Professor R. Caprilli, Cattedra di Gastroenterologia, Dipartimento di Scienze Cliniche, Università‘La Sapienza’ di Roma, Policlinico ‘Umberto I’, Viale del Policlinico, 155, 00161 Rome, Italy. E-mail: firstname.lastname@example.org
Background : The role of mesalazine in preventing the clinical recurrence of Crohn's disease after surgery has been shown in a meta-analysis of all published studies. No clear relationship, however, has been shown between dosage and response.
Aim : To evaluate whether 4.0 g/day of mesalazine may offer therapeutic advantages over 2.4 g/day in the prevention of both endoscopic and clinical post-operative recurrence of Crohn's disease.
Methods : The study was a double-blind, randomized, multi-centre, prospective, controlled clinical trial. Two hundred and six patients, submitted to first or second intestinal resection for Crohn's disease limited to the terminal ileum, with or without involvement of the caecum/ascending colon, were enrolled. Of these, 101 were randomly allocated to receive 4.0 g/day of mesalazine (Asacol, Giuliani SpA, Milan, Italy) and 105 to receive 2.4 g/day, starting 2 weeks after surgery. The primary outcome was endoscopic recurrence, at 12 months after surgery. Three different degrees of endoscopic recurrence were evaluated (endoscopic scores: > 0, > 1 and > 2). The secondary outcome was clinical recurrence, defined as a Crohn's disease activity index of more than 150 points or an increase in the Crohn's disease activity index of 100 points or more. For statistical analysis, chi-square, Wilcoxon and Cox regression model tests were used, when appropriate.
Results : Eighty-four patients in the 4.0 g/day group and 81 patients in the 2.4 g/day group were evaluable by endoscopy. Endoscopic recurrence of > 0 was significantly higher in the 2.4 g/day group than in the 4.0 g/day group (62% vs. 46%; P < 0.04). No difference was observed between the two groups with regard to the other two endoscopic outcomes (> 1 and > 2) or clinical recurrence.
Conclusions : A 4.0 g/day regimen of mesalazine does not offer a clinically significant advantage over a 2.4 g/day regimen in the prevention of post-operative endoscopic and clinical recurrence of Crohn's disease at 1 year of follow-up.
Several drugs have been tested in an attempt to prevent the post-operative recurrence of Crohn's disease, but mostly with disappointing results.1–14 To date, mesalazine is the most investigated drug in the post-surgical setting. A meta-analysis of five randomized controlled trials, including 729 patients, showed that treatment with mesalazine significantly reduced the risk of clinical recurrence at 1 year after surgery. The pooled estimate of the treatment effect was significant (overall risk difference, − 10%; 2P = 0.0041), and the number of patients needed to treat to prevent one post-operative recurrence was 10. Moreover, it was estimated that a further eight null trials are needed to state that mesalazine is ineffective.7,15,16
The evidence available is therefore sufficient to conclude that the risk of clinical recurrence is significantly reduced by mesalazine in patients operated on for Crohn's disease. The magnitude of the overall effect is small, but clinically relevant, at least at 1 year of follow-up. In the meta-analysis, no clear relationship was observed between dosage and response.
In this paper, we report the results of a large, Italian, multi-centre, randomized controlled trial in which two doses of mesalazine (4.0 g/day or 2.4 g/day of oral Eudragit-S-coated mesalazine; Asacol, Giuliani SpA, Milan, Italy) were compared by intention-to-treat analysis in the prophylaxis of endoscopic and clinical post-operative recurrence of Crohn's disease limited to the terminal ileum, with or without involvement of the caecum/ascending colon.
Materials and methods
Patients operated on for Crohn's disease (first or second intestinal resection) were enrolled in the study by 17 Italian collaborating centres.
The following criteria had to be fulfilled for patients to enter the trial: males/females aged between 18 and 65 years; disease limited to the terminal ileum (lesions not exceeding 1 m), with or without involvement of the caecum/ascending colon, evaluated by colonoscopy and small bowel follow-through within 1 month before surgery; first or second resection, and considered by the surgeon during the operation to be ‘radical’ (complete removal of the macroscopically involved intestinal segment); absence of skip lesions; diagnosis of Crohn's disease confirmed macroscopically and microscopically by standard criteria.
Exclusion criteria were as follows: localization of the disease to the jejunum, proximal ileum, transverse colon, left colon or ano-rectum; small bowel resection exceeding 1 m; known side-effects from sulfasalazine or salicylates; severe diseases unrelated to Crohn's disease (e.g. renal or liver dysfunction); treatment with drugs likely to affect intestinal pH (H2-receptor antagonists, proton pump inhibitors); pregnancy; questionable ability to co-operate; inability to give informed consent according to the Helsinki Declaration.
The study was a double-blind, randomized, multi-centre, prospective, controlled clinical trial performed to evaluate the efficacy of two regimens of oral mesalazine (4.0 g/day or 2.4 g/day) in the prevention of the post-operative recurrence of Crohn's disease.
The patient enrolment period lasted 3 years starting in January 1997. The established follow-up period was 1 year. Eligible patients, after providing written informed consent, were randomly allocated to receive 4.0 g/day or 2.4 g/day of oral Eudragit-S-coated mesalazine (Asacol). Patients in the 4.0 g/day group received five tablets of mesalazine (800 mg) divided into three doses (1 + 2 + 2 tablets), whereas those in the 2.4 g/day group received three tablets of mesalazine (800 mg) divided into three doses (1 + 1 + 1 tablets) plus two tablets of placebo identical in appearance. Patients were randomized in blocks of four according to a computer-generated randomization scheme provided by an independent institution at the beginning of the trial and forwarded to the Department of Clinical Trials at Giuliani SpA. All patients and investigators were blind with regard to treatment allocation. The co-ordinating centre was notified, within 10 days, of the enrolment and start of treatment for each patient. Treatment was started 2 weeks after surgery. No other pharmacological treatment was given, with the exception of anti-diarrhoeal drugs on demand.
Baseline information included the following: diagnosis, duration of the disease, site and extent of the lesions, clinical course, indication for surgery, previous treatments, clinical and laboratory findings, Crohn's disease activity index (CDAI) and risk factors.
Patients were seen for clinical and laboratory assessment 2 weeks after surgery, and then at 6 and 12 months. Laboratory analyses included the evaluation of the erythrocyte sedimentation rate, C-reactive protein, orosomucoids, full blood count, serum iron, proteins, electrolytes, liver and renal function tests and urinalysis. The CDAI was calculated at the same intervals. Colon ileoscopy was performed at 12 months. Clinical, laboratory and endoscopic examinations were brought forward if the recurrence of symptoms was reported before the scheduled follow-up.
The treatment blinding code was broken in June 2000 when all assessments had been completed; no serious adverse events required the breaking of the code before the end of the study. This protocol was approved by the Ethics Committee of the Institutional Board of Clinicians.
The number of patients randomized per centre varied: Florence, 40; Palermo, 40; Rome ‘La Sapienza’, 23; Naples, 21; Padua, 21; San Giovanni Rotondo (FG), Rome ‘San Filippo Neri’ and Forlì, between 10 and 20 patients each. The other centres enrolled < 10 patients each.
The primary outcome parameter was endoscopic recurrence. Recurrence was defined as the presence of typical endoscopic Crohn's disease lesions in the neoterminal ileum and/or anastomosis, and was graded according to the criteria of Rutgeerts et al.:17 0, no lesions; 1, less than five aphthous lesions; 2, more than five aphthous lesions with normal mucosa between the lesions, or skip areas of larger lesions, or lesions confined to the ileo-colonic anastomotic lining (< 1 cm); 3, diffuse aphthous ileitis with diffusely inflamed mucosa; 4, diffuse ileal inflammation with larger ulcers, nodules or narrowing. Hyperaemia and oedema alone were not considered as signs of recurrence. Three different degrees of endoscopic recurrence were evaluated: (i) an endoscopic score of > 0; (ii) an endoscopic score of > 1; and (iii) an endoscopic score of > 2 (severe recurrence). All three different degrees of endoscopic recurrence were evaluated as they were variably used in other studies.1–6,8–12,14,18
The secondary outcome parameter was clinical recurrence, defined as CDAI > 150 points or an increase in CDAI score of ≥ 100 points. The CDAI was calculated from the patients' diary card according to the original criteria proposed by Best et al.19
Patients with severe symptoms, requiring steroids or further surgery, were withdrawn from the study. Patients who stopped treatment for more than two consecutive weeks, those lost to follow-up and those who developed severe side-effects were considered as drop-outs. Adverse events and reported compliance with the drug were recorded at each visit.
Compliance (percentage of the required intake) was evaluated using an electronic drug exposure monitor, which consists of a standard medication bottle with a pressure-activated microprocessor in the cap. The microprocessor records each opening and lists the date and time of the opening. Non-compliance was defined as interruption of drug intake for more than two consecutive weeks, or consumption of less than 80% of the prescribed dose.
Sample size. Assuming that mesalazine, 2.4 g/day, would reduce severe endoscopic recurrence from 70% to 55% at 1 year of follow-up, we hypothesized that mesalazine, 4.0 g/day, would reduce the rate of severe endoscopic recurrence to 30%. We calculated that the number of patients needed to ensure a type I and type II error level of 5% was 85 patients per group plus 25% drop-outs (i.e. a further 43 patients). Therefore, the total number of patients required was 213.
Analysis of data. Outcome measures were analysed in all randomized patients who had taken at least one dose of the study medication (intention-to-treat population). Chi-square and Wilcoxon tests were used, when appropriate. The following variables were analysed at univariate analysis: sex, smoking habit, type of anastomosis (end-to-end, end-to-side, side-to-end, side-to-side), localization of anastomosis (ileo-caecal, ileo-transverse), previous resection, duration of disease, therapy, extension, use of contraceptives and serum albumin levels. A Cox regression model was used if, at univariate analysis, a P value of < 0.10 was obtained.
All analyses were carried out using a computer program (MS-DOS).
Over the 3-year period, 263 patients were screened. Of these, 51 were not enrolled due to violation of inclusion criteria or applicability of exclusion criteria. A further six eligible patients refused to take part in the study. Thus, 206 patients (97% of the eligible population) were randomized, 101 to mesalazine, 4.0 g/day, and 105 to mesalazine, 2.4 g/day (Figure 1). Of these 206 patients, 186 (95 in the 4.0 g/day group and 91 in the 2.4 g/day group) were clinically evaluated at the end of follow-up, and 165 (84 in the 4.0 g/day group and 81 in the 2.4 g/day group) underwent endoscopic examination.
The pre-trial characteristics of the 206 randomized patients are shown in Table 1. All features examined were similar in the two groups of patients. In particular, the two groups were homogeneous with regard to sex, age, smoking habits, duration of disease, site and extent of the lesions, clinical course and indications for surgery, CDAI score and acute phase reactants at operation, previous treatments and indication and type of surgery.
Table 1. Pre-trial characteristics of the study population
Age (years) (mean)
Duration of disease (years) (mean)
Clinical course (indications for surgery)
Type of surgery
Only four patients (two in the 4.0 g/day group and two in the 2.4 g/day group) reported adverse reactions to mesalazine and were withdrawn from the trial. In the 4.0 g/day group, two patients developed severe dyspepsia, whilst, in the 2.4 g/day group, an increase in transaminases was observed in one patient and lower limb cramps in the other. Interruption of treatment promptly led to complete disappearance of clinical signs in all four patients.
Patients adhering to protocol
Twenty patients (six in the 4.0 g/day group and 14 in the 2.4 g/day group) failed to return for clinical control. In the 4.0 g/day group, four patients refused to continue treatment, one died from suicide and the other was lost to follow-up. In the 2.4 g/day group, five patients refused to continue treatment, one was submitted to surgery for ovarian carcinoma and the remaining eight were lost to follow-up.
A total of 41 patients (17 in the 4.0 g/day group and 24 in the 2.4 g/day group) failed to attend for endoscopy.
Compliance was considered to be good, as all patients took more than 80% of the tablets prescribed and none interrupted treatment for more than two consecutive weeks.
The endoscopic findings at 1 year of follow-up are shown in Table 2.
Table 2. Endoscopic recurrence at 1 year of follow-up
> 0 (1–4)
> 1 (2–4)
> 2 (3–4)
Endoscopic recurrence with a score of > 0 (grades 1, 2, 3, 4) was observed in 39 patients (46%) in the 4.0 g/day group and in 50 patients (62%) in the 2.4 g/day group, the difference being statistically significant [risk difference (RD), − 0.153; 95% confidence interval (CI), from − 0.303 to − 0.003; z = − 1.996; P = 0.04].
Endoscopic recurrence with a score of > 1 was observed in 28 patients (33%) in the 4.0 g/day group and in 35 patients (43%) in the 2.4 g/day group. The difference was not statistically significant (RD, − 0.099; 95% CI, from − 0.246 to 0.049; z = − 1.311; P = 0.19).
Severe endoscopic recurrence (score > 2) was observed in 27% of patients in the 4.0 g/day group and in 37% of patients in the 2.4 g/day group. The difference was not statistically significant (RD, − 0.097; 95% CI, from − 0.238 to 0.045; z = − 1.333; P = 0.18).
Overall clinical recurrence was observed in 11 patients (12%) in the 4.0 g/day group and in 13 patients (14%) in the 2.4 g/day group. The difference was not statistically significant (RD, − 0.027; 95% CI, from − 0.124 to 0.069; z = − 0.549; P = 0.58).
Recurrence requiring further surgery was observed in only two cases, both in the 2.4 g/day group.
At univariate analysis, in the three endoscopic outcomes, only therapy (P = 0.04) was protective in the first endoscopic outcome (score > 0), whereas only ileo-transverse anastomosis increased the risk of recurrence in the second and third endoscopic outcomes (scores > 1 and > 2) (P = 0.01). Considering that, for each outcome, only one variable reached statistical significance, no multivariate analysis was carried out.
The results of this study show that a regimen of 4.0 g/day of mesalazine offers only a slight advantage over a regimen of 2.4 g/day in the prevention of post-operative endoscopic recurrence at 1 year of follow-up after radical surgery of Crohn's disease limited to the terminal ileum, with or without involvement of the caecum/ascending colon.
The total number of cases of post-operative endoscopic recurrence was lower in the 4.0 g/day group.
The results were significant when the presence of any lesion (endoscopic score of > 0) was used as the endoscopic outcome measure. However, when more than five aphthae (score of > 1 or > 2) were considered as the endoscopic outcome measure, the difference between the two regimens was not significant. In the 4.0 g/day group, there was a slight decrease in the proportion of severe endoscopic recurrence (score > 2), which is considered to predict a poor prognosis.18
With regard to clinical recurrence, the difference between the two dosages of mesalazine was not significant. The frequencies of clinical events were relatively low (12% vs. 14%), and comparable with those of other studies.2,3,5,14 It is difficult to show a significant advantage between the two doses of mesalazine at such a low level of clinical recurrence.
The trial presented here lacks a placebo control arm as, when it was planned, the meta-analysis of Cammàet al. had clearly shown a benefit of mesalazine in reducing clinical relapse in the post-surgical setting.7
Subsequently, a large European co-operative study showed that mesalazine, 4.0 g/day, did not significantly affect post-operative clinical recurrence.6 However, a sub-group of patients with small bowel Crohn's disease showed a significant benefit when treated with mesalazine, 4.0 g/day, with respect to placebo. Cottone and Cammà have updated the results of their meta-analysis to include the European trial, and the results are still in favour of treatment with mesalazine.15
The large number of patients in our study allowed us to perform a meta-analysis of six studies evaluating the role of mesalazine in the prevention of endoscopic recurrence of Crohn's disease. The results of this meta-analysis showed that mesalazine reduces, by 18%, the rate of endoscopic recurrence, which is a clinically relevant result; the number needed to treat is 5.5. We have also updated the meta-analysis on clinical recurrence including the results of this study; the data still remain in favour of mesalazine, with an overall risk difference of 15%, which is also clinically relevant; the number needed to treat is 6.6.20
Rutgeerts and co-workers have shown that metronidazole reduces the 1-year clinical recurrence of Crohn's disease by per protocol analysis and, on the basis of this finding, suggest that metronidazole should be taken into consideration in the prevention of post-operative recurrence.8,21 However, the risk difference (18%) by intention-to-treat analysis obtained in their study (60 patients) is comparable with the overall risk difference (15%) obtained with mesalamine in six trials with a total of 1141 patients. We believe that the risk difference emerging from a large series is more reliable than that obtained from a single study. Recently, a randomized controlled trial has reported the efficacy of ornidazol in the prevention of post-operative recurrence (80 Crohn's disease patients), although the tolerance was low.9 In this study, a close relationship was observed between the development of severe endoscopic Crohn's disease lesions in the neoterminal ileum early after surgery and the subsequent development of clinical relapse.
In conclusion, our results provide additional evidence of the efficacy of mesalazine in the prevention of the post-operative recurrence of Crohn's disease after radical surgery for disease limited to the terminal ileum, with or without involvement of the caecum/ascending colon. Considering that 4.0 g/day offers only a minimal advantage over 2.4 g/day, in terms of cost/benefit, a dose of 2.4 g/day should be recommended as first-choice regimen in the post-surgical setting.
Gruppo Italiano per lo Studio del Colon e del Retto (GISC) co-operative investigators: Piero Vernia, Livia Biancone, Francesco Pallone, Lucio Capurso, Stefano Minervini, Loredana Gili (Rome); Giuseppe Frieri (L'Aquila); Marilena Fazi (Florence); Daniela Valpiani, Nicola D'Imperio (Forlì); Michele Comberlato (Bolzano); Renata D'Incà, Luca Benazzato (Padua); Roberto Di Mitri, Ambrogio Orlando, Lorenzo Oliva, Angelo Casà (Palermo); Giovanni Lombardi (San Giovanni Rotondo); Tindaro Pontillo (Catania); Angelo Pera (Turin); Gilberto Poggioli, Silvio Laureti, Giorgio Zoli (Bologna); Gabriele Riegler, Antonio Rispo, Antonio Cozzolino, Mariateresa Tartaglione (Naples); Marcello Ingrosso (Bari).
This study was supported by a grant from Giuliani SpA, Milan, Italy.