been introduced and applied to patients with Crohn's disease on biologics/immunomodulators who have no symptoms and objective signs of inflammation.⁸

Also, in ulcerative colitis, mucosal healing has been associated with a reduced risk of relapse^9–11 and lower colectomy rates,^12,13 but its importance does not end here. In children and adolescents, it can often restore normal growth, and in patients of all ages it may be associated with reduced loss of important nutrients such as iron. Finally, in theory at least it is thought to have potentially positive effects on the risk of cancer.

For these reasons, more and more clinicians are now asking whether mucosal healing should be a therapeutic objective in patients with IBD, and if so, how this objective can be achieved. This section will explore some of the fundamental issues related to mucosal healing that have to be resolved—beginning with the definition of the term itself. We will also look at the therapeutic strategies most likely to achieve and maintain this goal and its possible impact on therapeutic decision making.

2 How should mucosal healing be assessed and when should it be measured?

Mucosal healing is currently assessed on the basis of the macroscopic appearance of the mucosa at ileocolonoscopy. The term healing in this context does not imply the complete absence of active inflammation. The term mucosal healing falls within the boundaries of a broader concept, that of intestinal healing, which also includes histological, transmural, and fistula healing. Endoscopically ascertained mucosal healing is considered a surrogate efficacy end point in clinical drug trials, and its use in clinical practice is also increasing. And yet we still do not have a validated definition of what this term means in patients with inflammatory bowel disease.

The endoscopic features of mucosal inflammation in a patient with Crohn's disease include erythema, swelling, nodularity, aphthoid ulcerations, ulcers of variable size and depth, and strictures. In ulcerative colitis the inflamed mucosa presents at endoscopy with vascular congestion, erythema, edema, granularity, friability, spontaneous bleeding, and macroscopic ulcerations of variable size and depth. In light of the complexity of these presentations, it is clear that any meaningful endoscopic assessment of mucosal lesions in inflammatory bowel disease and their responses to therapeutic interventions must be based on validated endoscopic indices of disease activity.^14,15

The endoscopic indices used thus far to grade mucosal lesions in Crohn's disease and in ulcerative colitis have been extensively discussed in another part (part B) of this Workshop. In particular, the Crohn's Disease Endoscopic Index of Severity (CDEIS)¹⁶ and the Simplified Endoscopic Activity Score for Crohn's Disease (SES-CD)¹⁷ both have their own strengths and weaknesses. They present lack of agreement about cut-off values for endoscopic response/remission, and do not provide a reliable estimate of the extension of the disease. Furthermore, one of the main limitations of the Rutgeerts score¹⁸ in the postoperative setting is the definition of grade 2 recurrence, which is much more subjective than the definitions of other grades. As a matter of fact, the definition of mucosal healing usually employed in clinical trials and in clinical practice is the disappearance of mucosal ulcerations. This “black-or-white” approach has a number of shortcomings, including a clear classification bias toward one specific type of lesion. It leaves no space to for the concept of mucosal (partial) improvement. The endoscopic disease activity indices developed for use in ulcerative colitis classify endoscopic lesions as normal/near normal, improved, no change or worse; assess the severity of bleeding without any consideration of mucosal ulcers (e.g., Baron score)¹⁹; or variably assess the pattern of mucosal lesions (e.g., Mayo score).²⁰ The Mayo endoscopic score is used frequently, but distinguishes only four grades of severity, and mild friability (score 1) is classified as a sign of active disease rather than of mucosal healing. As noted above, each of these indices has its own strengths and weaknesses, but all are subject to interobserver variability, and none has been fully validated with the notable exception of the recently presented UCEIS (ulcerative colitis endoscopic index of severity).²¹

Enrolment criteria have varied widely in the clinical trials that have included mucosal healing as a marker of efficacy in the assessments of conventional treatments for Crohn's disease. Study populations considered have entered the specific trial as active or quiescent disease, with “colonic or ileocolonic involvement” and “presence of severe lesions”, or “mucosal ulcerations”, or “mucosal ulcerations in at least 2 intestinal segments and a CDEIS of more than 8” or “active endoscopic inflammation”. Postsurgical populations have been enrolled with “recurrent pre-anastomotic ileal or colonic disease” or “severe refractory recurrence,” and the endoscopic findings have been reported as scores or with simple descriptive terms. As for mucosal healing, it has been assessed at various time points, and its definition is generally descriptive (e.g., normal appearing mucosa—only scars, disappearance of all endoscopic lesions, disappearance of any pathological findings, complete absence of mucosal ulcerations, no ulcers) although in some cases it is not defined at all.^14,22 From a clinician's point of view, this heterogeneity makes it difficult to compare among studies and/or medications and should be made more uniform.

As for ulcerative colitis, mucosal healing has generally been regarded as an important treatment goal, and endoscopic remission has been used as end point in trials focusing the induction or maintenance of remission. The entry criteria adopted in these studies vary from active symptoms, with or without endoscopic confirmation, disease activity indices or composite clinical and endoscopic disease activity indices. In 2007 an expert consensus statement recommended that complete mucosal healing (defined as “absence of friability, blood, erosions and ulcers in all the visualized segments”) should be incorporated into the primary end point of all trials on ulcerative colitis.¹⁵ Among the few studies that have included an end point of this type, the definitions have varied, from descriptive definitions (e.g., “rectal mucosa repaired with appearance of vascular pattern” or “normal sigmoidoscopic appearance”) to those based on endoscopic activity index scores.²³

All of the IBD activity indices used in recent clinical trials contain several subjective parameters, which may be responsible for some of the high placebo response rates seen in several of these trials. Therefore, among other biologic outcome parameters, endoscopic healing has been advocated in the evaluation of novel therapeutic agents in inflammatory bowel disease.

On the basis of currently available data, the optimal timing of assessments for mucosal healing is not clear. Randomized controlled trials on anti-TNF alpha in active Crohn's disease have evaluated healing after different periods of treatment (e.g., 4 weeks, 12 weeks, 26 weeks) in different patient populations,^24–26 and similar variability is seen in UC trials (e.g., 6 weeks, 8 weeks, 16 weeks).^10,27–30

Therefore, it seems premature to advocate routine endoscopic follow-up of all patients with Crohn's disease who achieve clinical remission with medical therapy, but in patients presenting with persistent symptoms despite objective clinical response, endoscopic assessment might be useful to rule out post-inflammatory irritable bowel syndrome as a cause of diarrhea and abdominal pain. Persistent or recurrent symptoms suggestive of irritable bowel syndrome have been reported in a substantial percentage of IBD patients with sustained clinical and endoscopic remission.³¹ In addition, ileocolonoscopy is recommended during the first year after surgery, when treatment decisions are tailored to the severity of recurrent endoscopic lesions, although there is no strong evidence that this strategy reduces the incidence/severity of recurrence. Finally, mucosal healing is a predictor for sustained clinical remission in patients who are stopping biologics but continue immunosuppressives (thiopurines).³²

In patients with ulcerative colitis, flexible sigmoidoscopy is a reliable and generally well tolerated means for assessing disease severity. Therefore, periodic endoscopy to evaluate treatment effectiveness is an option that should be discussed with all patients with ulcerative colitis patients.

In conclusion, each of the methods used in clinical trials to measure the endoscopic activity of Crohn's disease has its own intrinsic limitations, and lack validated cut-offs, thus reflecting the difficulties in arriving at a quantitative measure of “healing”. In most of clinical trials on ulcerative colitis, disease activity has been rated with composite clinical and endoscopic scores. The accuracy of these indices is limited by the wide spectrum of symptoms in patients with distal or extensive UC, by overlaps in the endoscopic findings associated with quiescent and mild-to-moderate disease, and by the absence of formal validation. Therefore, no robust recommendations for routine clinical practice can be made on the basis of available evidence.

As noted earlier, the assessment of mucosal healing currently requires endoscopy, which is expensive, time-consuming, and unpleasant for the patient, but new methods are being developed for this purpose that should overcome many if not all of these shortcomings, as discussed in part B of this Workshop.

• Mucosal healing is assessed by ileocolonoscopy or colonoscopy according to the macroscopic appearance of the mucosa. In light of the heterogeneity and the complexity of mucosal lesions endoscopic indices of severity have been developed.

• In Crohn's disease endoscopic indices present no validated cut-off value for mucosal healing and the current definition of mucosal healing employed in clinical trials and in clinical practice is “disappearance of mucosal ulcerations”.

• Mucosal healing has been incorporated in the assessment of treatment efficacy in ulcerative colitis. Current evidence does not allow to clarify optimal timing of mucosal healing assessment in inflammatory bowel disease.

3 What drugs are most likely to produce mucosal healing?

3.1 Crohn's disease

Clinical trials in the field of Crohn's disease have traditionally evaluated treatment efficacy in terms of clinical symptoms rated with validated scoring systems. Well-defined cut-offs for relapse and remission have been established for the most widely used scoring systems, such as the Crohn's disease activity index (CDAI), which has also been adapted for use in pediatric patients, and the Harvey Bradshaw index (HBI).¹⁴ Until recently, little emphasis has been placed on mucosal healing as a treatment goal. It was regarded as unimportant and nearly impossible to achieve with the treatments available until the 1990s. In the early 1990s, endoscopic evidence of mucosal healing was investigated in a study designed to validate the CDEIS. Less than one third of the patients treated for 7 weeks with high-dose corticosteroids displayed healing, and in almost 10% of the patients the lesions actually worsened.³³

At that time, the purine analogs were the only drugs known to alter the course of Crohn's disease, diminishing the need for steroids and promoting long-standing remission and fistula healing.³⁴ But at the end of the 1990s, therapy with the azathioprine was also shown to be associated with endoscopic evidence of mucosal healing in patients with postoperative recurrence of Crohn's ileitis and those with primary Crohn's ileocolitis.^35,36 At this point, interest began to grow in the concept of endoscopic healing, and its impact on the course of the disease became increasingly clear. Today–largely as a result of the introduction of biological therapies for inflammatory bowel disease–mucosal healing is regarded by many as an important measure of treatment efficacy and an obligatory end point for studies aimed at obtaining regulatory approval of treatments. It has been included as a secondary or even primary endpoint in clinical trials with biological agents, and it is also assuming an increasingly important role in the follow-up of individual patients in clinical practice.

3.1.1 The mucosal healing capacities of drug therapies in Crohn's disease

3.1.1.1 Corticosteroids

Although steroids are known to induce acute clinical response and even remission in patients with Crohn's disease, they do not seem to have any beneficial effects on mucosal lesions or the likelihood of postoperative recurrence of endoscopic lesions. Studies conducted by the Groupe d'Etudes Thérapeutiques sur les Affections Inflammatoires Digestives (GETAID) showed that oral prednisolone (1 mg/kg per day) produced clinical remissions in 92% of the patients within 7 weeks, but only 29% of these patients achieved endoscopic remission. Moreover, the clinical severity of the disease displayed no correlation with the severity or extension (involved surface area) of endoscopic lesions.^33,37 In a small study in Sweden, glucocorticoids displayed high efficacy in attenuating the symptoms of small-bowel Crohn's disease, but once again this improvement was not accompanied by any significant improvement in the small intestinal lesions observed at endoscopy.³⁸ More recently, Mantzaris et al. studied 77 patients with steroid-dependent forms of Crohn's ileocolitis or proximal colitis who had achieved clinical remission on conventional steroid therapy. After 1 year, only one out of four of the 39 patients treated with budesonide (6–9 mg per day) presented complete or near complete healing on endoscopy.³⁹

3.1.1.2 Thiopurines (azathioprine, 6-mercaptopurine)

Azathioprine's ability to induce mucosal healing in Crohn's disease was first demonstrated in 1997. D'Haens et al. studied 19 patients who had undergone ileocaecal resections for their Crohn's disease, which were followed by severe recurrent ileitis. Azathioprine therapy was maintained for at least 6 months after the patients had been completely weaned off corticosteroids, and 15 of the patients were then reassessed endoscopically or radiologically. The immunomodulator therapy produced sustained clinical remission in all 15 cases, and 6 (40%) exhibited endoscopic evidence of complete mucosal healing in the neoterminal ileum. Eight others displayed near-complete healing with only superficial erosions (n = 5) or partial healing (n = 3), and only 1 had no improvement at all. On the basis of these findings, the authors suggested that azathioprine should be the treatment of choice for severe recurrent ileitis in patients with Crohn's disease.³⁵

In another study, the same investigators endoscopically evaluated 20 patients with Crohn's colitis or ileocolitis in clinical remission achieved while they were taking azathioprine (for at least 9 months). None of the patients had taken corticosteroids for at least 3 months before the endoscopy. Complete to near-complete healing of the mucosa was observed in the large bowel of 80% of the patients and at the ileal level in up to 69%.³⁶

The GETAID conducted a randomized, double-blind study to identify the effects of azathioprine withdrawal on sustained remission. Baseline ileocolonoscopy was performed in 45 (54%) of the 83 patients. These patients had been in clinical remission on azathioprine for at least 42 months. Complete endoscopic remission defined as CDEIS = 0 was seen in 36% of the patients, although the mean CDEIS level was rather low at baseline (mean = 2.5).³⁹ In another study, sustained mucosal healing was investigated in patients who had achieved clinical remission with standard corticosteroid therapy. After 1 year of azathioprine therapy (2.0–2.5 mg/kg), 83% of the patients presented evidence of complete or near-complete healing, with a decrease in the mean CDEIS from 7 to 0.55. In a multivariate logistic regression analysis, initiation of AZA within 1 year of the Crohn's disease diagnosis was the only factor that predicted complete endoscopic healing.⁴⁰In the step-up top down trial, on the contrary, mucosal healing was observed in less than a third of patients who were initially treated with steroids and maintained with azathioprine.⁴¹ The reason for this discrepancy is not clear. Reinisch et al. performed a prospective trial in patients with postoperative recurrence of Crohn's disease comparing aminosalicylates with azathioprine.⁴² An improvement in the Rutgeerts score was observed in 63% of the patients on azathioprine versus in 34% on aminosalicylates.

3.1.1.3 Methotrexate

In the late 1980s, Kozarek et al. conducted a pilot study on the effects of methotrexate in 14 patients with Crohn's colitis, and after 12 weeks of treatment 5 (36%) of the patients had colonoscopy-documented mucosal healing.⁴³ Later studies confirmed that methotrexate was an effective drug for the induction and maintenance of clinical remission.^44–45 A single study assessed the effect of methotrexate on mucosal healing in patients clinically responding to the treatment.⁴⁶ This observational study suggested that methotrexate was inferior to infliximab (p = 0.008) or azathioprine (p = 0.011) in inducing mucosal healing.

3.1.1.4 Anti-TNF antibodies

The first anti-TNF antibody, infliximab, has been evaluated in a number of multicenter studies. The effects on mucosal healing were assessed in substudies.

In 1999, D'Haens et al. reported the disappearance of ulcers in the ileum in 74% of patients who had received a single dose of infliximab 4 weeks before endoscopy, and the rate of mucosal healing in the rectum was even higher (96%). In the same study, decreases in the CDEIS were significantly correlated with drops in the CDAI.⁴⁷ In a substudy of the ACCENT I project, episodic and scheduled-treatment infliximab regimens were compared in patients with Crohn's disease who had a clinical response to a single infusion of the biological agent (5 mg/kg).⁴⁸ Patients randomized to the scheduled-therapy regimen had fewer hospitalizations and surgeries and higher rates of mucosal healing. At week 10, complete mucosal healing was seen in 31% of the patients in this group (versus 0% of those on episodic treatment), and the difference was even more substantial at week 54 (mucosal healing rates of 50% and 7%, respectively). In addition, the mucosal healing rates were directly associated with the number of hospitalizations, and no hospitalizations at all were required in cases characterized by endoscopic remission at weeks 10 and 54. Nonetheless, this study found no consistent relationship between mucosal healing and clinical remission.⁶ Scheduled therapy also proved to be superior to the episodic approach in a retrospective study of 214 patients with Crohn's disease, in which the initiation of infliximab was associated with mucosal healing in 67.8% of responders and complete mucosal healing in 45% of all patients. The likelihood of mucosal healing did not appear to be related to the concomitant use of immunomodulators.⁷

In the SONIC trial, patients with active Crohn's disease who had never been exposed to biological or immunomodulatory treatment were randomized to azathioprine monotherapy, infliximab monotherapy, or the two drugs combined. Steroid tapering was obligatory after week 14. Endoscopy was performed on all participants at the time of enrolment and at week 26, the time point of the primary analysis. Mucosal healing defined as the absence of ulcers was found in 44% of the patients on combination treatment, 30% receiving infliximab monotherapy, and 16% of those on azathioprine monotherapy. Statistically, the combo regimen was not superior to the infliximab monotherapy group despite a strong trend.²⁵ The mucosal healing effects of other anti-TNF antibodies, adalimumab and certolizumab pegol, have also been assessed. In the EXTEND trial 129 patients with active Crohn's disease were placed on induction treatment with adalimumab 160 mg followed by 80 mg at week 2 and randomized to maintenance treatment with 40 mg adalimumab or placebo every 2 weeks. Endoscopic assessments were performed at baseline and at weeks 12 and 52. At week 12, the complete mucosal healing rate in the group on adalimumab maintenance therapy was roughly twice as high as that observed in the placebo maintenance group (27% versus 13%, respectively; p = 0.056). At week 52, 24% of adalimumab-treated patients still presented complete healing but this finding was not observed in any of the patients on placebo maintenance (p < 0.001).²⁶ In the MUSIC study, all 89 patients received open-label induction with certolizumab pegol (400 mg every 2 weeks for 3 doses followed by 400 mg every 4 weeks), and endoscopies were performed at baseline and at weeks 10 and 56. By week 10, the mean CDEIS score was already significantly reduced (mean decrease: 42%, p < 0.0001 versus baseline). Well over half (61%) had score decreases of >5 points. However, only in 4 patients all lesions had disappeared.⁴⁹ Of note in the MUSIC trial CDEIS criteria were used whereas in the ACCENT and EXTEND trial, mucosal healing was judged by complete ulcer disappearance.

3.1.1.5 Antibodies to cell adhesion molecules

Another biologic approach to the management of inflammatory bowel diseases involves the suppression of leukocyte migration into inflamed intestinal tissues by blocking cellular adhesion molecules. Natalizumab, a humanized monoclonal antibody to α-4 integrin, has been shown to be effective for the treatment of Crohn's disease,⁵⁰ and it also induces mucosal healing. In a small substudy of the ENACT-1 trial, 8 (22%) of the 37 patients with mucosal ulceration at study entry achieved complete mucosal healing at week 10 compared with only 1 (8%) of the 13 in the placebo group.⁵¹

3.1.1.6 Enteral nutrition

A primary therapy in children with Crohn's disease involves the use of an oral polymeric diet as the sole source of nutrition, and this approach has shown some efficacy in achieving mucosal healing.^52,53 In a randomized trial that included 37 children with active Crohn's disease, the rate of mucosal healing in patients treated with the polymeric diet was significantly higher than that achieved with corticosteroid therapy (74% versus 33%; p < 0.05).⁵⁴

• Mucosal healing in Crohn's disease can be achieved, with varying degrees of success and evidence, by means of corticosteroids, enteral nutrition (pediatrics), thiopurines, methotrexate and biological agents (infliximab, adalimumab, certolizumab pegol, natalizumab).

• Mucosal healing in Crohn's disease can be maintained with thiopurines and biological agents (infliximab, adalimumab).

3.2 Ulcerative colitis

In ulcerative colitis the inflammatory lesions are confined to the mucosa, the most superficial layer of the colon, and the disease almost always involves the distal colon, which is easier to examine with endoscopy. In the early 1990s, the results of a trial with 5-aminosalicylic acid (5-ASA) therapy clearly showed that remissions documented by clinical and endoscopic findings were likely to be longer-lasting than those that were exclusively clinical.⁵⁵ During the same period, another study demonstrated the positive impact of mucosal healing on ulcerative colitis relapse rates. Flares during the 12-month period following an episode of active colitis were observed in only 4% of the patients with clinical remission and mucosal healing and in 30% of those whose clinical remission was accompanied by persistent mucosal lesions.⁵⁶ Analogous pictures have emerged from a more recent Italian study, where relapse rates at 1 year were 23% in patients with clinical and endoscopic remission and 80% in those who still had endoscopic lesions,¹⁰ and from the ACT 1 and 2 trials. In the latter studies, rapid induction of mucosal healing with infliximab resulted in a four-fold increase in clinical remission rates at week 30.²⁸ In light of the documented impact of mucosal healing on the course of ulcerative colitis, endoscopic assessments have been incorporated in the evaluation of drugs for ulcerative colitis.¹⁵

3.2.1 The mucosal healing capacities of drug therapies in ulcerative colitis

3.2.1.1 5-Aminosalicylic acid

A randomized study reported in 1987 evaluated the efficacy of oral 5-ASA therapy (daily doses of 4.8 or 1.6 g) in patients with mildly to moderately active ulcerative colitis. After 6 weeks of treatment, flexible proctosigmoidoscopy showed complete and partial response rates of 24% and 50%, respectively, in patients receiving the higher dose of 5-ASA, whereas corresponding rates in the placebo group were only 5% and 13%.¹⁹ Sandborn et al. analyzed the results of two recent studies with Multi Matrix system (MMX) mesalazine and found that rates of complete mucosal healing (defined as a sigmoidoscopy score of 0) in patients treated with up to 8 weeks of MMX mesalazine were roughly twice as high as those observed in the placebo group (32% versus 16%, respectively).²⁷ Similar mucosal healing effects have been documented when 5-ASA is administered topically: enemas containing 4 g 5-ASA and 100 mg hydrocortisone were found to produce endoscopic healing rates of 93% and 54%, respectively.⁵⁷

3.2.1.2 Corticosteroids

The efficacy of steroids in controlling ulcerative colitis was first documented in the 1950s by Truelove and Witts. They reported mucosal healing rates of 52% in patients who received high-dose steroid therapy (cortisone 100 mg a day) for 6 weeks compared with only 32% of those treated for the same period with placebo.⁵⁸

3.2.1.3 Azathioprine

Compared with 5-ASA, azathioprine is significantly more effective in inducing clinical and endoscopic remission in corticosteroid-dependent ulcerative colitis. This outcome was observed in over half of the AZA-treated patients studied by Ardizzone et al., a therapeutic gain of roughly 35% with respect to 5-ASA.⁵⁹ The efficacy of azathioprine was also documented in a study of 42 patients with corticosteroid-dependent or corticosteroid-resistant ulcerative colitis. Complete endoscopic remission was seen in 22 of 32 (69%) patients tolerating azathioprine after 6 months. Ten patients were unable to tolerate azathioprine, and 6 of these achieved complete mucosal healing with methotrexate.⁶⁰

3.2.1.4 Anti-TNF antibodies

The effects of infliximab in active ulcerative colitis have been extensively investigated. In the large controlled ACT-1 and ACT-2 trials, all participants were examined endoscopically at enrolment and again at weeks 8 and 30. Fifty nine to sixty per cent of patients healed their mucosa on infliximab therapy at week 8, versus 32% on placebo (p < 0.001). It should be noted that the trial definition of mucosal healing was not “the absence of all lesions” but rather a Mayo endoscopic subscore of 0–1, which means that healing sometimes included the persistence of milder lesions. Using this definition, healing rates at week 30 were 48%–53% in the infliximab groups versus 27% in patients treated with placebo (p < 0.001), but complete endoscopic healing rates (i.e., a Mayo score of 0) were somewhat lower (25% and 28%–31% at weeks 8 and 30 vs. 8% and 9%, respectively), in the placebo group. Patients who achieved complete mucosal healing also had better outcomes: those who were healed at week 8 were 4 times more likely to remain in remission until week 30 (43.8% vs. 9.5% in those with persistent lesions at week 8). Most patients had repeat endoscopic investigations at week 54 during the follow-up phase of this study, and maintenance of healing was clearly demonstrated during continued infliximab treatment.²⁸

In the recently reported UC Success trial, which lasted 16 weeks, three different treatment regimens–azathioprine, infliximab, or infliximab plus azathioprine–were compared in patients with moderate–severe ulcerative colitis. None of the patients had been treated with biologic agents; some had already been treated with azathioprine, but it had been discontinued at least 3 months before enrolment. Rates of mucosal healing in the patients who received infliximab (55% with infliximab alone, 63% with infliximab plus azathioprine) were significantly higher than that observed in the azathioprine arm (37%; p < 0.05).³⁰

A very recent placebo-controlled study investigated the efficacy of adalimumab to induce clinical remission in moderately–severely active ulcerative colitis. Participants were examined endoscopically at enrolment and after 8 weeks of therapy. Mucosal healing (defined as a Mayo endoscopic subscore of 0–1) was observed in 46.9% of those treated with adalimumab 160/80 mg and 41.5% of those who received placebo.²⁹ The 52-week study assessed the efficacy of adalimumab for the induction and maintenance of clinical remission in active ulcerative colitis. Participants were examined endoscopically at enrolment and again at weeks 8 and 52. Compared with placebo-treated patients, those who received adalimumab had significantly higher rates of mucosal healing (Mayo endoscopic subscore of 0–1) at week 8 (41.1% versus 31.7%; p = 0.03) and at week 52 (25% versus 15.4%; p < 0.01). Sustained mucosal healing at both weeks 8 and 52 was also significantly higher in the adalimumab group (18.5% versus 10.6%; p = 0.01).⁶¹

3.2.1.5 Antibodies to cell adhesion molecules

In a placebo-controlled trial of the anti-α4β7 integrin antibody (MLN-02) in patients with active UC, Feagan et al. found that 28% of patients who received 0.5 mg/kg of MLN02 were in endoscopic remission by week 6, more than twice the rate (12%) observed among patients who received MLN-02 at a dose of 2.0 mg per kilogram and over three times higher than that (8%) observed in the placebo group.⁶²

• Mucosal healing in ulcerative colitis can be achieved, with varying degrees of success and evidence, by means of 5-aminosalicylates (oral and topical), corticosteroids (oral and topical), thiopurines, methotrexate and biological agents (infliximab, adalimumab, MLN-02).

• Mucosal healing in ulcerative colitis can be maintained with 5-aminosalicylates, thiopurines and biological agents (infliximab, adalimumab).

4 When to start treatment to achieve mucosal healing?

The natural history of inflammatory bowel disease is characterized by progressive complications. Therefore, the timing of initiation of therapy is of crucial importance in slowing down this process, taking into consideration that a longer disease duration inevitably leads to more irreversible damage in spite of medical treatment.

Indirect proof of this principle comes from comparison of studies on children and adults with inflammatory bowel disease. Earlier initiation of therapy is associated with higher remission rates in pediatric Crohn's disease than it is in adult patients, as shown by studies assessing thiopurines^63,64 and TNF alpha blockers.^65–68

In the “step-up/top-down” trial reported by D'Haens et al.,⁴¹ 129 patients with active, early-onset Crohn's disease were randomly allocated to classic treatment (i.e., a “step-up” approach consisting of remission induction with corticosteroids followed by readministration of corticosteroids with azathioprine for new flares and possible treatment with infliximab if active disease develops under immunosuppressant therapy) or a more aggressive regimen “top-down” approach in which combined-drug immunosuppression (infliximab plus azathioprine) was used from the outset to induce remission. Two years after randomization, complete mucosal healing (defined as complete disappearance of ulcers) was observed in 73% of the patients in the top-down group compared with only 30% of those in the step-up treatment arm.

The effects of disease duration on deep remission in Crohn's disease have been recently reported. In a subanalysis conducted during the EXTEND trial,⁶⁹ deep remission rates after 52 weeks of adalimumab maintenance therapy were significantly higher than those observed with placebo maintenance, with a trend toward relationship between deep remission rates and disease duration in the adalimumab arm.

Similar findings have been reported in other chronic inflammatory diseases, such as rheumatoid arthritis. The BeSt study, for example, showed that aggressive treatment of early-stage rheumatoid arthritis can slow disease progression. In this multicenter trial, 508 patients were randomized to disease-modifying anti-rheumatic drug monotherapy (group 1), step-up combination therapy (group 2), initial management with combination therapy that included tapered high-dose prednisone (group 3), or initial management with combination therapy that included infliximab (group 4). After 1 year of treatment, radiographic damage in the latter two groups was less severe than it was in patients managed with monotherapy or step-up combination regimens,⁷⁰ and this advantage persisted over the next 2 years.⁷¹

• Combined immunosuppression is associated with higher rates of mucosal healing than conventional therapy in early-onset adult Crohn's disease.

• An indirect relationship between higher rates of mucosal healing and Crohn's disease duration comes only from subgroup analyses data. This point has not been formally investigated in ulcerative colitis.

Can treatment step-up improve the chances of mucosal healing and what strategies should be recommended after mucosal healing has been achieved?

There is no evidence that treatment intensification leads to higher rates of mucosal healing in patients with Crohn's disease. In those with ulcerative colitis, however, the addition of enema therapy to the oral drug regimen may improve healing in the distal colon. A growing body of evidence supports the fact that combined oral and rectal 5-ASA therapy in patients with left-sided UC reduces the risk of relapse compared with monotherapy.⁷² A combination of oral and rectally administered mesalazine for the treatment of extensive mildly to moderately active ulcerative colitis has resulted superior to oral mesalazine alone also for the treatment of extensive mildly to moderately active ulcerative colitis.⁷³ In a randomized, double-blind study, remission rates at 8 weeks were 64% among patients treated with oral mesalazine (4 g/day) plus mesalazine enemas (1 g) compared with only 43% among those taking oral mesalazine alone (p = 0.03).

There are no prospective controlled data showing that treatment intensification improves outcomes in patients with persistent mucosal lesions in spite of sustained clinical remission. As far as corticosteroids are concerned, available data indicate that they do not have any effect in this setting. The GETAID group reported that in patients with Crohn's disease prolonging prednisolone therapy (1 mg/kg body weight/day) until the colonic mucosa was healed did not improve the outcomes observed after the steroids were tapered and finally discontinued.³⁷ Moreover, relapse-free intervals were similar patients who achieved clinical remission with and without endoscopic healing.

Healing of the mucosa has been associated with a prolongation of the symptom-free interval in comparison with the non-healed bowel, in an endoscopy substudy from D'Haens et al.⁵ who followed-up patients from the ACCENT 1 study and stratified them according to the degree of mucosal healing observed at the 54-week endoscopy. Irrespective of the treatment arm (infliximab scheduled or episodic), patients with complete or partial mucosal healing remained without clinical relapse for a median of 20 and 19 weeks, respectively. In contrast, those without mucosal healing developed clinical relapse after a median of only 4 weeks.

In conclusion, in patients whose clinical remission is not associated with mucosal healing or at least improvement of the mucosal lesions, intensification of treatment is not currently recommended.

Limited information is available on the optimal duration of immunosuppressive/biologic therapy in inflammatory bowel disease patients in prolonged clinical remission. Open-label experiences with anti-TNF alpha agents suggest that mucosal healing at the time of treatment withdrawal may predict better outcomes, whereas data on azathioprine show no influence.

Two studies have recently investigated relapse rates following thiopurine withdrawal in patients with Crohn's disease or ulcerative colitis who are in stable remission.^74,75 Five-year relapse rates indicated that only about one third of patients with inflammatory bowel disease remain in remission after discontinuing medication. However, neither of the studies provided endoscopic data on these patients. The GETAID azathioprine withdrawal trial⁴⁰ examined patients with Crohn's disease who had been in clinical remission on azathioprine for at least 42 months. Treatment was continued for 18 months with azathioprine or placebo. Kaplan–Meier estimates of the 18-month relapse rates were almost 3 times higher in the placebo-treated group (21% versus 8% for azathioprine). Thirty-eight percent of the patients in the placebo group had endoscopic disease activity when azathioprine was discontinued (baseline), but the presence of mucosal ulceration was not associated with the likelihood of disease relapse.

A substudy of the “step-up/top-down” trial looked at the endoscopic appearance of the gut mucosa after 2 years of treatment as a possible predictor of disease outcome in the following 2 years.⁴ At the end of year 2, 49 of the trial participants underwent ileocolonoscopy, and the presence of complete mucosal healing during that examination was the only predictor of sustained corticosteroid- and infliximab-free clinical remission during years 3 and 4 (p = 0.036; OR 4.35, 95% CI 1.1–17.2).

Finally, the STORI trial provides data about factors associated with clinical relapse following infliximab discontinuation in patients with Crohn's disease who had been treated with infliximab plus azathioprine and were in stable remission for at least 6 months. Twelve months after infliximab was discontinued, clinical relapse was observed in 45 of the 115 patients with luminal Crohn's disease, and one of the predictors of this outcome was the absence of mucosal healing when the drug was withdrawn.³²

• Treatment intensification to achieve mucosal healing in patients with Crohn's disease in clinical remission but not endoscopically healed is not recommended because of lack of evidence.

• In patients with ulcerative colitis, the addition of topical therapy to the oral regimen increases mucosal healing rates.

• Mucosal healing in Crohn's disease assessed at the time of anti-TNF alpha discontinuation is associated with better outcomes. Conversely, data on azathioprine show no influence. This point has not been formally investigated in ulcerative colitis.

5 Conclusions

There is a growing body of evidence that supports the importance of mucosal healing in both Crohn's disease and ulcerative colitis. In the era of biological therapy for inflammatory bowel disease, this parameter is emerging as a reliable predictor of disease control reflected by fewer surgical interventions and higher remission rates in Crohn's disease and lower colectomy rates and longer periods of disease control in ulcerative colitis. With newer forms of therapy, rapid, complete endoscopic remission is an attainable goal. Additional research is needed to determine whether/how the endoscopic appearance of the gut mucosa should be used in the therapeutic decision-making process, but its assessment is already an essential component of clinical practice and studies in the field of inflammatory bowel disease.

Conflict of interest

AA received consultancy, lecture fees and educational grants from MSD/Schering Plough, Abbott, and Nycomed. GD served as consultant, lecturer, investigator and author for MSD/Schering Plough, Jansen Biologics/Centocor, Ferring, Dr FalkPharma, Abbott, Millenium/Takeda, Pfizer, Shire, Vifor, UCB, and NovoNordisk. GVA received research support and honoraria from Abbott, MSD, Ferring, Shire and UCB. GPC: none. AG received research support from Abbott and Merck/Janssen; speaker honoraria and consultant fees from Abbott, Centocor/MSD/Janssen, and Nestle. MS received consultancy and lecture honoraria from MSD and Abbott. JCP received honoraria and consultancy fees from Abbott, Falk, and Medac. ML received speaker honoraria and consultancy fees from Abbott, MSD, Falk, Ferring and Ardeypharm.

Acknowledgments

Andreas Sturm organized this workshop, on behalf of the ECCO scientific committee.

Alessandro Armuzzi and Geert D'Haens drafted and edited the manuscript.

All authors read and approved the final manuscript.