Medical management of mild to moderate Crohn’s disease: evidence-based treatment algorithms for induction and maintenance of remission

Summary

Background  Patients with Crohn’s disease alternate between periods of active, symptomatic disease and periods of remission. The treatment goal for Crohn’s disease is to induce and then maintain remission of symptoms.

Aim  To review evidence from randomized, controlled, clinical trials on medical therapies for inducing and maintaining remission in patients with mild-to-moderate Crohn’s disease, and to suggest the best evidence-based approaches for induction and maintenance therapies.

Methods  PubMed search using the following terms: sulfasalazine or salicylazosulfapyridine or aminosalicylate or aminosalicylic acid or mesalamine or mesalazine or corticosteroid or prednisone or prednisolone or methylprednisolone or budesonide or antibiotic or metronidazole or ciprofloxacin or immunosuppressive or azathioprine or mercaptopurine or thiopurine or methotrexate and Crohn’s disease.

Results  Randomized, controlled trials demonstrated that sulfasalazine, budesonide, and conventional corticosteroids are effective for inducing remission of mild-to-moderate Crohn’s disease when administered for a period of 8–16 weeks. An ideal maintenance therapy does not currently exist.

Conclusions  Selection of maintenance therapy is based on a combination of evidence from controlled trials and patient features including disease severity and location, co-morbidities, previous response to treatment, and previous surgical resection. The options for maintenance therapy include therapy cessation and patient observation following successful induction, budesonide, or immunosuppressive therapy.

Introduction

Crohn’s disease (CD) is a chronic inflammatory condition that is characterized by repetitive cycles of active and quiescent disease. As a result, the clinician’s goals of therapy for this disease are to both induce and maintain remission of symptoms.

Numerous agents are available for the treatment of CD. The selection of an optimum treatment depends on several factors, including disease severity and location, co-morbidities, previous response to treatment, and the presence of previous surgical resection. Treatment of active disease has improved considerably; however, maintenance of remission remains an unmet medical need. In population-based studies, approximately 75% of individuals relapse over 5 years.¹ In clinical trials, our most effective agents are only 25% to 40% effective in maintaining remission over 1 year. Furthermore, the agents that are effective have the potential for important adverse events. Thus, the balance between efficacy and adverse events must be carefully considered when a maintenance therapy is prescribed.

This review focuses on the medical management of patients with mild to moderate CD. Mild to moderate CD has been defined by the American College of Gastroenterology (ACG) as, ‘ambulatory patients able to tolerate oral alimentation without manifestations of dehydration, toxicity (high fevers, rigors, prostration), abdominal tenderness, painful mass, obstruction, or >10% weight loss’.² Practice guidelines for the management of CD in adults were last published by the ACG in 2001,² while the American Gastroenterological Association (AGA) recently released a position statement on medical therapy with corticosteroids, immunomodulators, and infliximab for inflammatory bowel disease.³ Since the ACG guidelines were last published, there have been additional clinical trials reported that are relevant to the medical management of CD. Additionally, the AGA position statement, which was limited to corticosteroids, immunomodulators, and infliximab, does not adequately address the treatment of patients with mild to moderate CD. Therefore, in an effort to assist clinicians in managing mild to moderate CD, we have prepared this updated systematic review and proposed integrated, evidence-based, induction and maintenance treatment algorithms.

Methods

A search of the online bibliographic database PubMed (1966 to November 2005) was performed to identify potentially relevant English-language articles. The Medical Subject Heading terms sulfasalazine or salicylazosulfapyridine or aminosalicylate or aminosalicylic acid or mesalamine or mesalazine or corticosteroid or prednisone or prednisolone or methylprednisolone or budesonide or antibiotic or metronidazole or ciprofloxacin or immunosuppressive or azathioprine or mercaptopurine or thiopurine or methotrexate and Crohn’s disease were used to perform keyword searches of the database. Manual searches of the reference lists from the potentially relevant articles and the proceedings from annual AGA, ACG and UEGW meetings from 2000 to 2005 were performed to identify additional studies that may have been missed using the computer-assisted search strategy. Studies selected were randomized, controlled, medical therapy trials of patients with mild to moderate CD.

Induction therapy for mild to moderate CD

In clinical practice, active CD is generally defined by the presence of symptoms characteristic of the disease, such as chronic or nocturnal diarrhoea, abdominal pain, and rectal bleeding; in clinical trials disease activity is measured by a CD Activity Index (CDAI) score ≥150.^{2, 4} Multiple pharmacologic treatments have been evaluated for the reduction or resolution of CD symptoms. In 2003, a comprehensive evidence-based review and treatment algorithm for induction of mild to moderately active CD was published.⁵ Since that time, there has been limited new data examining induction of remission of mild to moderate CD. One new study, comparing budesonide with prednisolone,⁶ is addressed in the budesonide induction therapy section. Thus, due to limited new data, the following section reviewing induction of remission therapies will highlight the main results/conclusions for each drug.

Oral aminosalicylates

According to the Practice Parameters Committee of the ACG,² and the British Society of Gastroenterology,⁷ oral aminosalicylates (mesalazine 3.2–4.0 g/day or sulfasalazine 3–6 g/day) are indicated for use as a first-line therapy for patients with mild to moderately active CD.^{2, 7} Conversely, the European Crohn’s and Colitis Organization Consensus recently stated that oral aminosalicylates are not recommended for mild to moderate CD.⁸ Thus, it is obvious that there is conflicting evidence on the efficacy of oral aminosalicylates in active CD and their use in treating mild to moderate CD has been debated.^{9, 10}

Sulfasalazine

The results of several clinical trials demonstrate that sulfasalazine at daily doses of 3–6 g/day is more effective than placebo for inducing remission of mild to moderate CD.^11–14 However, subgroup analysis of these studies suggest that efficacy of sulfasalazine is greatest in patients with active colonic or ileocolonic disease.^{12, 13} Sulfasalazine also has a slower onset of action than prednisone or 6-methylprednisolone and is substantially less effective.^{12, 13} As an adjunctive therapy, sulfasalazine is not more effective than prednisone alone, nor is it steroid-sparing.^{13, 15}

Adverse events in patients treated with sulfasalazine include headache, epigastric pain, nausea, vomiting, cyanosis, skin rash, fever, hepatitis, autoimmune haemolysis, aplastic anaemia, leucopoenia, agranulocytosis, folate deficiency, pancreatitis, systemic lupus erythematosus, sulphonamide-induced toxic epidermal necrolysis, Stevens-Johnson syndrome, pulmonary dysfunction and male infertility.^{16, 17} Most of the side effects of sulfasalazine can be attributed to the systemic absorption of sulfapyridine and the adverse effects occur more frequently in patients who are genetically predisposed to ‘slow’ acetylation of sulfapyridine to N-acetylsulfapyridine in the liver.¹⁶ Some of the side effects (headache, nausea, vomiting, and epigastric pain) are dose-related and can be minimized by gradual dose escalation.¹⁷

In ulcerative colitis, the active component of sulfasalazine is believed to be 5-aminosalicylate (mesalazine),¹⁸ whereas in rheumatoid arthritis, the active component of sulfasalazine is believed to be systemically absorbed sulfapyridine.¹⁹ It is unknown whether the active component of sulfasalazine in CD is 5-aminosalicylate or sulfapyridine, and the possibility that sulfapyridine may contribute to the efficacy of sulfasalazine in CD could explain the observed differences in efficacy between sulfasalazine and mesalazine (see below).

Mesalazine

Although mesalazine has been used as first-line treatment of mild to moderate CD for many years, the efficacy of this treatment has not been consistently demonstrated in randomized, placebo-controlled studies. Results of two small placebo-controlled studies that evaluated mesalazine 3.2 g/day or 4 g/day suggest that mesalazine was more effective than placebo.^{20, 21} However, in the Tremaine et al. study,²¹ improvement over placebo was demonstrated by combining patients that showed partial or complete remission; mesalazine was not significantly different than placebo when complete remission was evaluated. Five placebo-controlled studies evaluating mesalazine (1–2 g/day in three studies and 4 g/day in two larger studies) failed to demonstrate that mesalazine was more effective than placebo.^22–25 In addition, results of two small studies showed that mesalazine 4 g/day was similar in efficacy to ciprofloxacin 1000 mg/day or 6-methylprednisolone 40 mg/day for treating active CD;^{26, 27} however, these studies were not powered to demonstrate non-inferiority (equivalence) and are thus not interpretable. In yet another large comparative trial, mesalazine 4 g/day was shown to be less effective than budesonide 9 mg/day.²⁸

Adverse events in patients treated with mesalazine occur infrequently. Rare, but serious adverse events include pulmonary toxicity, pericarditis, hepatitis and pancreatitis.^{29, 30} Interstitial nephritis has also been reported to occur infrequently in patients treated with mesalazine; however, it is unclear if mesalazine is the cause of the renal lesions.

Antibiotics

The bacterial flora in the intestine likely plays a role in the pathogenesis of CD. Thus, antibiotics have been explored as a therapeutic agent. Metronidazole and ciprofloxacin are the most widely used antibiotics in clinical practice for the treatment of CD; however, the efficacy of these agents has not been convincingly demonstrated in controlled clinical trials. Other antibiotics, such as clarithromycin and ethambutol, have been investigated for CD treatment in randomized controlled trials and did not offer benefit to patients.³¹

Metronidazole Only a few small studies have evaluated metronidazole for treatment of active, inflammatory CD. Metronidazole, administered at doses of 0.8 g/day, 1 g/day, 10 mg/kg/day, 20 mg/kg/day or 0.8 g/day in combination with cotrimoxazole was not more effective than placebo in inducing remission of mild to moderate CD.^31–35 Subgroup analysis did reveal that there was a trend toward better efficacy of metronidazole in patients with disease involving the colon.^{32, 35} A small comparison trial of metronidazole (0.8 g/day) versus sulfasalazine (3 g/day) in patients with active CD found no significant difference in the efficacy of the two treatments but this study was not powered to demonstrate non-inferiority (equivalence) and is thus not interpretable.³⁶

Adverse events in patients treated with metronidazole may include nausea, a metallic taste, disulfuram-like reaction and sensory peripheral neuropathy.^{37, 38} Peripheral neuropathy has been reported in patients receiving high-doses of metronidazole treatment for CD (i.e., 15–20 mg/kg·day);^38–41 this is usually observed only after a cumulative dose of >30 g has been received.⁴¹ The neuropathy is generally reversible, although in some cases it may take up to 2 years for symptoms to abate.⁴¹

Ciprofloxacin As with metronidazole, only a few small studies have evaluated ciprofloxacin for treatment of active CD. A single, underpowered (n =47) study that compared ciprofloxacin (1 g/day) to placebo suggested that ciprofloxacin was more effective than placebo at inducing remission.⁴² In contrast, no significant differences in remission rates were found between ciprofloxacin (1 g/day) and mesalazine (4 g/day) in a small non-blinded study.²⁶ Also, no significant differences were found between ciprofloxacin and metronidazole combined therapy (1 g/day each) versus methylprednisolone, although there was a trend for methylprednisolone to be more effective (remission rates at 12 weeks were 46% and 63%, respectively).⁴³ Additionally, in the largest randomized controlled trial yet performed in this area, combination therapy of ciprofloxacin with metronidazole and budesonide was not more effective than monotherapy with budesonide;⁴⁴ nor is combination therapy of ciprofloxacin and prednisolone more effective than monotherapy with prednisolone.⁴⁵ Ciprofloxacin has fewer side effects than metronidazole, but it can cause nausea, diarrhoea and skin rashes.⁴⁶ Tendonitis and Achilles tendon rupture have been associated with ciprofloxacin use, but these are rare complications.⁴⁷ In conclusion, the available data do not support the use of antibiotics as induction therapy for mild to moderate CD.

Conventional Corticosteroids

Conventional corticosteroids, such as prednisone, prednisolone, and 6-methylprednisolone, have all demonstrated efficacy in active CD^{12, 13} and are perceived as the most effective therapeutic option for inducing remission of mild to moderate CD. However, their use as a first-line therapy in mild to moderate disease is limited due to their numerous side effects.^{7, 12, 13} Common adverse effects include acne, hypertension, hirsutism, striae, moon face, cataracts and infection.^{13, 48} Additionally, corticosteroids can induce a loss of bone mineral density^{49, 50} and increase the relative risk of fractures.⁵¹ Conventional corticosteroid use is also associated with hyperlipidemia, hypokalemia, hyperglycemia and hypocalcemia.^{52, 53} Thus, weighing the risk/benefit ratio, conventional corticosteroids are not an ideal first-line therapy for mild to moderately active CD.

Non-systemic Corticosteroids

Budesonide Budesonide is a locally acting, topically delivered, corticosteroid that undergoes extensive first-pass hepatic metabolism (80% to 90%) and accordingly, has low systemic absorption.⁵⁴ Thus, the benefits of a corticosteroid in managing mild to moderate CD can be achieved with a reduced risk of systemic adverse effects. Recently, both an AGA position statement and the European Crohn’s and Colitis Organization Consensus recommended budesonide as a first-line therapy for mild to moderate CD of the ileum and proximal colon.^{3, 8} These recommendations were made based upon several randomized controlled trials. Greenberg et al. reported that budesonide 9 mg/day for 8 weeks was significantly more effective than placebo at inducing remission in active CD (51% vs. 20%, respectively).⁵⁵ A similar study found that budesonide 9 mg/day for 8 weeks resulted in remission in 48% of active CD patients but was not significantly different from placebo due to the high remission rate, 33%, in the placebo-treated patients.⁵⁶ A meta-analysis demonstrated that budesonide is more likely to induce remission than placebo and 5-aminosalicylic acid (5-ASA).⁵⁷ Budesonide, in comparison to placebo, induced remission more frequently (odds ratio = 1.85).⁵⁷ Budesonide 9 mg/day is also a more effective induction therapy than mesalazine 4 g/day as assessed by remission rates at 8 weeks (69% vs. 45%; P =0.001) and 16 weeks (62% vs. 36%; P <0.001).²⁸ Additionally, among five clinical trials (four randomized, double-blind, double-dummy trials^58–61 and one randomized open trial⁶²), budesonide was found to be only slightly less effective than prednisolone or 6-methylprednisolone at inducing remission of mild to moderate CD (Figure 1). However, it was noted by the authors that in a sub-analysis of patients with low disease activity, budesonide had similar remission rates as conventional corticosteroids.⁵⁷

Figure 1.  Frequency of remission for budesonide versus conventional corticosteroids for active Crohn’s disease. Reproduced with the permission of Blackwell Publishing from Ref. (57).

The number and severity of adverse events associated with budesonide are similar to placebo.^{55, 56} Additionally, budesonide has a more favourable adverse event profile than conventional corticosteroids.^58–61 Furthermore, in comparison to mesalazine, budesonide has a comparable adverse event profile, yet mesalazine-treated patients have a significantly greater number (P =0.04) of severe adverse events due to the progressive worsening of CD.²⁸ Thus, budesonide is safe and effective as an induction therapy for mild to moderate CD involving the distal small intestine and proximal colon.

Summary of induction therapies

Data from well-designed clinical trials demonstrate that sulfasalazine, budesonide and prednisone are effective medical therapies for induction of remission of mild to moderate CD (Table 1), while mesalazine and antibiotics have not demonstrated efficacy. Utilizing the evidence presented above, an evidence-based approach to managing active mild to moderate CD is proposed in Figure 2. For patients with mild to moderate CD of the ileum and/or proximal colon, the best treatment option is budesonide 9 mg/day administered for 8–16 weeks followed by tapering over 2–4 weeks in increments of 3 mg. For patients with mild to moderate CD confined to the colon and without systemic symptoms who are not known to be intolerant to sulfa compounds, the best option is sulfasalazine 3–6 g/day for up to 16 weeks. For patients with moderate CD and systemic symptoms, the evidence suggests treatment with prednisone 40–60 mg/day, followed by a tapering schedule beginning after 2–4 weeks (the initial taper is by 5 mg/week until a dosage of 20 mg/day is reached and then tapering continues by 2.5–5 mg/week until prednisone is stopped). Additionally, patients who initiate therapy with budesonide or sulfasalazine and fail to respond to either of these first-line therapies should be subsequently treated with prednisone 40–60 mg/day, followed by a tapering schedule over 8–16 weeks. Infliximab (5 mg/kg at weeks 0, 2 and 6, followed by 8 mg/kg every 8 weeks thereafter) may be an option for patients who fail to respond to budesonide or sulfasalazine, or for those who cannot tolerate prednisone treatment; these patients may need to be reclassified as moderate to severe patients.

Table 1.   Quality of evidence supporting induction of remission algorithm recommendations

Conclusions/recommendations	Quality of evidence*
CD, Crohn’s disease; AGA, American Gastroenterological Association. * Based on AGA guidelines for evidence-based position statements: A, homogeneous evidence from multiple well-designed, randomized (therapeutic) or cohort (descriptive) controlled trials, each involving a number of participants to be of sufficient statistical power; B, evidence from at least 1 large well-designed, clinical trial with or without randomization from cohort or case-control analytic studies or well-designed meta-analysis; C, evidence based on clinical experience, descriptive statistics, or reports of expert committees; D, not rated.
Sulfasalazine 3–6 g/day for induction of remission of mild to moderate CD involving the left colon	A
Budesonide 9 mg/day for induction of remission of mild to moderate CD involving the ileum and/or proximal colon	A
Prednisone for induction of remission of mild to moderate CD when patients fail to respond to sulfasalazine or budesonide	B
Infliximab for induction of remission of mild to moderate CD in patients who fail to respond to sulfasalazine or budesonide, or in patients who have undesirable side-effects following prednisone therapy	B
Prednisone for induction of remission of moderate CD with systemic symptoms	A

Figure 2.  Best evidence-based approach to managing active mild to moderate Crohn’s disease (CD) [adapted with the permission of Blackwell Publishing from Ref. (5)].

Maintenance therapy following treatment of mild to moderate CD

Because CD is a chronically relapsing disease, the goal of treatment following medically induced remission of disease (CDAI score <150⁴) is to prevent symptom recurrence for as long as possible. The pharmacologic options available are reviewed in the following section.

Oral aminosalicylates

Sulfasalazine As noted above, sulfasalazine has modest efficacy as an induction therapy; however, this agent is not efficacious as a maintenance therapy. Two large, placebo-controlled trials have demonstrated that sulfasalazine 1.5–3.0 g/day is not more effective than placebo at maintaining remission.^{12, 13}

Mesalazine Mesalazine (3 g/day) was found to be no different than placebo at maintaining remission for 1 year in two large studies.^{63, 64} In the Sutherland et al. study,⁶³ 25% of mesalazine-treated patients (n =118) experienced a relapse and 36% of the placebo-treated patients (n =128) experienced a relapse at 1 year (P =0.056). Similarly, Thomson et al.⁶⁴ reported no significant difference (P =0.45) in the cumulative proportion of patients relapsing at 1 year. Other studies have reported subgroup analyses suggesting that mesalazine may be effective at maintaining remission based upon a patient’s risk of relapse. For example, mesalazine was suggested to be effective in high-risk patients (in remission <3 months).⁶⁵ However, this finding has not been replicated. Also, caution should be exercised when interpreting this result as the absolute number of patients was small.⁶⁵ The International Mesalazine Study Group⁶⁶ and the Italian IBD Study Group have also reported beneficial results for mesalazine.⁶⁷ The International Mesalazine Study Group analysed 206 patients for 12 months and reported that oral 5-ASA 1.5 g/day was more effective than placebo at maintaining remission (relapse rate for 5-ASA: 22.4%, placebo: 36.2%; P =0.0395). It should be noted that in this study 30% of the patients were in surgically induced remission and the benefits of mesalazine were most evident in this patient population. Prantera et al.⁶⁷ randomized 125 patients with inactive CD to receive mesalazine 2.4 g/day for 12 months and found mesalazine to be more effective than placebo (relapse rates 34% and 55%, respectively; P =0.02). Again, a significant decrease in risk of relapse was found for patients with prior surgical resection. Thus, it is difficult to interpret these results for patients with medically-induced remission. As a result of the conflicting results,^68–72 5 meta-analyses have been performed.^73–77 Salomon et al.⁷⁴ evaluated five trials for medical maintenance of remission and concluded that aminosalicylates were ineffective. In contrast, Steinhart et al.⁷⁵ analysed 10 trials and concluded that mesalazine, but not sulfasalazine, was effective as a maintenance therapy. It should be noted that in this meta-analysis, patients with medically or surgically induced remission were not analysed separately. A third meta-analysis evaluated five clinical trials and like Steinhart et al., determined that 5-ASA significantly reduced the relapse frequency at 6, 12 and 24 months.⁷³ A fourth meta-analysis,⁷⁶ found that mesalazine significantly lowers the relative risk of relapse in patients with surgically induced remission (−13.1%), but does not lower the relative risk of relapse in patients with medically induced remission (−4.7%). Finally, a recent Cochrane meta-analysis based on six eligible trials,⁷⁷ found no evidence indicating that 5-ASA preparations are superior to placebo for maintaining medically induced remission in CD patients at 12 or 24 months. At 12 months, the odds ratio was 1.00 and at 24 months the odds ratio was 0.98.

Olsalazine Olsalazine, a dimer comprised of two 5-ASA molecules linked by an azo bond, has also been evaluated as a treatment for maintenance of medically induced remission in CD patients. The data reveals that there was no significant difference in the frequency of the relapse rate between the two groups as 48.5% of olsalazine-treated (2 g/day) patients (n =167) and 45% of placebo-treated patients (n =160) experienced a relapse over 1 year.⁷⁸ The investigators also reported more olsalazine-attributed adverse events than placebo, with significantly greater gastrointestinal symptoms causally related to olsalazine treatment (olsalazine 40.7% vs. placebo 26.9%; P =0.01).

Balsalazide No data exist for balsalazide. Based on the lack of efficacy of mesalazine and olsalazine, it would not be expected that balsalazide would be effective.

In summary, sulfasalazine and olsalazine have not proven efficacious as a maintenance therapy for mild to moderate CD and the data examining mesalazine as a maintenance therapy is inconsistent at best. This information is summarized in Table 2.

Table 2.   Outcomes from clinical trials examining oral aminosalicylic acid as a maintenance therapy for patients with Crohn’s disease in medically induced remission [adapted, in part, from Ref. (76)]

Reference	Drug	No. of patients	Study length	Relapse (n)
5-ASA, 5-aminosalicylic acid. * Significantly different (P <0.05) from placebo.
Sutherland et al. 1997 (63)	placebo	152	24 months	61
	mesalazine 3 g/day	141		41*
Thomson et al. 1995 (64)	placebo	148	24 months	38
	mesalazine 3 g/day	138		33
Gendre et al. 1993 (65)	placebo	81	24 months	36
	mesalazine 2 g/day	80		30
International Mesalazine Study Group 1990 (66)	placebo	123	12 months	27
	5-ASA 1.5 g/day	125		17
Prantera et al. 1992 (67)	placebo	61	12 months	32
	5-ASA 2.4 g/day	64		19
Arber et al. 1995 (68)	placebo	31	24 months	15
	5-ASA 1 g/day	28		6
Bresci et al. 1994 (69)	placebo	33	48 months	22
	5-ASA 2.4 g/day	33		17
Brignola et al. 1992 (70)	placebo	22	4 months	13
	5-ASA 2 g/day	21		11
Modigliani et al. 1996 (71)	placebo	37	12 months	20
	mesalazine 4 g/day	48		27
De Franchis et al. 1997 (72)	placebo	59	24 months	29
	5-ASA 3 g/day	58		34
Mahmud et al. 2001 (78)	placebo	160	12 months	72
	olsalazine 2 g/day	167		81

Immunosuppressive agents

Azathioprine and mercaptopurine, two commonly utilized immunosuppressive agents, are not indicated for use as induction therapy in patients with mild to moderately active CD due to their relatively slow onset of action (≥3 months).⁷⁹ However, these agents have been found to be effective as a maintenance therapy following corticosteroid-induced remission of patients with mild to moderate CD.² Azathioprine 2.5 mg/kg was more effective than placebo at maintaining remission for up to 15 months.⁸⁰ Following 15 months of therapy, the proportion of azathioprine-treated patients in remission was 42% compared with 7% of placebo-treated patients (P =0.001). Furthermore, withdrawal of azathioprine maintenance therapy leads to a greater rate of relapse than continuation of azathioprine.⁸¹ Eighty-three patients, who were on azathioprine therapy and in remission for at least 42 months, were randomized to continue azathioprine therapy (n =40) or switched to placebo (n =43). At 18-months follow-up, three azathioprine patients (8%) had relapsed and nine placebo-treated patients (21%) had relapsed. Additional evidence for the use of thiopurines as maintenance therapy comes from a meta-analysis of five trials examining azathioprine for maintenance of remission. The results revealed a common odds ratio of 2.3 for response to azathioprine. However, it should be noted that the study design of the individual studies varied in length, from 26 weeks to 24 months, differed in outcome measures, and some studies allowed concurrent steroid therapy to maintain symptom control, while others did not.⁸²

Azathioprine therapy may also be advantageous for weaning CD patients off conventional corticosteroid therapy. Seventy-five percent of azathioprine-treated patients were able to reduce or discontinue steroid usage compared with 36% of placebo-treated patients,⁷⁹ a finding which was further supported by meta-analyses.^{76, 82}

A concern of long-term, maintenance therapy is the potential for adverse side effects. Approximately 2% to 8% of patients report mercaptopurine-induced toxicity events including pancreatitis, bone marrow depression, allergic reactions and infectious complications.⁸³ Others have also reported pancreatitis, allergy, and opportunistic infection, plus additional adverse events such as leucopoenia and neutropenia.^{82, 84, 85} Furthermore, there is an increase in the rate of lymphoma during treatment with azathioprine.^86–89 In summary, there is support for the use of azathioprine 2.0–2.5 mg/kg/day or mercaptopurine 1.0–1.5 mg/kg/day as maintenance therapy for mild to moderate CD (see Table 3). However, these agents are currently reserved for patients who are steroid-dependent or have complications such as fistulizing disease because of their potential for serious toxicity.

Table 3.   Randomized, double-blind, placebo-controlled trials of azathioprine and mercaptopurine as maintenance therapy for Crohn’s disease

Reference	Drug	No. of patients	Study length	Outcome
* Outcome assessed as investigator grading of improvement/response to therapy. Grading was +3 for excellent improvement, through 0 for no change, to −3 for severe worsening. Included are the patients that completed the first year of study, which was designed to be a 2-year, crossover study. The first year results were analysed separately as if the study had been designed to be a parallel study; † Significantly different from placebo (P <0.05); ‡ Outcome assessed as relapse rates [Crohn’s disease Activity Index (CDAI) score >150 with an increase of at least 60 points] and median time to relapse; § Outcome assessed as estimated common odds ratio (a value >1 indicates a therapeutic benefit compared with placebo).
Present et al. 1980 (79)	placebo	36	12 months	26/36 (72%)*
	mercaptopurine 1.5 mg/kg	36		5/36 (14%)†
Candy et al. 1995 (80)	placebo	30	15 months	Relapse rate = 93%
	azathioprine 2.5 mg/kg	33		Relapse rate = 58%†
Lemann et al. 2005 (81)	placebo	43	18 months	21%; 15.9 months‡
	azathioprine (subjects in remission >42 months for azathioprine)	40		7.5%; 17.3 months‡†
Pearson et al. 1995 (meta-analysis)(82)	placebo azathioprine or	183	26 weeks to 24 months	3.09 (CI, 2.45–3.91)§
	mercaptopurine	136

Methotrexate, an immunosuppressive agent that has been successfully used as a treatment for rheumatoid arthritis, has also been found to be an effective induction therapy for steroid dependent/resistant CD.^90–92 However, like azathioprine and mercaptopurine, it has a slow onset of action. Thus, it is not recommended as an induction treatment for mild to moderately active CD. Methotrexate is an effective therapy for maintaining methotrexate-induced remission.⁹³ In a study by the North American Crohn’s Study Group investigators, low-dose methotrexate (15 mg weekly, intramuscularly) or placebo was randomly administered to patients (n =76) who had achieved methotrexate-induced remission in a previous study.⁹³ At the end of 40 weeks of treatment, 26 patients (65%) treated with methotrexate remained in remission over the entire duration of follow-up, compared with 14 placebo-treated patients (38.9%) (P =0.04). Safety information on methotrexate is largely based on studies evaluating treatment of rheumatoid arthritis.⁹⁴ Nausea is the most common minor side effect of methotrexate treatment, and it tends to occur for a period of 24–48 h following weekly injections. Leucopoenia and associated opportunistic infection have been reported, but rarely. Other concerns include hypersensitivity pneumonitis (occurs in up to 1% of patients) and hepatotoxicity (clinically important hepatotoxicity rarely occurs when treatments are administered weekly versus daily). Methotrexate cannot be used by pregnant women, as it is teratogenic. No high quality data indicate that methotrexate is associated with malignancy.

Ciclosporin, another immunosuppressive agent, has not been found to be an effective induction or maintenance treatment for CD.^{95, 96}

Conventional corticosteroids

Conventional corticosteroids are not indicated for maintaining remission of CD due to their lack of efficacy and extensive systemic side effects.² While one small study⁹⁷ reported methylprednisolone to be more beneficial than placebo at maintaining remission, two large studies^{12, 13} have demonstrated prednisone and 6-methylprednisolone to be no different than placebo at maintaining medically induced remission (Table 4). The National Cooperative Crohn’s Disease Study reported that prednisone 0.25 mg/day was not effective at a 2-year follow-up at preventing relapses among patients in remission.¹² Similarly, the European Cooperative Crohn’s Disease Study determined that 6-methylprednisolone 8 mg once daily was no better than placebo at maintaining remission at 2 years.¹³ Additionally, a Cochrane Database Review also concluded that conventional corticosteroids were not an effective maintenance therapy for up to 2 years.⁹⁸ Interestingly, a subset of patients exists that initially respond to conventional corticosteroid therapy for active CD and then relapse during tapering of conventional corticosteroids. Following another course of induction therapy with conventional corticosteroids, these patients can then be maintained asymptomatic with long-term corticosteroid therapy.^{13, 99, 100} These patients are classified as steroid-dependent.

Table 4.   Randomized, double-blind, placebo-controlled trials of conventional corticosteroids as maintenance therapy for Crohn’s disease

Reference	Drug	No. of patients	Study length	Outcome
* Outcome assessed as relapse rates based upon life-table analysis. Additionally, outcome ranking by the Wilcoxon Rank Sum method revealed worsening of disease in 36 of 101 (36%) placebo-treated patients and in 26 of 61 (43%) of prednisone-treated patients; † Outcome assessed as relapse rates.
Summers et al. 1979 (12)	placebo	101	12 months	∼25%*
	prednisone 0.25 mg/kg/day	61		∼25%
Malchow et al. 1984 (13)	placebo	52	24 months	Not statistically different according to life table analysis of ‘failure and relapse’
	6-methylprednisolone 8 mg/day	66
Brignola et al. 1988 (97)	placebo	9	6 months	78% (7 of 9 relapsed) †
	methylprednisolone 0.25 mg/kg/day	9		11% (1 of 9 relapsed)

The adverse events associated with conventional corticosteroids are well documented and have been previously highlighted. Obviously, there is heightened concern for adverse events associated with long-term conventional corticosteroid use, including suppression of adrenal function. Thus, the clinician must also consider the adverse events when recommending the appropriate maintenance therapy. In summary, conventional corticosteroids are not indicated for use as a maintenance therapy for mild to moderate CD due to their limited efficacy and adverse effects associated with long-term use.

Non-systemic corticosteroids

Budesonide Budesonide, compared with placebo, has been shown to prolong the time to relapse in CD patients with medically-induced remission. This observation is based on a pooled analysis¹⁰¹ of four randomized, double-blind, placebo-controlled trials with similar design.^102–105 These four studies, which were predesigned for completion of a pooled analysis, showed that budesonide 6 mg/day is effective at prolonging the time patients are in remission compared with placebo. The median time to relapse was 268 days for budesonide-treated patients and 154 days for placebo-treated patients (P =0.007).¹⁰¹ In addition, budesonide 6 mg/day was effective at maintaining remission for 3 and 6 months; however, at 1 year, relapse rates did not differ between the two groups.¹⁰¹

Additionally, budesonide 6 mg/day may be an option for steroid-dependent patients. In one study, patients receiving prednisolone (10–30 mg/day for 6 months) had their medication tapered to discontinuation and were randomly assigned to budesonide 6 mg/day or placebo. After 6 months, relapse rates were 32% and 65% (P <0.001) in the budesonide and placebo-treated patients, respectively.¹⁰⁶ Budesonide 6 mg/day has also been found to be more effective than mesalazine 3 g/day at maintaining remission and an improved quality of life in patients with steroid-dependent CD.¹⁰⁷ Relapse rates were 55% for budesonide-treated patients and 82% for mesalazine-treated patients (P =0.045) and the mean time to relapse was 241 days and 147 days, respectively (P =0.003).¹⁰⁷ Importantly, the incidence and types of adverse events were similar between the two groups.

The efficacy of budesonide for maintenance of remission has also been compared with other therapeutic agents, such as conventional corticosteroids and mesalazine. For example, it was reported that a variable dose regimen of budesonide (0–9 mg/day adapted to disease activity) resulted in similar control of disease activity as prednisolone (0–40 mg/day adapted to disease activity) over a 2-year period, but with significantly fewer emergent corticosteroid side effects. In addition, bone mineral density loss was significantly lower in steroid-naïve patients treated with budesonide compared with prednisolone.^{108, 109}

Budesonide as a maintenance therapy has been demonstrated to be well tolerated. In placebo-controlled trials and a meta-analysis examining the safety of budesonide, the adverse event profile was similar between budesonide and placebo treatment. Specifically, the overall number of adverse events and corticosteroid-associated side effects were not significantly different between the two groups.^{57, 102, 104, 105} Additionally, a change in therapy from conventional corticosteroids to budesonide results in a significant reduction in the number of corticosteroid-associated adverse events.^{106, 110} Modest dose-dependent adrenal suppression has been shown to occur with budesonide treatment; however, these effects are significantly less than those seen with systemic steroids, such as prednisolone. In addition, budesonide is associated with better bone mass preservation than conventional steroids and is not associated with occurrence of metabolic bone disease.¹⁰⁸ A safety profile of budesonide (2–21 mg/day; 9 mg/day the most common daily dosage) from a compassionate use program reported only 232 adverse events (59.5% were gastrointestinal) among 4092 patients.¹¹¹ Furthermore, the incidence of serious adverse events associated with corticosteroid use was rare.

In summary, budesonide 6 mg/day is well tolerated and effective at prolonging the median time to relapse in patients with mild to moderate CD (Table 5). Although relapse rates at 1 year did not differ between patients receiving budesonide or placebo, budesonide 6 mg/day was effective at maintaining remission of CD for 3–6 months.

Table 5.   Randomized, controlled trials of budesonide as maintenance therapy for Crohn’s disease

Reference	Drug	No. of patients	Study length	Outcome*
CDAI, Crohn’s disease Activity Index. * Outcome assessed as relapse rates and median time to relapse, unless otherwise indicated; † Median time to relapse significantly different (P <0.05) from comparator; ‡ Relapse rate significantly different (P <0.05) from comparator.
Greenberg et al. 1996 (102)	placebo	33	52 weeks	67%; 39 days
	budesonide 6 mg/day	36		61%; 178 days†
Lofberg et al. 1996 (103)	placebo	27	52 weeks	63%; 146 days
	budesonide 6 mg/day	32		59%; 271 days†
Ferguson et al. 1998 (104)	placebo	27	52 weeks	60%; 310 days
	budesonide 6 mg/day	22		48%; 275 days
Hanauer et al. 2005 (105)	placebo	55	52 weeks	58%; 169 days
	budesonide 6 mg/day	55		47%; 360 days
Sandborn et al. 2005 (pooled-analysis) (101)	placebo	145	52 weeks	62%; 154 days
	budesonide 6 mg/day	145		60%; 268 days†
Cortot et al. 2001 (106)	placebo	58	13 weeks	65%; 75 days
	budesonide 6 mg/day	59		32%; >160 days†‡
Mantzaris et al. 2003 (107)	mesalazine 3 g/day	28	52 weeks	82%; 147 days
	budesonide 6 mg/day	29		55%; 241 days†‡
Schoon et al. 2005 (108)	prednisone 14.9 mg/day	70	24 months	Comparable efficacy (CDAI∼150)
	budesonide 6.8 mg/day	66

Summary and comments on maintenance therapies

There are relatively few well-designed clinical trials for the maintenance therapy of patients with mild to moderately active CD that take into account the induction therapy used to achieve a clinical remission. To date, sulfasalazine has not been shown to be effective and conventional corticosteroids are too toxic to be considered as a safe and effective maintenance therapy. Additionally, the effectiveness of mesalazine has been inconsistent after induction with medical therapy or subsequent to surgical resections. In particular, mesalazine has not been efficacious at maintaining remissions after induction with conventional corticosteroids. Azathioprine, mercaptopurine and methotrexate have been shown to be effective in maintaining remission after corticosteroid induction therapy for moderate to severe disease, but have not been evaluated without corticosteroids in mild to moderate disease. Budesonide, at 6 mg/day, significantly prolongs the time to relapse and has a favourable safety profile, but does not maintain remission at 1 year. These data are summarized in Table 6.

Table 6.   Quality of evidence supporting maintenance of remission algorithm recommendations

Conclusions/recommendations	Quality of evidence*
CD, Crohn’s disease; AGA, American Gastroenterological Association. * Based on AGA guidelines for evidence-based position statements: A, homogeneous evidence from multiple well-designed, randomized (therapeutic) or cohort (descriptive) controlled trials, each involving a number of participants to be of sufficient statistical power; B, evidence from at least 1 large well-designed, clinical trial with or without randomization from cohort or case-control analytic studies or well-designed meta-analysis; C, evidence based on clinical experience, descriptive statistics, or reports of expert committees; D, not rated.
Budesonide 6 mg/day for maintaining remission of mild to moderate CD symptoms for up to 3–6 months in patients receiving induction budesonide therapy	A
Immunosuppressive therapy for maintenance of remission of mild to moderate CD symptoms in patients receiving induction budesonide therapy	D: there are no studies evaluating immunotherapy in patients with budesonide-induced remission. Recommendation is based on studies demonstrating effectiveness in maintaining steroid-induced remission with prednisone or prednisolone.
Immunosuppressive therapy for maintenance of remission of mild to moderate CD symptoms in patients receiving induction sulfasalazine therapy	D: There are no studies evaluating immunotherapy in patients with sulfasalazine-induced remission.
Immunosuppressive therapy for maintenance of remission of mild to moderate CD symptoms in patients receiving induction prednisone or prednisolone therapy	A: azathioprine; B: mercaptopurine and methotrexate

Thus, an ideal maintenance therapy after treatment for mild to moderate CD has not been identified. Considering that there is a substantial placebo response when the CDAI is used to assess disease activity and clinical remission in CD, and there is an absence of controlled clinical trials following induction therapy with sulfasalazine, mesalazine, and antibiotics, there is a substantial ‘evidence gap’ for a large segment of the population of patients with mild to moderate CD. Unfortunately, the available evidence is not sufficient to make recommendations for patients induced with these agents.

Based on the evidence that is available from controlled trials after treatment for mild to moderate CD, we propose the following treatment algorithm for the selection of maintenance therapy, taking into account the clinical features of the patient including disease location, co-morbidities, and previous response to medical treatment (Figure 3).

Figure 3.  Best evidence-based approach for maintenance therapy in patients with mild to moderate Crohn’s disease (CD).

For patients with mild to moderate CD of the ileum and/or proximal colon induced into remission with budesonide, options for maintenance therapy include cessation of therapy and monitoring for recurrent symptoms or continuation of budesonide 6 mg/day for up to 3 months. There are non-placebo-controlled data regarding the ability to maintain remissions for up to 2 years with flexible dosing of budesonide between 6 and 9 mg/day.¹⁰⁸ There are no controlled-trial data regarding the potential for immunosuppressive therapy as maintenance after budesonide; however, azathioprine, mercaptopurine and methotrexate have been effective after conventional steroids.

For patients with mild to moderate CD restricted to the colon and induced into remission with sulfasalazine, there are no controlled trial data to make recommendations for maintenance therapy. Thus, options include cessation of therapy and monitoring for recurrent symptoms and possibly immunotherapy (noting that there are no controlled-trial data regarding the potential for immunosuppressives as a maintenance therapy following sulfasalazine induction therapy). Maintenance therapy with sulfasalazine following sulfasalazine induction has never been evaluated in controlled trials. Whether the outcome of such a trial would be more favourable than the negative results obtained with sulfasalazine for maintenance of medically induced remission is unknown.

For patients with moderate CD induced into remission with conventional steroids, controlled trials have demonstrated that azathioprine, mercaptopurine and methotrexate can maintain remission for 1 year. Prolongation of azathioprine maintenance therapy has been shown to be efficacious for greater than 4 years.¹¹²

Thus, the pros and cons of these treatment options should be discussed with each individual patient based upon their disease course, anticipated risk of recurrence and disease complications, as well as the potential side effects from immunosuppressive therapy and the individual’s history of adherence to therapy and willingness to pursue therapeutic monitoring.

Conclusions

Controlled clinical trials have demonstrated budesonide, sulfasalazine and conventional corticosteroids to be effective medical therapies for mild to moderately active CD. In patients with mild to moderate CD of the ileum and/or proximal colon, budesonide 9 mg/day administered for 8–16 weeks has been shown to be more effective at inducing clinical remissions than placebo or mesalazine and is equally efficacious and safer than conventional corticosteroids. For patients with mild to moderate CD confined to the colon, sulfasalazine 3–6 g/day for up to 16 weeks has been shown to be more effective than placebo, but less effective than conventional corticosteroids. Patients who initiate treatment with budesonide or sulfasalazine, and fail to respond, should be subsequently treated with prednisone 40–60 mg/day followed by a tapering schedule, or infliximab. For patients with moderate CD, or those with systemic symptoms, evidence suggests treatment with prednisone 40–60 mg/day, followed by a tapering schedule after 2–4 weeks.

With regard to maintenance therapy after induction therapy for mild to moderate CD, the evidence is less robust and leaves profound gaps regarding prognostication and therapeutic approaches. Evidence for aminosalicylates therapy is insufficient to recommend either sulfasalazine or mesalazine as a proven maintenance therapy and conventional corticosteroids are either ineffective or too toxic to recommend as maintenance agents. Patients with mild to moderate CD of the ileum and/or proximal colon who have been induced into remission with budesonide can be maintained on budesonide 6 mg/day for 3–6 months. Immunosuppressives have not been evaluated as maintenance agents after budesonide therapy but are effective after treatment with conventional corticosteroids.

Given that mild to moderate CD accounts for the largest proportion of patients, half of whom may never require therapy with conventional corticosteroids, additional controlled trials are needed to fill the ‘gaps’ necessary to provide evidence-based recommendations in the multiple clinical scenarios of even these most common presentations of CD.

Acknowledgements

Declaration of personal interests: W. J. Sandborn has served as a consultant for Centocor, Prometheus, Proctor and Gamble, and Shire, and has received research funding from Centocor, Proctor and Gamble, and Shire. B. G. Feagan has served as a speaker for AstraZeneca and a consultant for Abbott, Bristol-Meyers Squibb, Celgene, Celltech, Centocor, Combinatorx, Elan/Biogen, ISIS, Janssen-Ortho, Millenium, Napo Pharma, Osiris, Proctor and Gamble, Protein Design Labs, Santarus, Schering Canada, Schering Plough, Serono, Synta, Teva Pharmaceuticals, and UCB Pharma. He has also been an advisory board member for AstraZeneca, Celltech, Elan/Biogen, Protein Design Labs, Schering Canada, and Synta and has received research funding from Abbot, Berlex, Boehringer Engelheim, Centocor, Millenium, Novartis, Otsuka, Protein Design Labs, Schering-Plough, Synta, Tillotts. G. R. Lichtenstein has served as a speaker for AstraZeneca, Axean, Centocor, Faro, Proctor and Gamble, Salix, Schering-Plough, Shire, and Solvay and a consultant for AstraZeneca, Axean, Celltech, Centocor, Elan, Faro, Human Genome Sciences, Proctor and Gamble, Prometheus, Protein Design Labs, Protomed Scientific, Salix Pharmaceuticals, Schering-Plough, Serono, Shire, SmithKline Beecham, Solvay, Synta, UCB, and Wyeth. He has also received an honorarium from Falk and research funding from Abbott, Berlex, Celgene, Celltech, Centocor, Elan, Genetics Institute, Human Genome Sciences, Inkine, Intesco, ISIS, Millenium, Otsuka, Protein Design Labs, Protomed Scientific, and Salix.

Declaration of funding interests: Editorial support was provided by Stella Y. Chow of Complete Healthcare Communications, Inc. (Chadds Ford, PA, USA) and funded by Prometheus Laboratories, Inc.