Ferring pharmaceuticals provided an unrestricted grant to perform this study.
High mucosal healing rates in 5-ASA-treated ulcerative colitis patients: Results of a meta-analysis of clinical trials†
Article first published online: 14 MAR 2012
Copyright © 2012 Crohn's & Colitis Foundation of America, Inc.
Inflammatory Bowel Diseases
Volume 18, Issue 11, pages 2190–2198, November 2012
How to Cite
Römkens, T. E.H., Kampschreur, M. T., Drenth, J. P.H., van Oijen, M. G.H. and de Jong, D. J. (2012), High mucosal healing rates in 5-ASA-treated ulcerative colitis patients: Results of a meta-analysis of clinical trials. Inflamm Bowel Dis, 18: 2190–2198. doi: 10.1002/ibd.22939
- Issue published online: 15 OCT 2012
- Article first published online: 14 MAR 2012
- Manuscript Accepted: 14 FEB 2012
- Manuscript Received: 12 FEB 2012
- mucosal healing;
- ulcerative colitis;
- systematic review;
- clinical trials
Recently, mucosal healing (MH) is regarded as an important treatment goal in ulcerative colitis (UC). 5-Aminosalicylates (5-ASA) are the standard treatment in mild-to-moderate UC, but the effect on MH is less known. The aim of this study was to systematically review the medical literature in order to compare different preparations of 5-ASA for the effect on MH.
We conducted a structured search of PubMed and the Cochrane Central Register of Controlled Trials to identify randomized controlled clinical trials with 5-ASA in UC providing data about MH. We calculated the sample size-weighted pooled proportion of patients with MH, and performed meta-analysis of head-to-head comparisons.
Out of 645 hits, we included 90 treatment arms, involving 3977 patients using oral 5-ASA (granulate and tablets) and 2513 patients using rectal 5-ASA (suppositories, enema, and foam). Overall, 43,7% of 5-ASA treated patients achieved MH (oral 36,9%; rectal 50,3%). In oral studies, 49% of patients using granulate (7 treatment-arms) achieved MH compared to 34,9% using tablets (43 treatment-arms). In rectal studies the proportion of MH was 62% for suppositories (eight treatment arms), 51% for foam (nine treatment arms), and 46% for enema (23 treatment arms), respectively.
5-ASA preparations achieved MH in nearly 50% of UC patients. There were no significant differences in MH between the various 5-ASA agents, either in the oral or the rectal treatment groups. (Inflamm Bowel Dis 2012;)
Ulcerative colitis (UC) is a chronic inflammatory bowel disease (IBD) limited to the colonic mucosa with an unknown etiology. It is characterized by diarrhea, rectal bleeding, abdominal pain, and, without treatment, usually follows a relapsing-remitting course. The degree of disease activity can vary from mild to highly severe.1–3 Formulations containing 5-aminosalicylic acid (5-ASA) are the keystone for induction and subsequent maintenance of remission in mild-to-moderate active UC.4–7 The exact mechanism of action of 5-ASA remains unclear. However, the antiinflammatory effect in UC appears to be related to local exposure to the mucosal surfaces of the colon.8, 9
Over the years a variety of scoring systems have been developed to assess disease activity in UC patients, based both on clinical symptoms as well as on endoscopic healing and remission. In the late 1990s greater interest in the latter scoring system was raised by the finding that the biological agent infliximab could possibly achieve endoscopic healing of mucosal ulcerations in both Crohn's disease (CD) and UC.10–14 Since then mucosal healing (MH) has been more often proposed as an important goal in the biological treatment of CD and UC patients. Subsequently, numerous studies provided convincing evidence that 5-ASA formulations also may possess the ability to heal the colonic mucosa.4, 14, 15 Indeed, when achieved, MH is associated with a lower risk of relapse, or future surgical interventions, a reduced risk of colorectal malignancy, and improved quality of life. These key elements emphasize the significance of MH in UC as an indicator of treatment efficacy and a marker of long-term prognosis.4, 15 In addition, the use of MH may facilitate standardization of scoring systems and might serve as an objective determinant of treatment response in future clinical trials.
We performed the present study predominantly to assess the ability of 5-ASA preparations to accomplish MH in UC patients. Furthermore, we were interested in possible differences in efficacy between various 5-ASA formulations and dosages in inducing MH. We hypothesized that both local and oral 5-ASA-formulations are effective in inducing MH in UC patients with mild to moderate active disease.
The aims of the present study were to systematically review 1) the efficacy on MH of oral and rectal formulations of 5-ASA in UC; 2) definitions for MH in UC; and 3) to perform meta-analyses for different formulations and dosages of 5-ASA on MH.
MATERIALS AND METHODS
The search strategy was designed to identify English-language randomized controlled clinical trials (RCT) of 5-ASA in adult UC patients that provided data about induction of MH. We performed a structured search of PubMed and the Cochrane Central Register of Controlled Trials. The following search terms and MeSH terms were used: “5-aminosalicylic acid,” “5-ASA,” “mesalamine,” “mesalazine,” “olsalazine,” “balsalazide,” “inflammatory bowel disease,” and “IBD.”
Studies were eligible if they were 1) of randomized controlled design; 2) reported on induction of endoscopic MH; 3) studied 5-ASA preparations monotherapy; and 4) in patients with mild to moderate active UC. Studies were excluded if they 1) reported on maintenance of remission only; 2) studied combination therapy; or 3) studied combined populations of UC and CD. Only studies in the English language were included.
Two independent reviewers (T.R. and M.K.) screened the titles and abstracts of the articles identified by the electronic searches and completed an inclusion form for eligible studies. We applied a preliminary screen and we identified studies that could meet the inclusion criteria for full review. Full copies of all included studies were retrieved and reviewed by two authors (T.R. and M.K.). In case of disagreement they discussed with a third author (M.vO.) in order to reach consensus.
Data Extraction and Quality Assessment
Full papers were reviewed for quality and extracted data using a purpose-designed database. Demographic variables extracted were study design, participating countries, description of patients, description of severity of disease, type and dose of medication, and definition of MH. The extracted outcome data were frequencies regarding different definitions, scores, and cutoffs for MH, number of patients with MH, and comparison of the proportion of MH between these definitions. Important quality criteria for studies concerning MH in UC were the availability of an endoscopic measurement of MH and the use of a reliable definition for MH.
Data Synthesis and Analysis
All abstracted data were entered into a structured database and analyses were performed using SAS Statistical Software, v. 8.2 (Cary, NC), and Cochrane Review Manager 5. Kappa statistics were used to analyze agreement among reviewers for every step in literature acquisition. Frequency and evidence tables were provided for characteristics of the study treatment arms. For analysis of the primary outcome, we calculated the sample size-weighted pooled proportion of patients with MH and compared that between different formulations in oral and rectal treatment arms of the included studies. We carried out a meta-analysis of head-to-head comparisons of formulations or dosages within the oral and rectal treatment arms. Based on heterogeneity between included studies, we used random effect models. For all comparisons forest plots were provided.
Our literature search is summarized in a flow diagram (Fig. 1). Our initial search identified 645 studies, of which we excluded 459 studies for various reasons. We applied full-text screening for 186 articles. The main reasons for exclusion were inappropriate data on induction of MH, inclusion of both CD and UC patients in one statistical analysis without presenting separate data, or structural use of comedication for IBD. Eventually, we included 49 articles (90 treatment arms) in our meta-analysis, involving 6490 patients using 5-ASA preparations (n = 3977 oral; n = 2513 rectal). All essential details of the included studies are depicted in the Appendix.1–9, 16–55
In each step of the literature search, presented in Figure 1, the calculated kappa for agreement between both reviewers was above 0.8, indicating good agreement. Figure 2 shows the distribution of the various 5-ASA formulations over the 90 study arms. The overall baseline characteristics of oral and rectal treated UC patients are depicted in Table 1. As expected, the proportion of patients with pancolitis was higher in the oral-treated UC patient group compared with the rectal group. There were no other major differences between groups, although patients treated with rectal 5-ASA were slightly younger.
|Patients Characteristics Baseline||Oral Arms (N=50) (3977 Patients)||Rectal Arms (N=40) (2513 Patients)|
|Mean age (yrs)||42.3||40.2|
|Mean % males||53.2||52.6|
|Mean duration UC (months)||63.9||62.4|
|Mean % pancolitis||27.3||0|
|Mean % smoking||10.1||8.9|
|Industry funded (n (%))||46 (92)||27 (68)|
Table 2 displays all definitions and cutoffs used for MH in the included study arms. We identified 19 different definitions and cutoffs for MH. Overall, by using the MH definition of the original articles, 43.7% of UC patients treated with 5-ASA achieved MH. This was 36.9% in the oral group and 50.3% in the rectal group. Pooled data from head-to-head comparisons showed a higher proportion of MH in patients treated with a higher mean dose of 5-ASA (Fig. 4a,b), although this was only statistically significantly different in the oral group. Head-to-head comparisons of treatment duration of 5-ASA and efficacy in terms of MH are not available from the included studies. In the rectal-treated group MH seemed comparable in studies with a short treatment period (<6 weeks) to studies with a longer treatment period (>6 weeks); in the oral group MH rates seemed slightly higher when longer treated.
|Endoscopic Score||Treatment Arms (N)||% Mucosal Healing (Range)|
|Endoscopic index ≤ 1||2||50.5 (48.4-52.2)|
|Endoscopic index <3||6||36.5 (28.9-45)|
|Endoscopic index ≤4||9||57.2 (42-72.7)|
|Sigmoidoscopy score of 0||5||55.2 (17-80)|
|Sigmoidoscopy score ≤1||6||67.2 (53-77.6)|
|Baron Score||11||45.3 (14.3-65)|
|Truelove and Richards||6||61.9 (43.3-93)|
|Combined clinical and endoscopic score||27||30.7 (0-74)|
Using the original articles definition of MH, granulate treatment seemed associated with higher MH rates compared with treatment with tablets in the oral-treated group (49% vs. 34.9%), using comparable dosage of 5-ASA (mean dose 3.5 g in granulate vs. 3.4 g/day in tablets) (Fig. 3). However, head-to-head comparisons of the few studies using both 5-ASA tablets and granulate revealed no significance difference between the treatments (Fig. 5a). In addition, we looked at the effect of the various brands and delivery systems of the oral 5-ASA agents with respect to MH, which seemed comparable. Between these groups there are no head-to-head data available and extensive subgroup analyses in these small numbers of patients may be misleading.
Similarly, in the rectal 5-ASA treatment group induction of MH differed between the various preparations using the original articles definitions. Suppositories achieved MH in 62% of UC patients versus 51% in foam and 46% in enema (Fig. 3) (mean dose respectively 1.2 vs. 2.1 vs. 2.7 g/day). Naturally, patients with proctitis, using only suppositories, represent a different patient group with different characteristics. Therefore, only head-to-head comparisons of studies using both 5-ASA foam and enema were performed. Again, MH rates were not statistically significantly different between these rectal 5-ASA preparations (Fig. 5b).
Based on this meta-analysis, we conclude that 5-ASA formulations are able to heal the colonic mucosa in nearly half of all patients with mild-to-moderate UC, regardless of the used definition of MH or the 5-ASA formulations.
Theoretically, the different dose and delivery systems of 5-ASA formulations used may influence treatment distribution and efficacy. However, in our analysis in mild to moderate UC we found no difference in MH between the various 5-ASA agents and delivery systems, either in oral- and rectal 5-ASA-treated patients. In the head-to-head comparisons of rectal 5-ASA, patients with proctitis only, using suppositories, were not included because this is a different patient group with different characteristics. Our finding that different 5-ASA agents and delivery systems have no differential effect on MH is in concordance with a recently published Cochrane review.59 There the authors conclude that future trials should focus more on improving patient adherence rather than detecting differences in efficacy.
We found a significant 5-ASA dose–response relationship in the oral group, but only a trend in the rectal treatment group. A previous meta-analysis for rectal 5-ASA preparations is in concurrence with our findings and similarly failed to detect a significant dose–response relationship.56 On the other hand, an advantage for higher oral 5-ASA doses was demonstrated earlier for UC patients with more severe disease activity.31, 58 This detailed patient information could not be subtracted from the original articles analyzed in our meta-analysis. Furthermore, a recently published British guideline on UC indicates comparable remission rates on high-dose versus low-dose 5-ASA. This guideline does not elaborate on which 5-ASA formulation to use and mentions that “a greater clinical improvement,” but not necessarily remission, is associated with 5-ASA doses >3 g/day.57
In summary, in our study both rectal and oral treatment with 5-ASA shows MH rates of nearly 50%, irrespective of the formulations used in the head-to-head comparison. Therefore, we recommend first-line topical 5-ASA treatment in cases of patients with mild to moderate proctitis or proctosigmoiditis, irrespective of the used formulation and dosage. In cases of mild to moderate left-sided or pancolitis, oral 5-ASA is recommended as first-line treatment, with no significant difference between the different formulations, but with higher MH rates by using ≥3g/day.
The therapeutic goals for UC have changed from predominantly treatment of symptoms toward achievement of endoscopic remission. The use of MH in the treatment of UC patients may facilitate standardization of scoring systems and might serve as an objective determinant of treatment response in future clinical trials. However, before MH can be introduced as a treatment goal, a uniform definition for MH must be developed and validated. Unfortunately, at this moment this uniform accepted definition is lacking. Between all the included RCTs in our study we identified 19 different definitions and cutoffs for endoscopic MH. In 2007, the clinical trials task force of the International Organization of Inflammatory Bowel Disease (IOIBD) recommended a definition of genuine endoscopic healing in UC as the absence of friability, blood, erosions, and ulcers in all visualized segments, whereby an abnormal vascular pattern in the absence of these other features is still compatible with endoscopic healing.60 We support this and call for acceptance of this definition for future studies in order to compare efficacy data. However, because until now authors used heterogeneous definitions on MH in UC, we accepted the definitions of MH as indicated in the individual articles to facilitate quantitative pooling of results in our analysis. Naturally, it would have been preferable to convert the definition for MH from each trial to reduce interstudy heterogeneity and improve the accuracy of the MH rate in this meta-analysis, but this would have implied a completely different study strategy. As might be expected, studies with less stringent definitions reported higher MH rates.1, 7, 20, 21, 24, 36, 37, 44, 52
Our study has several strengths and some limitations. First, our study provides the most extended overview of the efficacy of various 5-ASA formulations in UC patients on MH. Therefore, according to us, we present substantial data on this issue. Second, by broadening the primary search term to “inflammatory bowel disease,” we made attempts to identify all useful trials to obtain data that strengthened our analysis. “Sulfasalazine” was not used as a separate search term, but no RCT that met the inclusion criteria was excluded. Third, two independent reviewers reviewed all eligible studies, with very good agreement scores based on the presented kappa value, which was calculated in every literature review step, presented in Figure 1. Finally, we only included RCTs published as full articles, without restrictions by year of publication. Although we restricted our analysis to English-language RCTs, there was no large trial excluded. In the primary search we found 53 non-English studies in both the Cochrane database and PubMed database. None of them were available as full text. By selection on title and abstract, only one controlled study in 50 UC patients using 5-ASA mentions data on MH.61 In general, it can be misleading to draw conclusions by pooling data of several partly heterogenic studies, as we did. Using different definitions, dosage, type of delivery system, and duration of treatment can alter the result. We tried to overcome this problem as much as we could by performing head-to-head comparisons of available data on the mentioned topics, which showed no major differences, except for dosage. Statistical heterogeneity between the included studies was low, according to the I2 in the presented Forrest plots. Still, heterogeneity of the included studies must be taken into account by interpreting the results.
In addition, possible confounding aspects in this meta-analysis are the permitted comedication, comorbidity, the length of treatment, and nonuniformity in the definition of the extent of disease. Single trials that reported interim endpoints at different time intervals observed higher endoscopic remission rates with prolonged treatment.16, 17
In our study the duration of treatment among the accepted trials ranged from 2–12 weeks. There seemed to be no major differences in the rectal-treated group, but in the oral group MH rates seemed to be slightly higher in patients with longer treatment, but head-to-head-comparisons of different treatment durations and the effect on MH are not available from the included studies.
In summary, we showed that 5-ASA preparations are highly effective in inducing MH in mild to moderate UC. For this reason, 5-ASA remains to be considered the first-line therapy for patients with mild to moderately active UC, irrespective of disease extension and of 5-ASA formulations used.
We thank Margriet Moret-Hartman for statistical support.
- 3Mesalazine (5-aminosalicylic acid) micropellets show similar efficacy and tolerability to mesalazine tablets in patients with ulcerative colitis—results from a randomized-controlled trial. Aliment Pharmacol Ther. 2004; 20: 1353–1363., , .
- 7Clinical trial: randomized-controlled clinical study comparing the efficacy and safety of a low-volume vs. a high-volume mesalazine foam in active distal ulcerative colitis. Aliment Pharmacol Ther. 2007; 26: 1237–1249., , , et al.
- 16[No authors listed.] Topical 5-aminosalicylic acid versus prednisolone in ulcerative proctosigmoiditis. A randomized, double-blind multicenter trial. Danish 5-ASA Group. Dig Dis Sci. 1987; 32: 598–602.
- 20Comparison of the efficacy and safety of Eudragit-L-coated mesalazine tablets with ethylcellulose-coated mesalazine tablets in patients with mild to moderately active ulcerative colitis. Aliment Pharmacol Ther. 2006; 23: 1017–1026., , , et al.
- 30Delayed-release oral mesalamine at 4.8 g/day (800 mg tablet) for the treatment of moderately active ulcerative colitis: the ASCEND II trial. Am J Gastroenterol. 2005; 100: 2478–2485., , , et al.Direct Link:
- 33Mesalazine 4 g daily given as prolonged-release granules twice daily and four times daily is at least as effective as prolonged-release tablets four times daily in patients with ulcerative colitis. Inflamm Bowel Dis. 2001; 7: 237–242., , , et al.
- 34Balsalazide is superior to mesalamine in the time to improvement of signs and symptoms of acute mild-to-moderate ulcerative colitis. Am J Gastroenterol. 2002; 97: 3078–3086., , , et al.Direct Link:
- 44Mesalamine foam enema versus mesalamine liquid enema in active left-sided ulcerative colitis. Am J Gastroenterol. 2008; 103: 3106–3114., , , et al.Direct Link:
- 46Safety and efficacy of two dose formulations of alicaforsen enema compared with mesalazine enema for treatment of mild to moderate left-sided ulcerative colitis: a randomized, double-blind, active-controlled trial. Aliment Pharmacol Ther. 2006; 23: 1403–1413., , , et al.
- 555-Aminosalicylic acid or sulfasalazine retention enemas in distal ulcerative colitis. A randomized therapeutic trial. Current Therapeutic research, vol 42, no 5; 1987., , , et al.
- 56Rectal 5-aminosalicylic acid for induction of remission in ulcerative colitis. Cochrane Database Syst Rev. 2010; CD004115., , , et al.