Mild to moderate Crohn's disease – defining the basis for a new treatment algorithm


Dr W. J. Sandborn, Division of Gastroenterology and Hepatology, Mayo Clinic, 200 First St. SW, Rochester, MN 55905, USA.


Previously, clinicians have had few choices in treating mild to moderate Crohn's disease. They currently treat these Crohn's disease patients with oral mesalamine and antibiotics. This treatment approach is based on the safety of these agents, and the perception that they are effective. This perception regarding efficacy may be influenced by publication bias.

This review examines the efficacy and safety data of the conventional corticosteroids, mesalamine, sulfasalazine, budesonide and antibiotics for inducing the remission of mild to moderate Crohn's disease from randomized controlled trials, and proposes an evidence-based treatment approach.

Sulfasalazine has demonstrated modest efficacy when Crohn's disease is confined to the colon. Mesalamine has no clear benefit over placebo in treating active Crohn's disease. Conventional corticosteroids effectively induce remission but are associated with unwanted adverse effects. Budesonide has similar efficacy to conventional steroids with far fewer adverse effects. Antibiotics have not consistently demonstrated efficacy.

We propose a new evidence-based approach which suggests inducing remission of mild to moderate Crohn's disease with budesonide 9 mg/day for patients with ileal and/or right colonic involvement; sulfasalazine for those with disease limited to the colon; and conventional steroids for high disease activity, those who failed budesonide and those with left-sided disease who are allergic or intolerant to sulfasalazine.


Despite medical and surgical advances, symptomatic relapses of Crohn's disease occur frequently. Periods of remission are punctuated by symptomatic flares of disease that can negatively affect a patient's quality of life. The goals of therapy are to control symptoms rapidly and effectively with an agent that maximizes remission rates and minimizes adverse events due to treatment.

Factors that influence the choice of therapy in Crohn's disease are multifactorial. Disease severity and location, the presence of complications, and response to previous treatment should be considered. In addition, the therapeutic efficacy of an individual agent must be weighed against issues of safety and tolerability.

For patients with mild to moderate Crohn's disease, defined by the American College of Gastroenterology 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’,1 sulfasalazine and systemic steroids were the mainstays of therapy during the 1970s and 1980s. Although systemic corticosteroids have a faster onset of action and demonstrated greater efficacy in inducing remission than sulfasalazine, many physicians avoided or at least postponed the use of corticosteroids in many patients with mild to moderate Crohn's disease because of their adverse event profile.2, 3

Instead, physicians used sulfasalazine as the drug of first choice, even though it was less effective than corticosteroids, in order to avoid corticosteroid side-effects. In the 1990s, 5-aminosalicylate in the form of oral mesalamine received regulatory approval for the treatment of ulcerative colitis. Most clinicians believe that because it is comparable to sulfasalazine in inducing and maintaining remission of ulcerative colitis, mesalamine should also be efficacious and better tolerated in patients with mild to moderate Crohn's disease. For this reason, in the 1990s the first-line treatment for mild to moderate Crohn's disease changed from sulfasalazine to oral mesalamine. However, many physicians who prescribe oral mesalamine for Crohn's disease are unaware of the randomized controlled trials in Crohn's disease patients comparing it with placebo or with budesonide.

Budesonide capsules, an enteric-coated (pH- and time-dependent release) formulation of budesonide, were developed to bring the efficacy of systemic steroids in inducing remission of Crohn's disease, with a much-improved safety and tolerability profile.

The purpose of this article is to examine the available data from randomized, controlled clinical trials with regard to the efficacy and safety of the conventional corticosteroids, oral mesalamine and sulfasalazine, budesonide capsules and antibiotics that have been used to induce remission of mild to moderate Crohn's disease, and to propose a novel, evidence-based treatment algorithm. This algorithm is intended for use in patients who are treatment naïve, as well as those who have been previously treated but are currently untreated, and those who have relapsed while on therapies that are modestly effective or ineffective. Our goal is to devise a treatment algorithm that leads to the induction of remission of mild to moderate Crohn's disease with agents that offer optimal efficacy and safety.

Strategy for identification of controlled clinical trials

A search of the online bibliographic databases MEDLINE (1966-March 2003) and Current Contents/Science Edition (1996-March 2003) was carried out to identify potentially relevant English language articles. The Medical Subject Heading (MeSH) terms ‘sulfasalazine’ or ‘sulfasalazine’ or ‘salicylazosulfapyridine’ or ‘salicylazosulfapyridine’ or ‘5-aminosalicylate’ or ‘5-aminosalicylic acid’ or ‘mesalamine’ or ‘mesalazine’ or ‘corticosteroids’ or ‘prednisone’ or ‘prednisolone’ or ‘6-methylprednisolone’ or ‘budesonide’ or ‘antibiotics’ or ‘metronidazole’ or ‘ciprofloxacin’ and ‘Crohn's disease therapy' were used to perform keyword searches of the database. Manual searches of the reference lists from the potentially relevant papers and the proceedings from annual American Gastroenterological Association and American College of Gastroenterology meetings from 1990 to 2003 were performed in order to identify additional studies that may have been missed using the computer-assisted search strategy. Studies selected were randomized trials of patients with mild to moderate Crohn's disease.


Efficacy vs. placebo and conventional corticosteroids

Sulfasalazine (SASP) has been compared with placebo and with conventional steroids in several controlled clinical trials (Table 1).

Table 1.  Randomized, controlled, double-blind trials of SASP
ReferenceDrugs comparedNo. of patientsDurationResponse
Summers et al. 19792SASP 1 g/15 kg
Prednisone 0.25–0.75 mg/kg
as per CDAI
Azathioprine 2.5 mg/kg
17 weeksRemission:
SASP: 43%
Prednisone: 60%
Azathioprine: 40%
Placebo: 30%
Malchow et al. 19843SASP 3 g
48 mg/day, weekly
tapering to 8 mg/day
6-Methylprednisolone +  SASP
Active disease
16 weeksRemission:
SASP: 50%
MP: 83%
MP  +  SASP: 76%
Placebo: 34%
Anthonisen et al.
SASP 1.5 g/day 1st 3 days,
then 3 g/day
 314 months (four
1-month crossovers)
Overall clinical condition:
SASP greater therapeutic than placebo among
those with no previous
van Hees et al. 19817SASP 4–6 g/day
26 weeksDecrease in inflammatory
activity index > 25%:
SASP: 62%
Placebo: 8%
Singleton et al. 19794SASP plus prednisone
8 weeksOverall results:
SASP added to
prednisone not better
than prednisone alone
Rijk et al. 19915
SASP 6 g/day (4 g/day if adverse effects occurred)
plus 30 mg/day
prednisone tapered
SASP 6 g/day (4 g/day if adverse effects occurred)
plus placebo
16 weeksOverall results:
Significantly greater decrease in Crohn's Disease
Activity Index (CDAI)
at all time points with
SASP plus prednisone
vs. SASP alone

In two large studies, SASP was compared with placebo and conventional corticosteroids.2, 3 In one study, Summers et al.2 compared the efficacy of SASP, prednisone and azathioprine with placebo, and showed that prednisone (0.25–0.75 mg/kg adjusted to disease activity) was superior to SASP (1 g/15 kg), and that SASP, but not azathioprine (2.5 mg/kg), was superior to placebo, in inducing clinical remission of Crohn's disease. Sub-group analyses suggested that patients who had been treated with steroids previously failed to respond to SASP, while those who had not taken steroids at entry responded to SASP significantly better than placebo. Patients with involvement limited to the colon responded to SASP better than to placebo, while those who had disease located only in small bowel were less likely to benefit. In a similar study, Malchow et al. randomized patients to treatment with 6-methylprednisolone (48 mg/day weekly, tapering to 8 mg/day) alone, 6-methylprednisolone in combination with SASP (3 g), SASP alone, or placebo.3 6-Methylprednisolone proved the most effective therapy for inducing remission; in sub-group analyses it was significantly more effective than SASP for patients who had been previously treated with steroids, for those who had disease only in the small bowel, and for those with disease in the small bowel and colon. The combination of the steroids plus SASP was most effective in previously untreated patients and when the disease was located in the colon. SASP alone was the least effective active treatment regimen.

Other controlled trials in patients with active Crohn's disease have confirmed some of the conclusions of these sub-group analyses by demonstrating that SASP is not an effective adjunctive therapy to prednisone,4 whereas prednisolone is an effective adjunctive therapy to SASP.5 This is likely because prednisone and prednisolone are more effective therapies than SASP.

In a small, 4 week study, Anthonisen et al.6 showed that SASP (1.5 g/day for 3 days, then 3 g/day) was superior to placebo in patients who had not undergone previous surgery but not in those who had had previous resection. And in another small, comparative study of SASP (4–6 g/day) with placebo, van Hees et al. showed that 62% of patients treated with SASP had a favourable response after 26 weeks compared with 8% of placebo-treated patients.7

Tolerability of sulfasalazine

Adverse events (see Table 4 later) occur in 10–30% of patients with inflammatory bowel disease treated with SASP.8, 9 The majority of these side-effects can be attributed to the systemic absorption of sulfapyridine, and they occur more commonly in patients who are genetically predisposed to ‘slow’ acetylation of sulfapyridine to N-acetylsulfapyridine in the liver.10 Headache, nausea, vomiting and epigastric pain often appear to be dose-related, and it is frequently possible to desensitize patients by discontinuing SASP for 1–2 weeks, and then restarting at 0.125–0.25 g/day and increasing by 0.125 g/week up to a maintenance dose of 2 g/day.9

Table 4.  Side-effects with sulfasalazine, mesalamine and conventional corticosteroids
SulfasalazineHeadache, epigastric pain, nausea, vomiting, cyanosis, skin rash, fever, hepatitis,
 autoimmune haemolysis, aplastic anaemia, leukopenia, agranulocytosis, folate deficiency,
 pancreatitis, systemic lupus erythematosis, sulfonamide-induced toxic epidermal necrolysis,
 Stevens–Johnson syndrome, pulmonary dysfunction, male infertility9, 10
MesalamineHeadache, nausea, dyspepsia and rarely: pulmonary toxicity, pericarditis, hepatitis,
 pancreatis,62–65 interstitial nephritis,66–69 hypersensitivity reactions3
Conventional corticosteroidsMoon face, acne, bodyweight gain, infection, hypertension, hirsutism, petechial bleeding, striae,
 hyperglycemia, myopathy, cataracts, hypokalemia, growth failure in children


Efficacy vs. placebo

Six randomized, double-blind, placebo-controlled studies have been conducted using varying doses of mesalamine in patients with mild to moderately active Crohn's disease (Table 2).11–16 Two studies investigated a 1.5-g daily dose of oral mesalamine (equivalent to 3.0 g/day of SASP).11, 12 A study by Mahida et al.12 investigated treatment with mesalamine 1.5 g/day or placebo for 6 weeks; Rasmussen et al.11 studied the same 1.5 g/day treatment regimen for 16 weeks. Both studies showed no significant differences between mesalamine and placebo treatment groups for inducing remission of Crohn's disease.

Table 2.  Randomized, double-blind, placebo-controlled trials of mesalamine
ReferenceDrug comparedNo. of patientsDurationResponse
Rasmussen et al. 198711Mesalamine 1.5 g
16 weeksImprovement:
Mesalamine 40%
Placebo 30%, ns
Mahida et al. 199012Mesalamine 1.5 g (Pentasa)
6 weeksImprovement:
Mesalamine 40%
Placebo 35%, ns
Tremaine et al. 199415Mesalamine 3.2 g
(Asacol, 0.8 g tablets
4 times daily)
16 weeksPartial or complete
Mesalamine 45%
Placebo 22%, P = 0.042
Singleton et al. 199313Mesalamine 1 g
  2 g
  4 g
16 weeksRemission:
Mesalamine 4 g 43%,
P < 0.01 vs. placebo
Mesalamine 2 g 24%, ns
Placebo 18%
Singleton et al. 199414
Hanauer et al. 200116
Mesalamine 2 g,
  4 g
16 weeksRemission:
No significant change in
CDAI from baseline for any
Change in CDAI from
baseline to end-point:
Mesalamine 4 g: − 41
Placebo: − 35, ns
Hanauer et al. 200116Mesalamine 4 g
16 weeksChange in CDAI from
baseline to end-point:
Mesalamine: − 72
Placebo: − 64, ns

Tremaine et al.15 compared treatment with a 3.2-g mesalamine dose for 16 weeks with placebo in 38 patients with mild to moderate Crohn's disease and found that a significantly greater number of mesalamine-treated patients (n = 9) vs. placebo-treated patients (n = 4) had partial remission [defined as either a decrease in Crohn's disease activity index (CDAI) score ≥ 70 points or complete remission (CDAI score < 150 points)]. However, this study demonstrated no benefit for mesalamine if a conventional definition of remission (CDAI < 150 points) is utilized. Moreover, it is not stated in the publication whether the partial remission end-point was prespecified as the primary end-point, or was a secondary end-point.

Singleton et al.13 compared three doses of mesalamine (1, 2 or 4 g daily) with placebo in a 16-week trial and found a significant difference between the 4 g/day mesalamine group and placebo. This same group then conducted a very similar second trial using the same formulation of mesalamine, and randomized patients to either 2 or 4 g mesalamine or placebo.14 The authors remarked that even though the reductions in CDAI in the mesalamine-treated patients were similar in the second trial to those in the first, these did not differ significantly from placebo.

More details about the second Singleton study, as well as an additional trial which compared the 4 g/day dose of mesalamine with placebo, have become available.16 The results in both of the additional trials showed no significant difference between mesalamine 4 g/day and placebo in remission rates. These two double-blind, placebo-controlled trials, which were not published in full, were used to compile a meta-analysis which has been published as an abstract16 consisting of these two trials and the initial Singleton et al. study.13 In study no. 1 (the second Singleton study), patients were randomized to receive 2 or 4 g mesalamine daily or placebo;14 after 16 weeks there was no significant change from baseline in the mean CDAI score for any of the treatment groups (change in CDAI score was the primary end-point, results of the secondary end-point, remission, not stated). The second study (which has never previously been reported) had the same design and compared mesalamine 4 g daily (no 2 g dose) with placebo.16 Again, after 16 weeks the mean change from baseline in CDAI was not significant for any of the treatment groups. The overall analysis of data from the three trials (original Singleton study, second Singleton study and third previously unpublished study) showed an 18-point difference in the improvement in mean CDAI scores in favour of the 4 g mesalamine group that was statistically significant; however, this difference is not clinically meaningful.

Efficacy vs. systemic and nonsystemic steroids

Mesalamine at doses of 2–4.5 g/day has been compared with conventional corticosteroids and with the nonsystemic steroid, budesonide, in the treatment of active mild to moderate Crohn's disease (Table 3).

Table 3.  Randomized, controlled trials of mesalamine vs. systemic corticosteroids and budesonide
ReferenceDrug comparedNo. of patientsDurationResponse
Thomsen et al. 199817Mesalamine 4 g
Budesonide 9 mg
16 weeksRemission:
Mesalamine: 45%
Budesonide: 69%
P = 0.001 at 8 weeks
Scholmerich et al. 199018Mesalamine 2 g
48 mg/day, weekly tapering
to 8 mg/day
24 weeksChange in CDAI:
Mesalamine: − 58
6-Methylprednisolone: − 151,
P < 0.001
Martin et al. 199019Mesalamine 3 g
Prednisone 40 mg/day for
2 weeks, weekly tapering
to 20 mg/day
12 weeksRemission:
Mesalamine: 47%
Prednisone: 46%, ns
Gross et al. 199520Mesalamine 4.5 g
48 mg/day, weekly tapering
8 weeksRemission:
Mesalamine: 40%
6-Methylprednisolone: 56%, ns
Prantera et al. 199921Mesalamine tablets or
microganules 4 g/day each
40 mg/day
12 weeksRemission:
Mesalamine tablets: 60%
Mesalamine granules: 79%
6-Methylprednisolone: 61%,
ns between groups

One large, randomized, double-blind study compared treatment with budesonide 9 mg daily or mesalamine 4 g daily for 16 weeks.17 At 8, 12 and 16 weeks budesonide was significantly more effective than mesalamine at inducing remission of Crohn's disease.

A number of small trials and one large trial have compared mesalamine with systemic corticosteroids. In the largest study, the German 5-ASA Study group compared a 2-g daily dose of mesalamine with 6-methylprednisolone (48 mg/day tapering to 8 mg)18 and found that systemic steroid treatment was significantly better than mesalamine at inducing remission of Crohn's disease. In two very small, double-blind studies, where mesalamine, at doses of 3 g and 4.5 g daily, was compared with prednisone (40 mg/day tapering to 20 mg)19 and 6-methylprednisolone (48 mg/day tapering to 8 mg),20 respectively, rates of remission were not significantly different between mesalamine and steroid treatments at study end-point. These small studies are underpowered noninferiority (equivalence) studies, and therefore from the statistical standpoint they are not interpretable. In another small, comparative study, no significant difference was found among remission rates in patients receiving mesalamine tablets (4 g/day), mesalamine microgranules (4 g/day) or 6-methylprednisolone (40 mg/day) for 12 weeks. Again, the study lacked sufficient statistical power to detect clinically meaningful differences between the three treatments.21

Tolerability of mesalamine

In general, adverse events occur infrequently in patients treated with mesalamine. Common and rare side-effects of mesalamine treatment are listed in Table 4.

In the studies evaluated,11–16 the majority of side-effects seen with mesalamine were mild to moderate in intensity and similar to placebo. In the comparison of mesalamine with budesonide,17 while the incidence of adverse events was similar, the number of severe adverse events was significantly greater with mesalamine treatment than with budesonide [22% vs. 12% (P = 0.04)] due to worsening of Crohn's disease in the mesalamine-treated patients.

Systemic corticosteroids

Prednisone and prednisolone, which act systemically to deliver anti-inflammatory effects, have demonstrated efficacy that is superior to placebo, sulfasalazine and antibiotics in inducing remission of Crohn's disease (Table 5). However, because they have systemic effects, their use has been associated with an increased incidence of a number of adverse events.

Table 5.  Comparative efficacy of systemic steroids
StudyNSystemic steroidComparatorOutcome
Summers et al. 19792295Prednisone 0.25–0.75 mg/kg3–5 g/day SASP or
 placebo or
Prednisone significantly
 superior efficacy to SASP,
 placebo and azathioprine
 No additional improvement
 with adjunctive SASP
 Additive benefit when
 prednisone added to SASP
Malchow et al. 198432156-Methyl prednisolone 48 mg/day3–5 g/day SASP or
Prednisolone superior
 to SASP and placebo
 No additional improvement
 with adjunctive SASP
Rijk et al. 1991560Prednisone 30 mg/day tapered6 g/day SASPAdditive benefit when
 prednisone added to SASP
Prantera et al. 19964241Prednisone 30 mg/day1 g/day ciprofloxacin
 1 g/day metronidazole
Prednisone significantly
 greater efficacy than antibiotics
Prantera et al. 199921946-methylprednisolone 40 mg/dayMesalamine
 4 g/day capsules
 of mesalamine
Similar efficacy both
 treatment groups
 4 g/day micogranules
Rutgeerts et al. 199427176Prednisone 40 mg/day tapered9 mg/day budesonide
Similar efficacy both
 treatment groups
Campieri et al. 199724177Prednisolone 40 mg tapered9 mg/day budesonide
Similar efficacy both
 treatment groups
Gross et al. 199626676-Methylprednisolone 48 mg/day
 9 mg/day
Similar efficacy both
 treatment group
Bar-Meir et al. 199829201Prednisone 40 mg/day
 9 mg/day
Similar efficacy both
 treatment groups

Tolerability of corticosteroids

The toxicity of corticosteroids can be significant and varies depending on route of administration, dosage, duration of therapy, gender and individual sensitivities to corticosteroids.22 Common adverse (Table 4) events include moon face, mood changes, insomnia, gastrointestinal intolerance, oedema, weight gain and elevated blood glucose levels.23


Once it was demonstrated that corticosteroids were effective in inducing remission of Crohn's disease, oral drug delivery systems were developed to provide the efficacy of a corticosteroid without the associated corticosteroid toxicity. Budesonide capsules (ENTOCORT EC), a product of this effort, have been evaluated for induction of remission of active mild to moderate Crohn's disease in several randomized, controlled, double-blind, multicentre trials (Table 6).17, 24–29

Table 6.  Randomized, controlled, double-blind trials of budesonide
ReferenceDrug comparedNo. of patientsDurationResponse
  1. BUD,= budesonide; *P < 0.01 vs. placebo.

Greenberg et al. 199425BUD capsules 3 mg/day
   9 mg/day
15 mg/day
10 weeksRemission:
BUD 15 mg 43%*
BUD 9 mg 51%*
BUD 3 mg 33%, ns
Placebo 20%
Tremaine et al. 200228BUD capsules 9 mg q.d.s.
   4.5 mg b.d.
8 weeksRemission:
BUD 9 mg/day 50%
Placebo 33%, ns
Change in CDAI:
BUD 9 mg/day
significantly decreased
vs. placebo
Thomsen et al. 199817BUD capsules 9 mg/day
Mesalamine 4 g/day
16 weeksRemission:
BUD 9 mg 62%
Mesalamine 36%, P < 0.001
Rutgeerts et al. 199427BUD capsules 9 mg/day
1st 8 weeks, then 6 mg/day
Prednisolone 40 mg/day
1st 2 weeks, then tapered
to 5 mg/day by week 10
10 weeksRemission:
BUD capsules 53%
Prednisolone 66%, ns
Campieri et al. 199724BUD capsules 9 mg q.d.s.
4.5 mg b.d.
Prednisolone 40 mg/day,
tapered to 5 mg/day at 9 weeks
8 weeksRemission:
BUD 9 mg 60%
BUD 4.5 mg b.d. 42%
Prednisolone 60%,
ns between groups
Gross et al. 199626Budenofalk 9 mg/day
48 mg/day tapered
weekly to 8 mg/day
8 weeksRemission:
BUD 56%
73%, ns
Bar-Meir et al. 199829Budenofalk 9 mg/day
Prednisone 40 mg/day
1st 2 weeks, tapered to
5 mg/day at end-point
8 weeksRemission:
BUD 51%
Prednisone 53%, ns

Budesonide vs. placebo

An 8-week study by Greenberg et al.25 compared 3, 9 and 15 mg budesonide daily with placebo and showed that remission rates for patients treated with both the 9 and 15 mg dose of budesonide were significantly superior to placebo. Another 8-week trial failed to demonstrate that treatment with budesonide 9 mg once daily or 4.5 mg twice daily was significantly better than placebo.28

In a recently published meta-analysis of budesonide studies, Kane et al.30 found that remission was induced significantly more frequently by budesonide than placebo, with an odds ratio of 1.85. Thus, a patient is significantly more likely to go into remission when treated with budesonide 9 mg daily than with placebo.

Budesonide vs. mesalamine

As previously described, one trial has compared budesonide 9 mg daily with mesalamine 4 g daily in inducing remission of Crohn's disease and showed superior efficacy of budesonide compared with mesalamine at all time points evaluated.17 In their meta-analysis, Kane et al. reported that budesonide was more likely to induce remission than mesalamine, with a calculated odds ratio of 1.73.30 Accordingly, a patient is 73% more likely to achieve remission with budesonide than mesalamine.

Budesonide vs. conventional corticosteroids

Trials by Rutgeerts et al.27 and Campieri et al.24 showed that budesonide has similar efficacy to prednisolone with regard to inducing remission of Crohn's disease. Two other double-blind studies compared an oral pH-modified release (pH > 6.4) formulation of budesonide 9 mg/day (Budenofalk, which is not approved in the United States) with prednisone (40 mg tapered to 5 mg/day)29 or 6-methylprednisolone (48 mg/day tapered to 8 mg/day).26 No significant differences in remission rates were seen between the groups after 8 weeks of treatment in these two studies (Table 6). Although these three trials suggested that budesonide formulations have equivalent efficacy to conventional corticosteroid therapy, they were also not designed as formal non-inferiority studies. To further quantify the relative efficacy after the alternative treatment approaches Kane et al. performed a meta-analysis of these findings.30 These authors concluded that conventional corticosteroids induced remission of mild to moderate Crohn's disease more frequently than budesonide (with an odds ratio of 0.87). In sub-group analysis, when patients had a lower CDAI (200–300) remission rates were not clinically different.

Tolerability of budesonide

The side-effects frequently seen with conventional corticosteroid use are minimized in Crohn's disease patients treated with budesonide 9 mg/day due to its high first-pass metabolism and low systemic absorption (10–15%) (Table 7).

Table 7.  Relative rates of occurrence of adverse events with budesonide capsules 9 mg/day in controlled clinical trials
Side-effectvs. placebovs. mesalaminevs. corticosteroids
 side-effects, i.e. acne,
 ankle oedema,
 buffalo hump hirsutism
 and moon face
Similar rate,25, 28
 except moon face
 (7% budesonide vs. 2%
 placebo P = 0.001)25, 28
Incidence and types of all
 AEs similar,17
Significantly higher in
 prednisone group vs. BUD
 (P = 0.003).24, 26, 27, 29
Significantly higher
 incidence of moon face
 (P = 0.0005).24, 26, 27, 29
Significantly higher in
 6-methylprednisolone group
 vs. BUD (P = 0.0015).29
Significantly higher in the
 prednisone group vs. BUD
 (P = 0.0018).26
HPA axis suppressionSimilar rate.25, 28
 Significantly higher in
 9 mg/day budesonide
 group (P = 0.006),25, 28
10% budesonide
 0% mesalamine (P = 0.02)17
Significantly lower
 (P = 0.0022426, 27, 29)
Decreased bone mineral
(not measured)(not measured)Significantly less
 depression of bone formation
 with budesonide than
 methylprednisolone at 10 weeks.70
General adverse eventsSimilar proportion of
 patients withdrew due
 to AEs (5% vs. 4.5%)25,28
 7.5% vs. 7.5%25, 28
Incidence and types of AEs
 similar, except mesalamine
 group had significantly
 higher incidence
 of severe AEs (P = 0.04)17
Similar proportion of
 patients withdrew due to AEs

In their meta-analysis, Kane et al. found that budesonide capsules had a similar tolerability profile to placebo and mesalamine, and a greatly improved profile compared with prednisolone.30 In an ongoing, uncontrolled, compassionate use study of more than 4000 Crohn's disease patients treated with doses of up to 21 mg budesonide per day, some for more than 5 years, serious adverse events were experienced by 3% of patients; most of these were gastrointestinal and unrelated to treatment.31



Few randomized, double-blind, controlled studies have evaluated either metronidazole or ciprofloxacin for induction of remission of mild to moderate Crohn's disease. Most trials in the literature studied a small number of patients (Table 8). In one of the larger studies, Sutherland et al.32 found that 20 mg/kg of metronidazole was more effective than placebo for reducing the mean CDAI score but not for inducing remission. In sub-group analysis, the beneficial effect of metronidazole was limited to patients with ileocolonic or colonic disease. No effect was demonstrated in patients with disease located only in the ileum.

Table 8.  Randomized controlled trials of metronidazole and ciprofloxacin
ReferenceDrugs comparedNo. of patientsDurationResponse
Sutherland et al. 199132Metronidazole 10 mg/kg
  20 mg/kg
30 36
16 weeks
Change in CDAI from baseline
to end-point:
Metronidazole significantly better
than placebo in patients with
disease confined to large intestine
or affecting both small and large
intestine. Not significantly better
in patients with small bowel
disease only
Allan and Cooke 197733Metronidazole 20 mg/kg
116 monthsNo benefit of metronidazole
Ambrose et al. 198534Metronidazole 800 mg/day
Metronidazole  +  cotrimoxazole
2 monthsSymptomatic improvement
(clinical symptoms as well
as laboratory values):
No benefit of metronidazole
Blichfeldt et al. 197835Metronidazole 1.0 g/day
224 months (two
 crossover periods)
Overall clinical condition
(objective and subjective criteria
as well as laboratory values):
No benefit of treatment except in
n = 6 patients with colonic disease only
Ursing et al. 198236
Rosen et al. 198237
Metronidazole 800 mg/day
SASP 3 g/day
788 months (two
 crossover periods)
Change in CDAI and
orosomucoid level from baseline
to end-point:
No treatment benefit 1st 4 months.
Second 4 months, 15 patients on
SASP and switched to
metronidazole showed significant
decrease in CDAI.
No improvement for patients on
metronidazole and switched to
Arnold et al. 200239Ciprofloxacin 1 g/day
6 monthsRemission:
Ciprofloxacin: 76%
Placebo: 25% (P < 0.001)
Steinhart et al. 200240Ciprofloxacin  +  metronidazole
(both 1 g/day)
(All patients on budesonide
9 mg/day)
8 weeksRemission:
Antibiotics: 33%
Placebo: 38%, ns
Remission in patients with disease
of the colon:
Antibiotics: 53% (9/17)
Placebo: 25% (4/16), ns
Colombel et al. 199941Ciprofloxacin 1 g/day
Mesalamine 4 g/day
6 weeksRemission:
Ciprofloxacin: 56%
Mesalamine: 55%, ns
Prantera et al. 199642Ciprofloxacin (500 mg b.d.)
and metronidazole (250 mg
four times daily)
(0.7–1 mg/kg/day tapering to
40 mg, followed by tapering of 4 mg weekly)
12 weeksRemission:
Antibiotics: 46%
Methylprednisolone: 63%, ns

In three other smaller placebo-controlled studies,33–35 metronidazole (0.8–1 g/day) showed no treatment benefit over placebo, except in a sub-group analysis in one study35 for patients who had isolated colonic disease.

A small randomized, double-blind, crossover cooperative trial in Sweden showed no significant difference in response between patients treated with SASP (1.5 g b.d.) or metronidazole (0.4 g b.d.).36, 37 In sub-group analyses, the investigators found that patients with disease in the small intestine only reacted less favourably than did patients with disease located in the colon, with or without concomitant small bowel lesions, regardless of the drug given. These results should be interpreted with caution given the scientific limitations of crossover studies for evaluation of chronic diseases.38


In a recently published trial, ciprofloxacin was shown to be effective when added to existing treatment regimens in moderately active, but therapy-resistant Crohn's disease patients.39 However, in another larger placebo-controlled study by Steinhart et al. the authors found that the addition of both ciprofloxacin and metronidazole to budesonide did not provide any additional adjunctive therapeutic benefit to budesonide monotherapy in patients with active mild to moderate Crohn's disease.40

Two small active, nonblinded drug comparison studies showed no benefit of ciprofloxacin over its comparators.41, 42 In a study of 40 patients, Colombel et al.41 compared ciprofloxacin with mesalamine and showed no difference between treatments. Prantera et al.42 found higher rates of clinical remission in patients treated with 6-methlyprednisolone (0.7–1.0 mg/kg tapering of 4 mg weekly) than with the combination of ciprofloxacin (500 mg b.d.) plus metronidazole (250 mg q.d.s.); although this difference was not statistically significant, the study was at high risk of a Type II error (false-negative result).

Tolerability of antibiotics

The most common side-effects of short-term metronidazole treatment include nausea, a metallic taste and disulfuram-like reaction.43 Sensory peripheral neuropathy is also a common side-effect of metronidazole, and in some cases this complication either takes months to disappear or does not completely resolve with cessation of therapy.44 Emergence of resistant micro-organisms when antibiotics are taken long-term is also of concern.45–47


The efficacy of SASP was first reported in ulcerative colitis patients in 1942.48 Its success in treating ulcerative colitis held out the potential that it would be highly effective in inducing remission of Crohn's disease. Large, methodologically rigorous studies have demonstrated that SASP is only modestly effective for inducing remission of mild to moderate Crohn's disease. Furthermore, SASP is clearly inferior to conventional corticosteroids.2, 3 In addition, no significant adjunctive benefit has been shown when SASP is added to steroid treatment,4 most likely because of the superior efficacy of steroids. The greatest benefit of SASP is likely realized when Crohn's disease is confined to the colon. This observation may be explained by the fact that SASP is dependent on colonic bacteria for activation and it is topically acting.2, 3

The active moiety of SASP was found, in ulcerative colitis patients, to be the poorly absorbed 5-aminosalicylate (5-ASA) molecule, and not the well absorbed sulfapyridine portion of the compound.10, 49–51 In addition, the sulfapyradine moiety was determined to be responsible for the majority of side-effects seen with SASP, which occur in a third of patients.10 Based on these observations, oral drug delivery systems were devised to eliminate the requirement for a sulfa moiety, to avoid absorption of 5-ASA (mesalamine) in the proximal small bowel, and to target drug release in the distal small bowel and colon.

Clinicians have enthusiastically accepted the newer mesalamine formulations for the treatment of active Crohn's disease. One study estimates that as many as 75% of physicians use mesalamine as first-line therapy in their mild to moderate Crohn's disease patients (data on file, AstraZeneca, 2002). However, the data generated from clinical trials do not provide strong evidence for this practice. One small placebo-controlled study that compared 3.2 g/day mesalamine with placebo in 40 patients demonstrated efficacy with an unconventional end-point (combined partial and complete remission).15 Another study evaluating 4 g/day mesalamine showed a statistically significant effect of mesalamine in 150 patients.13 However, two large studies performed subsequently trials comparing the 4 g/day of the same mesalamine formulation with placebo in 450 patients found no significant differences between these two treatments.14, 16 Thus, it is not clear from the data available that mesalamine has any efficacy in the treatment of active Crohn's disease.

Consistent with this notion is the observation that mesalamine 4 g/day was inferior to budesonide 9 mg/day for inducing remission of mild to moderate Crohn's disease at all time points during the 16-week Thomsen trial.17 The safety profiles of these two drugs during this study were similar. While some very small comparative trials showed similar efficacy of mesalamine to systemic corticosteroids,19–21 these studies lacked sufficient statistical power to determine noninferiority (equivalence), and are thus not interpretable. Thus, while the safety and tolerability profile of the newer mesalamine formulations is superior to sulfasalazine, the questionable efficacy of these formulations makes it impossible to conclude that they have a better therapeutic index.

The systemic corticosteroids, prednisone and prednisolone, are unequivocally effective for induction of remission of mild to moderate Crohn's disease. Two large, placebo-controlled trials have demonstrated that prednisone administered at 0.25–0.75 mg/kg and a tapering dose of 48 mg/day of 6-methylprednisolone are effective in inducing remission of Crohn's disease.2, 3 However, these systemically acting steroids are associated with a variety of well documented adverse effects including moon face, mood changes, gastrointestinal intolerance, fluid retention and insomnia.

A newer corticosteroid, budesonide, which is primarily a topically acting therapy with reduced systemic effects, allows similar efficacy to prednisone and prednisolone in Crohn's disease patients, with far fewer adverse effects. Budesonide has a high affinity for the glucocorticoid receptor and undergoes extensive first-pass metabolism resulting in decreased systemic efficacy.

Budesonide (9 mg/day) has been demonstrated to be more effective than mesalamine (4 mg/day)17 and placebo in both clinical trials and meta-analysis,25, 30 with no significant differences between treatments in total adverse events or glucocorticosteroids (GCS)-associated adverse events (except moon face in one study). Budesonide 9 mg daily has also been shown to be only slightly less efficacious than standard corticosteroid treatment with prednisone, prednisolone and 6-methylprednisolone, with a significantly improved tolerability and safety profile.24, 26, 27, 29 Thus, the data clearly indicate that budesonide 9 mg once daily is effective in inducing remission of mild to moderate active Crohn's disease when the disease involves the ileum and/or ascending colon.

Although metronidazole and ciprofloxacin are widely used by clinicians for the treatment of active Crohn's disease, controlled trials have not consistently demonstrated efficacy. Sub-group analyses have suggested that patients with colonic disease might benefit.32, 35 However, this observation is based on a small number of studies that evaluated an inadequate number of patients and thus cannot be considered reliable.

Recommendations: defining the basis for a new treatment algorithm

The traditional, nonevidence-based approach to inducing remission of Crohn's disease has been to use step-wise therapy that starts with mesalamine or, less frequently, with sulfasalazine. Antibiotics such as metronidazole or ciprofloxacin are subsequently added in those who fail to respond. Physicians then progress to budesonide and finally systemic steroids if these interventions are unsuccessful.

Since the main goals of therapy in inducing remission of mild to moderate Crohn's disease are to control symptoms and maximize remission while minimizing adverse events, it is important to identify, based on evidence, which agents are most effective (effectiveness demonstrated by two adequate and well controlled trials), and which have the best safety and tolerability profiles. Based on these precepts, a new evidence-based approach to inducing remission of mild to moderate Crohn's disease should be considered (Figure 1). This new treatment algorithm is intended for use in patients who are treatment naïve, as well as those who have been previously treated but are currently untreated, and those who have relapsed while on therapies that are modestly effective or ineffective.

Figure 1.

Evidence-based algorithm for inducing remission of mild to moderate Crohn's disease. *If the patient is not improving, one needs to reclassify the patient as having moderate to severe disease and evaluate for treatment with infliximab, immunomodulators or surgery.

This approach takes into account the fact that neither mesalamine (at doses of up to 4 g/day) nor antibiotics have consistently demonstrated efficacy in inducing remission of Crohn's disease. Thus, mesalamine and antibiotics are not included in the proposed evidence-based treatment algorithm. We also elected not to include total enteral nutrition in the treatment algorithm. Although we recognize that total enteral nutrition is a modestly effective therapy for mild to moderately active Crohn's disease, it is less effective and has a slower onset of action than conventional corticosteroids, and the formulations are expensive and poorly tolerated by patients. The primary indication for total enteral nutrition is growth retardation due to caloric deficit in children and adolescents, and enteral nutrition as a primary therapy plays little role in the management of disease activity in adult patients with active Crohn's disease.52–61

Budesonide 9 mg/day in Crohn's disease patients with ileal and/or right colonic involvement has been shown in clinical trials to consistently demonstrate efficacy compared with both placebo and mesalamine, and should thus be used as first-line treatment for inducing remission of mild to moderate Crohn's disease in patients with ileal and/or right colonic involvement. Budesonide is used at a dose of 9 mg/day for 8–16 weeks and then discontinued over 2–4 weeks by tapering in 3 mg increments.

Sulfasalazine 3–6 g/day also has demonstrated efficacy for this indication, and a sub-group analysis suggests that this benefit is limited to patients with colonic involvement. For this reason, we believe that sulfasalzine 3–6 g/day for up to 16 weeks should be used as the first-line therapy for patients where disease is limited to the colon. Headache, nausea/vomiting and epigastric pain often appear to be related to the sulfasalazine dose, and it is frequently possible to desensitize patients by discontinuing sulfasalazine for 1–2 weeks, and then restarting at 0.125–0.25 g/day and increasing by 0.125 g/week. Complete blood count with differential and liver transaminases should be obtained every 2 weeks for 3 months, then monthly for 3 months, and subsequently every 3 months for the duration of therapy. A urinalysis with microscopic examination and serum creatinine should also be monitored periodically.

Conventional steroids should be reserved for patients with high disease activity, for those with left-sided disease who are allergic or intolerant to sulfasalazine, and for those who have failed treatment with budesonide. When conventional steroids fail, the patient should be reclassified as having severely active or refractory disease, and therapy should then be refocused on immunosuppressive therapy, infliximab or surgery.

In our opinion, clinical practice should shift from the current non-evidence-based treatment algorithm for induction of remission of mild to moderate Crohn's disease that emphasizes first-line therapy with mesalamine and antibiotics to an evidence-based treatment algorithm that recommends budesonide for patients with disease involving the ileum and right colon, sulfasalazine for left colonic disease and prednisone for patients who are more severely ill and those who fail budesonide or sulfasalazine therapy. We speculate that acceptance of this therapeutic plan will greatly improve patient well-being and outcomes and will result in reduced cost of care.

The issue of optimal postinduction therapy needs to be addressed; however, the complexity of this topic warrants discussion in a separate paper. Sulfasalazine, oral mesalamine and low dose prednisone are all ineffective for maintenance of medically induced remission. Budesonide 6 mg prolongs the time to relapse but does not meet the conventional criteria of maintenance of remission at one year. Antibiotics have not been studied for this indication. Infliximab is effective for both induction and maintenance of remission. Azathioprine, 6-mercaptopurine and methotrexate are effective maintenance agents but are not useful as induction agents because of the slow onset of action (2–6 months). The routine use of azathioprine, 6-mercaptopurine, methotrexate and infliximab in all patients is currently not accepted because of the occurrence of rare but important toxicities such as bone marrow suppression, opportunistic infection, non-Hodgkin's lymphoma, and, in the case of methotrexate, hepatotoxicity. Presently, in view of the lack of therapies that are both safe and effective in maintaining remission of Crohn's disease, one long-term strategy is to suspended therapy in mild to moderate patients once remission is achieved and then to restart treatment with the return of symptoms. This topic should be addressed in full, however, in a future evidence-based review of published data on long-term treatment of Crohn's disease.


Dr William Sandborn has consulted for AstraZeneca, Shire Pharmaceuicals & Proctor & Gamble. He has received research support from AstraZeneca & Proctor & Gamble. He has participated in continuing medical education events indirectly sponsored by AstraZeneca, Shire Pharmaceuticals, Falk Pharma and Proctor & Gamble. Dr Brian Feagan has consulted for AstraZeneca and has received research support from AstraZeneca. He has participated in continuing medical education events indirectly sponsored by AstraZeneca & Falk Pharma.