Short-term adverse effects of 5-aminosalicylic acid agents in the treatment of ulcerative colitis
Dr E. V. Loftus Jr, Division of Gastroenterology and Hepatology, Mayo Clinic, 200 First Street, SW, Rochester, MN 55905, USA.
Aim : To determine whether there is a difference in short-term adverse events in patients with ulcerative colitis treated with mesalazine, olsalazine or balsalazide.
Methods : MEDLINE was searched for articles published until 2002. Randomized trials of oral mesalazine, olsalazine or balsalazide for the treatment of active disease or the maintenance of remission were included. Outcomes of interest were the frequencies of patients experiencing adverse events and those withdrawn due to adverse events.
Results : Forty-six trials were included. One study of mesalazine vs. sulfasalazine for active colitis showed significantly fewer patients with adverse events with mesalazine. Both balsalazide vs. sulfasalazine studies for active disease showed significantly fewer withdrawals with balsalazide. One trial of balsalazide vs. sulfasalazine for maintenance showed significantly fewer patients with adverse events with balsalazide. Otherwise, no significant differences in safety outcomes were noted.
Conclusion : All three 5-aminosalicylic acid agents are safe in the short term. In mesalazine-treated patients, the frequencies of adverse events or withdrawals due to adverse events were comparable with those in placebo-treated patients and lower than those in sulfasalazine-treated patients. Overall, adverse events or withdrawals were not significantly more frequent with olsalazine than with placebo or sulfasalazine. Adverse events and study withdrawals on balsalazide were less frequent than those on sulfasalazine.
It has been 60 years since the initial landmark report of the efficacy of sulfasalazine for the treatment of ulcerative colitis.1 For many years, sulfasalazine was the mainstay of therapy for mild to moderately active ulcerative colitis. Unfortunately, treatment with this agent is problematic for many patients due to side-effects, such as headache, dyspepsia, nausea or allergy. The determination that 5-aminosalicylic acid (5-ASA), not sulfapyridine, was the active moiety in sulfasalazine led to the development of new drugs delivering 5-ASA to the colonic mucosa without the side-effects of sulfapyridine.2 There are now several 5-ASA-based agents available for the treatment of ulcerative colitis. These include sulfasalazine, mesalazine or 5-ASA itself, olsalazine (Dipentum), a dimer of 5-ASA, and balsalazide (Colazal or Colazide), a pro-drug consisting of 5-ASA joined to an inert benzoic acid derivative by an azo bond. Oral mesalazine is available in several forms, including tablets coated with a methacrylate resin, Eudragit S, which dissolves at a luminal pH ≥ 7 (Asacol), tablets coated with Eudragit L, dissolving at a luminal pH ≥ 6 (Claversal, Salofalk, Mesasal), and 5-ASA in ethylcellulose microspheres with a time-dependent release of the drug (Pentasa). The non-sulpha-based 5-ASA agents are thought to be associated with fewer side-effects than sulfasalazine, with the possible exception of olsalazine, which produces a dose-dependent secretory diarrhoea in some patients.
It is often difficult to determine from individual reports whether a particular therapy is potentially harmful. Well-designed studies on safety have the following characteristics: (i) clearly identified groups similar with respect to important determinants of outcome; (ii) similar measurement of outcome and exposures in the groups being compared; and (iii) sufficiently long and complete follow-up.3 In general, randomized clinical trials are the most robust studies to determine whether a particular therapy is harmful, provided that the above criteria are met, followed by other types of cohort studies and case–control studies.3 Case series and case reports are anecdotal in nature, do not provide a denominator to identify the magnitude of the risk and do not meet the above criteria. They are therefore more useful in generating hypotheses concerning potential harmful effects than determining the risk of an event.
Systematic reviews are helpful in evaluating potential harmful outcomes by utilizing comprehensive literature searches, pre-specified inclusion criteria for the selection of studies and duplicate extraction of data by independent investigators. These techniques minimize bias in the identification of appropriately designed studies and in the extraction of study data.
The data on the potential harmful effects of 5-ASA are limited. A meta-analysis of randomized clinical trials of 5-ASA in the treatment of ulcerative colitis,4 subsequently updated,5 focused mostly on efficacy, but safety was examined only briefly. These meta-analyses suggested that patients receiving 5-ASA therapy for active ulcerative colitis were equally as likely as those on placebo to experience an adverse event or be withdrawn from a trial due to an adverse event.5 Patients receiving 5-ASA for active disease were less likely than those taking sulfasalazine to experience an adverse event or be withdrawn from a trial due to such an event. Patients treated with either 5-ASA or sulfasalazine to maintain remission of ulcerative colitis demonstrated no difference in withdrawal or adverse event rates.5
Since the publication of the most recent meta-analysis, at least 10 randomized clinical trials of mesalazine or balsalazide for the treatment of ulcerative colitis have been published. There have been no systematic reviews focusing on the safety of these agents during the same period of time. We therefore undertook a systematic review to address the focused question: ‘Is there a difference in side-effects in patients with ulcerative colitis treated with mesalazine, olsalazine or balsalazide?
The on-line bibliographic database MEDLINE was searched for potentially relevant articles in any language published between 1966 and August 2002. The Medical Subject Heading terms ‘ulcerative colitis’ and ‘aminosalicylic acids’ or ‘mesalazine’ or ‘olsalazine’ were used to perform initial searches, and keyword searches of ‘mesalazine’, ‘olsalazine’ and ‘balsalazide’ were also performed.
Study selection criteria
Two independent reviewers (E.V.L. and S.V.K.) identified relevant studies. Reports were analysed further if they were prospective, randomized, double-blind, controlled clinical trials of oral mesalazine, oral olsalazine or oral balsalazide for the treatment of active ulcerative colitis or for the maintenance of remission of ulcerative colitis. Observational studies, open-label studies and single-blind studies were excluded. Only fully published reports were considered — studies available in only preliminary/abstract form were excluded. Three studies in which no safety data were reported were excluded from the analysis.6–8 Studies of combination 5-ASA therapy (i.e. oral and rectal) were considered only if both patient and investigator were blind with respect to the dosage of oral 5-ASA therapy. Likewise, dosing studies of 5-ASA therapy were considered only if both patient and investigator were blind to the dosage.
Data extraction and analysis
The investigators abstracted data from each of the relevant studies on to standardized data collection forms. Disagreement between investigators was resolved by consensus. The abstracted data included the type and dosage of treatment in each treatment arm, the number of patients enrolled and the number of patients actually receiving treatment. The outcomes included the number of patients who experienced an adverse event (any, serious or severe) and the number of patients who were withdrawn from the study due to an adverse event. If available, data on the total number of adverse events and the number of individual adverse events, such as headache, diarrhoea or abdominal pain, were abstracted.
The search strategies yielded a total of 351 potentially relevant studies. After further review, 46 randomized, double-blind clinical trials of mesalazine, olsalazine or balsalazide therapy for ulcerative colitis were included.9–54 The characteristics of 23 trials for the induction of remission of active ulcerative colitis are listed in Table 1, and those of 23 studies for the maintenance of remission are shown in Table 2.
Table 1. Randomized clinical trials of mesalazine, olsalazine or balsalazide for the treatment of active ulcerative colitis
|Selby et al.9|| 40||Steroids, IS||Olsalazine||2.0 g||Placebo|| || 2 weeks|
|Meyers et al.10|| 66||Severe UC||Olsalazine||0.75, 1.5, 3 g||Placebo|| || 3 weeks|
|Schroeder et al.11|| 87||Steroids, sulfa||Asacol||1.6, 4.8 g||Placebo|| || 6 weeks|
|Ewe et al.12|| 40||Severe UC, steroids, IS||Olsalazine||1.5 g||Sulfasalazine||3 g|| 4 weeks|
|Hetzel et al.13|| 30||Severe UC, steroids,||Olsalazine||2 g||Placebo|| || 6 weeks|
| sulfa, abx, IS|| || || || || |
|Riley et al.17|| 61||Steroids||Asacol||0.8, 2.4 g||Sulfasalazine||2 g|| 4 weeks|
|Willoughby et al.19|| 56||Steroids, IS, abx||Olsalazine||3 g||Sulfasalazine||3 g|| 5 weeks|
|Feurle et al.20||105||Severe UC, sulfa, abx, IS||Olsalazine||2 g||Placebo|| || 4 weeks|
|Rachmilewitz21||220||Other UC rx||Mesasal||1.5 g||Sulfasalazine||3 g|| 8 weeks|
|Rao et al.22|| 37||Severe UC, steroids||Olsalazine||2 g||Sulfasalazine||3–4 g|| 4 weeks|
|Zinberg et al.24||15||Steroids, sulfa, IS||Olsalazine||3 g||Placebo|| || 4 weeks|
|Sninsky et al.25||158||Steroids, sulfa, abx, IS||Asacol||1.6, 2.4 g||Placebo|| || 6 weeks|
|Ferry et al.30|| 56||Severe UC, abx||Olsalazine||30 mg/kg||Sulfasalazine||60 mg/kg|| 3 months|
|Hanauer et al.31||374||Steroids, sulfa, IS||Pentasa||1.0, 2.0, 4.0 g||Placebo|| || 8 weeks|
|Munakata et al.37||118||Steroids, IS||Pentasa||1.5 g||Sulfasalazine||3 g|| 4 weeks|
|Safdi et al.42|| 60||Steroids||Asacol||2.4 g||Rowasa or |
|4 g|| 6 weeks|
|Green et al.45||101||Steroids, IS||Balsalazide||6.75||Asacol||2.4 g||12 weeks|
|Kruis et al.46||168||Severe UC, steroids, IS, abx||Claversal||3 g||Olsalazine||3 g||12 weeks|
|Vecchi et al.50||130||Severe UC, steroids, IS||Salofalk||4 g||Salofalk oral |
|4 g|| 6 weeks|
|Green et al.51|| 57||Steroids, IS||Balsalazide||6.75 g||Sulfasalazine||3 g||12 weeks|
|Levine et al.52||154||Severe UC, IS, abx, topical rx||Balsalazide||2.25–6.75 g||Asacol||2.4 g|| 8 weeks|
|Mansfield et al.53|| 50||Steroids, IS||Balsalazide||6.75 g||Sulfasalazine||3 g|| 8 weeks|
|Pruitt et al.54||173||Severe UC, steroids, IS||Balsalazide||6.75 g||Asacol||2.4 g|| 8 weeks|
Table 2. Randomized clinical trials of mesalazine, olsalazine or balsalazide for the maintenance of remission in ulcerative colitis
|Ireland et al.14||164||Steroids, abx, IS||Olsalazine||2 g||Sulfasalazine||2 g|| 6 months|
|Kiilerich et al.28||223||Steroids, IS||Olsalazine||2 g||Sulfasalazine||2 g||12 months|
|McIntyre et al.15|| 79||Steroids, IS||Balsalazide||2 g||Sulfasalazine||2 g||24 weeks|
|Mulder et al.16|| 75||Steroids, IS||Pentasa||1.5 g||Sulfasalazine||3 g||12 months|
|Riley et al.18||100||Steroids||Asacol||> 0.8 g||Sulfasalazine||> 2 g||48 weeks|
|Rutgeerts23||334||Steroids, abx, IS||Claversal||0.75 g||Sulfasalazine||1.5–2.0 g||12 months|
|Giaffer et al.26||133||Steroids, IS||Balsalazide||2.0 g||Balsalazide||4.0 g||12 months|
|Green et al.27||108||None||Balsalazide||1.5 g||Balsalazide||3.0 g||12 months|
|Rijk et al.29|| 46||Steroids, abx||Olsalazine||2 g||Sulfasalazine||4 g||48 weeks|
|Wright et al.32||101||Steroids, sulfa||Olsalazine||2 g||Placebo|| ||12 months|
|Travis et al.33||199||None||Olsalazine||0.5 g||Olsalazine||1, 1.5 g||12 months|
|Kruis et al.36||162||Other UC rx||Olsalazine||0.5, 1.25, 2 g||Sulfasalazine||2 g|| 6 months|
|Nilsson et al.39||322||Other UC rx||Olsalazine||2 g||Sulfasalazine||2 g||6–18 months|
|Ardizzone et al.34|| 88||None||Claversal||1 g||Sulfasalazine||2 g||12 months|
|Fockens et al.35||169||Steroids, IS||Pentasa||1.5 g||Pentasa||3.0 g||12 months|
|Miner et al.38||205||Steroids, sulfa, IS||Pentasa||4 g||Placebo|| ||12 months|
|Anonymous40||264||Steroids, topical rx||Asacol||0.8–1.6 g||Placebo|| || 6 months|
|Hawkey et al.43||323||Steroids||Asacol||1.6 g||Zileuton |
|2.4 g||26 weeks|
|Kruis et al.41||120||Other UC rx||Salofalk||1.5 g||E. coli Nissle||0.2 g||12 weeks|
|Green et al.44|| 99||Steroids, IS||Balsalazide||3 g||Asacol||1.2 g||12 months|
|Ardizzone et al.47||112||Steroids, IS||Asacol||1.2 g||Placebo|| ||12 months|
|Rembacken et al.48||116||None||Asacol||1.2–2.4 g||E. coli |
|Kruis et al.49||133||Steroids, abx, IS||Balsalazide||1.5–3.0 g||Salofalk||1.5 g||26 weeks|
Treatment of active ulcerative colitis
The number of patients enrolled in trials of therapy for active disease ranged from 15 to 374 (Table 1). The duration of the trials ranged from 2 to 12 weeks. Eight trials for active ulcerative colitis were placebo controlled,9–11, 13, 20, 24, 25, 31 nine utilized sulfasalazine as a control,12, 17, 19, 21, 22, 30, 37, 51, 53 three compared mesalazine and balsalazide,45, 52, 54 two compared oral mesalazine with topical or combined oral/topical mesalazine42, 50 and one compared mesalazine and olsalazine.46
Five randomized trials of olsalazine vs. placebo for active ulcerative colitis met the inclusion criteria (Table 3).9, 10, 13, 20, 24 In two studies, the adverse event rates were not reported.13, 24 In the remaining studies, 23–83% of olsalazine-treated patients noted an adverse event, compared with 5–80% of patients who received placebo. In all cases, the event rate was numerically higher in the olsalazine arm; however, none of these differences was reported to be statistically significant. When individual events were specified, diarrhoea accounted for 0–40% of adverse events. The prevalence of study withdrawals due to adverse events was in the range 0–28% in olsalazine-treated patients vs. 0–27% in those treated with placebo, and no statistically significant differences in these rates were noted. Of 10 study withdrawals occurring in olsalazine-treated patients, five were due to diarrhoea.
Table 3. Adverse events and study withdrawals due to adverse events in trials for active ulcerative colitis
|Olsalazine vs. placebo|
| Selby et al.9||Olsalazine, 2 g||Placebo|| 5 (25%)|| 1 (5%)||0|| 0|
| Meyers et al.10||Olsalazine, 0.75 g||Placebo|| || ||1 (7%)|| 1 (5%)|
|Olsalazine, 1.5 g||Placebo||38 (83%)||16 (80%)||1 (6%)|| 1 (5%)|
|Olsalazine, 3 g||Placebo|| || ||1 (7%)|| 1 (5%)|
| Hetzel et al.13||Olsalazine, 2 g||Placebo||NR||NR||2 (13%)|| 4 (27%)|
| Feurle et al.20||Olsalazine, 2 g||Placebo||12 (23%)|| 9 (17%)||3 (6%)|| 0|
| Zinberg et al.24||Olsalazine, 3 g||Placebo||NR||NR||2 (28%)|| 0|
|Mesalazine vs. placebo|
| Schroeder et al.11||Asacol, 1.6 g||Placebo|| 8 (73%)||23 (61%)||1 (9%)|| 2 (5%)|
|Asacol, 4.8 g||Placebo||21 (55%)||23 (61%)||1 (3%)|| 2 (5%)|
| Sninsky et al.25||Asacol, 1.6 g||Placebo||NR||NR||0|| 0|
|Asacol, 2.4 g||Placebo||NR||NR||2 (4%)|| 0|
| Hanauer et al.31||Pentasa, 1 g||Placebo||15 (16%)||20 (22%)||5 (5%)||11 (12%)|
|Pentasa, 2 g||Placebo||13 (13%)||20 (22%)||9 (9%)||11 (12%)|
|Pentasa, 4 g||Placebo||17 (18%)||20 (22%)||7 (7%)||11 (12%)|
|Olsalazine vs. sulfasalazine|
| Ewe et al.12||Olsalazine, 1.5 g||Sulfasalazine, 3 g|| 4 (20%)||12 (60%)||0|| 1 (5%)|
| Willoughby et al.19||Olsalazine, 3 g||Sulfasalazine, 3 g||NR||NR||2 (8%)|| 2 (7%)|
| Rao et al.22||Olsalazine, 2 g||Sulfasalazine, 3–4 g|| 2 (10%)|| 4 (23%)||0|| 0|
| Ferry et al.30||Olsalazine, 30 mg/kg||Sulfasalazine, 60 mg/kg||11 (39%)||13 (46%)||0|| 4 (14%)|
|Mesalazine vs. sulfasalazine|
| Riley et al.17||Asacol, 0.8 g||Sulfasalazine, 2 g||NR||NR||0|| 2 (11%)|
|Asacol, 2.4 g||Sulfasalazine, 2 g||NR||NR||0|| 2 (11%)|
| Rachmilewitz21||Mesasal, 1.5 g||Sulfasalazine, 3 g||16 (14%)||25 (24%)||7 (6%)|| 8 (8%)|
| Munakata et al.37||Pentasa, 1.5 g||Sulfasalazine, 3 g|| 7 (13%)||19 (33%)*||3 (6%)|| 9 (16%)*|
|Balsalazide vs. sulfasalazine|
| Green et al.51||Balsalazide, 6.75 g||Sulfasalazine, 3 g||27 (96%)||27 (93%)||2 (7%)|| 9 (31%)*|
| Mansfield et al.53||Balsalazide, 6.75 g||Sulfasalazine, 3 g||17 (65%)||21 (88%)||1 (4%)|| 9 (38%)*|
|Balsalazide vs. mesalazine|
| Green et al.45||Balsalazide, 6.75 g||Asacol, 2.4 g||24 (48%)||35 (71%)*||1 (2%)|| 1 (2%)|
| Levine et al.52||Balsalazide, 2.25 g||Asacol, 2.4 g||27 (54%)||26 (51%)||5 (10%)|| 5 (10%)|
|Balsalazide, 6.75 g||Asacol, 2.4 g||23 (43%)||26 (51%)||1 (2%)|| 5 (10%)|
| Pruitt et al.54||Balsalazide, 6.75 g||Asacol, 2.4 g||45 (54%)||57 (64%)||3 (4%)|| 5 (7%)|
|Mesalazine vs. other|
| Safdi et al.42||Asacol, 2.4 g||Mesalazine enema, 4 g|| 9 (41%)|| 3 (17%)||1 (5%)|| 0|
|Asacol, 2.4 g plus||Mesalazine enema, 4 g|| 9 (45%)|| 3 (17%)||0|| 0|
| mesalazine enema, 4 g|| || || || || |
| Kruis et al.46||Claversal, 3 g||Olsalazine, 3 g||29 (36%)||41 (47%)||9 (11%)||11 (13%)|
| Vecchi et al.50||Salofalk, 4 g||Salofalk, 2 g plus Salofalk enema, 2 g|| 5 (8%)|| 4 (6%)||1 (1%)|| 1 (2%)|
In clinical trials of active ulcerative colitis comparing mesalazine with placebo,11, 25, 31 the fraction of patients with adverse events ranged from 13% to 73% in the treatment arms vs. 22% to 61% in the placebo arms (Table 3). In all but one case, the percentage of patients with adverse events was numerically higher in the placebo arm. Of 333 mesalazine-treated patients in these trials, 74 experienced adverse events (22%), compared with 43 of 128 placebo-treated patients (34%). The fraction of patients withdrawn due to adverse events was in the range 0–9% in the treatment arm and 0–12% in the placebo arm (Table 3). Twenty-five of 436 mesalazine-treated patients withdrew due to adverse events (6%) vs. 13 of 177 patients in the placebo arms (7%). There was no statistically significant difference in adverse events or withdrawal in any of the studies.
Four trials compared olsalazine with sulfasalazine for the treatment of active ulcerative colitis (Table 3).12, 19, 22, 30 Adverse event rates were in the range 10–39% in patients receiving olsalazine vs. 23–60% in sulfasalazine-treated ulcerative colitis patients. Although, in all cases, the frequency of adverse events was numerically higher in those treated with sulfasalazine, no statistically significant differences were reported. Diarrhoea accounted for 0–50% of patients reporting adverse events with olsalazine. Between 0% and 8% of olsalazine-treated patients withdrew from the study due to adverse events, compared with 0–14% of those treated with sulfasalazine. The only two study withdrawals reported in patients who received olsalazine were due to diarrhoea.
In trials of active ulcerative colitis comparing mesalazine or balsalazide with sulfasalazine,17, 21, 37, 51, 53 the adverse event rates were in the range 13–96% in the treatment arms vs. 24–93% in the sulfasalazine-treated groups (Table 3). In one trial, the percentage of patients with adverse events was significantly higher in the sulfasalazine arm.37 Of 221 patients treated with mesalazine or balsalazide for whom data were available, 67 experienced adverse events (30%), compared with 92 of 215 patients treated with sulfasalazine (43%). The fraction of patients withdrawn due to adverse events was in the range 4–7% in the mesalazine/balsalazide treatment arms vs. 8–38% in the sulfasalazine groups. In all cases, the percentage of withdrawals was numerically higher in the sulfasalazine groups, and in three trials these differences were statistically significant.37, 51, 53 Thirteen of 262 patients in the mesalazine/balsalazide arms were withdrawn (5%), compared with 37 of 234 treated with sulfasalazine (16%).
Three studies compared balsalazide and mesalazine for active ulcerative colitis (Table 3).45, 52, 54 Between 43% and 54% of balsalazide-treated patients experienced adverse events vs. 51–71% of mesalazine-treated patients. A total of 119 of 237 balsalazide-treated patients developed adverse events (50%), compared with 118 of 189 mesalazine-treated patients (62%). In one study, this difference was statistically significant (48% vs. 71%, P = 0.024).45 However, when the study authors considered only adverse events that were thought to be possibly or probably related to therapy, the differences were no longer statistically significant (21% vs. 11%, P = 0.24).45 The fraction of patients withdrawn from these trials due to adverse events ranged from 2% to 10% in both treatment arms (Table 3). Four per cent (10/237) of balsalazide-treated patients were withdrawn vs. 6% (11/180) of mesalazine-treated patients.
In trials comparing oral mesalazine with other 5-ASA treatment modalities for active ulcerative colitis,42, 46, 50 between 8% and 41% of mesalazine-treated patients experienced side-effects vs. 6–47% of patients treated with topical or combination therapy or olsalazine. These differences were not statistically significant. One-quarter of patients treated with oral mesalazine developed adverse events (43/169) vs. 30% of those treated with other therapies (57/189). Between 0% and 11% of patients receiving oral mesalazine withdrew due to adverse events, compared with 0–13% in the other arms. A total of 11 of 169 oral mesalazine-treated patients were withdrawn due to adverse events (7%), compared with 12 of 189 patients in the other treatment arms (6%).
Maintenance of remission trials
The number of patients enrolled in maintenance therapy trials ranged from 67 to 334, and the study duration ranged from 12 weeks to 12 months (Table 2). Four maintenance trials were placebo controlled,32, 38, 40, 47 10 employed sulfasalazine as a control,14–16, 18, 23, 28, 29, 34, 36, 39 four were dose-ranging trials of mesalazine, olsalazine or balsalazide,26, 27, 33, 35 two compared balsalazide and mesalazine44, 49 and three compared mesalazine and another treatment.41, 43, 48
In four placebo-controlled trials of 5-ASA agents,32, 38, 40, 47 the fraction of patients with adverse events was in the range 20–41% in the 5-ASA-treated patients vs. 39–44% in the placebo arms (Table 4). Of the 280 patients treated with mesalazine with available data, 86 experienced adverse events (31%), compared with 79 of 189 placebo-treated patients (42%). Study withdrawals due to adverse events occurred in 2–14% of patients receiving mesalazine vs. 4–33% of placebo-treated patients. In the placebo-controlled olsalazine trial, the withdrawal rate was numerically higher in patients receiving olsalazine, with diarrhoea severe enough to necessitate withdrawal from the study in 16% of those receiving active drug vs. 2% of those receiving placebo.32 In all but one case,47 withdrawals occurred numerically less frequently in mesalazine-treated patients, but these differences were not statistically significant. A total of 23 of 334 mesalazine-treated patients were withdrawn due to an adverse event (7%) vs. 40 of 247 placebo-treated patients (16%).
Table 4. Adverse events and study withdrawals due to adverse events in trials for the maintenance of ulcerative colitis remission
|Mesalazine/olsalazine vs. placebo|
| Wright et al.32||Olsalazine, 2 g||Placebo||12 (24%)|| 2 (4%)|| 6 (12%)|| 0|
| Miner et al.38||Pentasa, 4 g||Placebo||21 (20%)||45 (44%)||14 (14%)||34 (33%)|
| Anonymous40||Asacol, 0.8 g||Placebo||29 (32%)||34 (39%)|| 4 (4%)|| 4 (5%)|
|Asacol, 1.6 g||Placebo||36 (41%)||34 (39%)|| 2 (2%)|| 4 (5%)|
| Ardizzone et al.47||Asacol, 1.2 g||Placebo||NR||NR|| 3 (5%)|| 2 (4%)|
|Olsalazine vs. sulfasalazine|
| Ireland et al.14||Olsalazine, 1 g||Sulfasalazine, 2 g||21 (26%)||20 (24%)||16 (19%)|| 9 (11%)|
| Kiilerich et al.28||Olsalazine, 1 g||Sulfasalazine, 2 g||NR||NR|| 9 (8%)|| 6 (5%)|
| Rijk et al.29||Olsalazine, 2 g||Sulfasalazine, 4 g|| 9 (39%)|| 8 (35%)|| 3 (13%)|| 3 (13%)|
| Kruis et al.36||Olsalazine, 0.5 g||Sulfasalazine, 2 g|| 2 (5%)|| 4 (10%)|| 2 (5%)|| 1 (2%)|
|Olsalazine, 1.25 g||Sulfasalazine, 2 g|| 3 (8%)|| 4 (10%)|| 2 (5%)|| 1 (2%)|
|Olsalazine, 2 g||Sulfasalazine, 2 g|| 6 (18%)|| 4 (10%)|| 1 (3%)|| 1 (2%)|
| Nilsson et al.39||Olsalazine, 2 g||Sulfasalazine, 2 g||39 (24%)||26 (16%)||12 (7%)|| 8 (5%)|
|Balsalazide vs. sulfasalazine|
| McIntyre et al.15||Balsalazide, 2 g||Sulfasalazine, 2 g|| 2 (5%)||10 (26%)*|| 0|| 2 (5%)|
|Mesalazine vs. sulfasalazine|
| Mulder et al.16||Pentasa, 1.5 g||Sulfasalazine, 3 g||NR||NR|| 0|| 3 (9%)|
| Riley et al.18||Asacol, > 0.8 g||Sulfasalazine, > 2 g||NR||NR|| 0|| 1 (2%)|
| Rutgeerts23||Claversal, 0.75 g||Sulfasalazine, 1.5–2 g||24 (14%)||20 (12%)|| 9 (5%)|| 7 (4%)|
| Ardizzone et al.34||Claversal, 1 g||Sulfasalazine, 2 g||NR||NR|| 2 (5%)|| 3 (7%)|
|Balsalazide vs. mesalazine|
| Green et al.44||Balsalazide, 3 g||Asacol, 1.2 g||30 (61%)||30 (65%)|| 3 (6%)|| 1 (2%)|
| Kruis et al.49||Balsalazide, 3 g||Salofalk, 1.5 g||18 (38%)||20 (45%)|| 3 (6%)|| 4 (9%)|
|Balsalazide, 6 g||Salofalk, 1.5 g||21 (53%)||20 (45%)|| 2 (5%)|| 4 (9%)|
| Green et al.27||Balsalazide, 6 g||Balsalazide, 3 g||NR||NR|| 3 (6%)|| 6 (11%)|
| Giaffer et al.26||Balsalazide, 4 g||Balsalazide, 2 g||NR||NR||13 (19%)|| 8 (12%)|
| Travis et al.33||Olsalazine, 2 g||Olsalazine, 0.5 g||34 (55%)||30 (45%)||12 (19%)|| 6 (9%)|
|Olsalazine, 1 g||Olsalazine, 0.5 g||26 (40%)||30 (45%)|| 6 (9%)|| 6 (9%)|
| Fockens et al.35||Pentasa, 3 g||Pentasa, 1.5 g||10 (12%)||10 (11%)|| 4 (5%)|| 5 (6%)|
|Mesalazine vs. other|
| Kruis et al.41||Salofalk, 1.5 g||E. coli Nissle 1917, 0.2 g|| 8 (13%)|| 5 (9%)|| 1 (2%)|| 2 (3%)|
| Hawkey et al.43||Asacol, 1.6 g||Placebo||NR||NR|| 4 (4%)|| 6 (5%)|
|Asacol, 1.6 g||Zileuton, 2.4 g||NR||NR|| 4 (4%)|| 9 (8%)|
| Rembacken et al.48||Asacol, 1.2–2.4 g||E. coli Nissle 1917, two capsules b.d|| 7 (12%)|| 9 (16%)|| 2 (3%)|| 0|
In five trials of olsalazine vs. sulfasalazine for the maintenance of remission,14, 28, 29, 36, 39 adverse event rates were in the range 5–39% in ulcerative colitis patients receiving olsalazine vs. 10–35% in sulfasalazine-treated patients. Diarrhoea accounted for nearly 50% of all adverse events in olsalazine-treated patients in all but one study. No statistically significant differences in rates were noted. The frequency of study withdrawals due to adverse events varied between 3% and 19% in the olsalazine treatment arms, compared with 2–11% in those receiving sulfasalazine, and there were no significant differences. In those who received olsalazine, diarrhoea was the adverse event precipitating study withdrawal in 27 of 45 cases.
In five trials comparing mesalazine or balsalazide with sulfasalazine,15, 16, 18, 23, 34 only two reported the number of patients with adverse events15, 23 and all five reported study withdrawal data15, 16, 18, 23, 34 (Table 4). The fraction of patients with adverse events was in the range 5–14% in the treatment arms and 12–26% in the sulfasalazine arms. In only one trial was the percentage of patients with adverse events reported to be significantly different (26% for sulfasalazine vs. 5% for balsalazide, P = 0.017).15 Thirteen per cent of those treated with mesalazine or balsalazide experienced an adverse event (26/208) vs. 15% of those in the sulfasalazine arms (30/205). Between 0% and 5% withdrew due to adverse events in the treatment arms vs. 2–9% in sulfasalazine-treated patients. A total of 11 of 343 patients treated with 5-ASA withdrew from trials due to adverse events (3%), compared with 16 of 333 of those treated with sulfasalazine (5%).
Two trials compared balsalazide and mesalazine directly for the maintenance of remission in ulcerative colitis (Table 4).44, 49 The percentage of patients with adverse events was in the range 38–61% in balsalazide-treated patients vs. 45–65% in mesalazine-treated patients. No significant differences were reported. Exactly half of those treated with balsalazide experienced an adverse event (69/138), compared with 56% of mesalazine-treated patients (50/90). Study withdrawals occurred in 5–6% of patients receiving balsalazide vs. 2–9% of patients in the mesalazine arms. In total, 6% of patients in both groups were withdrawn (balsalazide, 8/138; mesalazine, 5/90).
In two dose-ranging maintenance trials of balsalazide, the number of patients with adverse events was not reported (Table 4).26, 27 Study withdrawals were in the range 6–19% in those treated with higher doses vs. 11–12% in those on lower doses. In a dose-ranging trial of mesalazine, adverse events occurred in 11–12% and withdrawals occurred in 5–6%.35 In a trial comparing three doses of olsalazine, adverse event rates were in the range 40–55% and study withdrawals occurred in 9–19%.33
Mesalazine was compared with other therapies in three trials (Table 4).41, 43, 48 Twelve per cent of mesalazine-treated patients (15/119) vs. 9% of probiotic-treated patients (14/115) developed adverse events, and study withdrawals occurred in 2% (3/119) and 3% (2/115), respectively.41, 48 In a trial comparing mesalazine with Zileuton, study withdrawals occurred in 4% (4/99) and 8% (9/113), respectively.43 No significant differences with respect to safety were reported.
Individual adverse events
The frequency of individual adverse events in the clinical trials was not uniformly available. In some cases, the frequency of individual adverse events was not reported at all; in other cases, only the frequency of individual events that led to study withdrawal was reported. Where data were available, the most commonly observed adverse events were diarrhoea, nausea/vomiting, headache, abdominal pain/dyspepsia, rash, fever, fatigue/weakness and arthralgia/myalgia (Table 5). As noted above, when individual adverse events were specified, diarrhoea was the most common event noted in olsalazine-treated patients. Most studies did not specify the severity of adverse events, and definitions of ‘serious’ or ‘severe’ events appeared to vary. Nevertheless, events labelled as serious, severe or requiring hospitalization were distinctly unusual.
Table 5. Frequency of individual adverse events in all trials of olsalazine, mesalazine and balsalazide for ulcerative colitis
|Diarrhoea||10 (2–28)||2 (1–9)||5 (2–19)|
|Nausea/vomiting|| 7 (1–15)||3 (2–16)||8 (3–17)|
|Headache|| 5 (1–14)||5 (1–30)||4 (1–13)|
|Abdominal pain/dyspepsia|| 4 (1–13)||4 (1–27)||6 (2–11)|
|Rash|| 4 (2–13)||2 (1–8)||2 (1–4)|
|Fever|| 3||3 (1–10)||4|
|Fatigue/weakness|| 2 (1–9)||4 (1–7)||4|
|Arthralgia/myalgia|| 1 (1–1)||7 (2–24)||3 (1–5)|
|Hepatic biochemical abnormalities|| ||2 (1–8)|| |
|Pruritus|| ||1 (1–7)||1|
Based on our systematic review of the safety of mesalazine, olsalazine and balsalazide in the treatment of ulcerative colitis, we conclude that all agents appear to be safe over the duration of the treatment trials. The fraction of mesalazine-treated patients experiencing adverse events or withdrawing from trials due to adverse events was similar to, or lower than, that seen in placebo-treated patients. Although the adverse event rate in olsalazine-treated patients was higher than that in placebo-treated patients, these differences were not reported to be statistically significant. However, these individual trials were powered to detect differences in efficacy, not adverse events. No fully published placebo-controlled trials of balsalazide exist. In almost all cases, adverse events and withdrawals due to adverse events occurred less frequently in patients treated with mesalazine or balsalazide than in those treated with sulfasalazine. In the case of olsalazine, adverse event rates were numerically lower than those seen with sulfasalazine in virtually all of the trials for the induction of remission, but this was not seen in the trials for the maintenance of remission of ulcerative colitis. In one of five trials comparing balsalazide and mesalazine, mesalazine-treated patients experienced significantly more adverse events than patients treated with balsalazide,45 but there were no significant differences in the frequency of patients with adverse events thought to be drug related. Moreover, four other trials comparing balsalazide and mesalazine found no difference in adverse events between the two treatments, and the rates of study withdrawals in the two groups were similar.44, 49, 52, 54
It should be noted that the rate of adverse events in different trials can vary widely as a result of the methodology of collection of the data. Prospective trials that routinely query patients for adverse events on a scheduled basis identify events that are missed in trials that rely on spontaneous event reporting. The methodology used was not always clear in the studies reviewed. The primary reason why we did not perform a meta-analysis was concern about the heterogeneity of the definition of adverse events and of the collection of such data. Because of this variability in adverse event reporting, the number of patients who withdrew from studies due to an adverse event may be a more reliable marker of clinically significant side-effects.
Our systematic review was restricted to fully published randomized controlled trials. Abstracts were not included because frequently they did not contain sufficient data on adverse event rates or withdrawals due to adverse events. This introduces the possibility that publication bias could have significantly influenced our conclusions (i.e. important trends in adverse event rates may have been missed); however, we do not believe this to be the case.
Rare, but potentially serious, adverse events are not well identified in prospective studies because of the relatively small number of patients enrolled and the limited duration of the treatment trials. When serious events have occurred following mesalazine therapy (e.g. hepatitis, blood dyscrasias, pancreatitis, pleuropericarditis, interstitial nephritis), they have usually been reported as isolated cases. The potential relationship of the event to the drug and the magnitude of the increased risk can only be assessed in studies of very large numbers of patients. This usually occurs as post-marketing surveillance, relying on spontaneous event reporting, or in population-based case–control studies, in which large numbers of patients can be included. A French pharmacovigilance study of mesalazine microgranules (Pentasa) reported between 6.6 and 9.0 adverse events per million treatment days over a 2-year period, including cases of pancreatitis, hepatitis, pericarditis and haematological disturbances.55 A British pharmacovigilance study of mesalazine (Asacol, Pentasa or Salofalk) and sulfasalazine over an 8-year interval identified 393 adverse events per million prescriptions for mesalazine and 514 events per million for sulfasalazine.56 The odds ratio for any adverse event on sulfasalazine (relative to mesalazine) was 1.31 (95% confidence interval, 1.22–1.40). However, the relative risk varied by individual adverse event. Interstitial nephritis and pancreatitis were significantly more likely to be reported with mesalazine, whereas blood dyscrasias occurred more frequently with sulfasalazine.56 There are no pharmacovigilance studies available for olsalazine or balsalazide. These pharmacovigilance studies rely on spontaneous reporting and, as such, are likely to underestimate the actual rate of adverse events; on the other hand, the design of these studies makes a causal relationship between the medication and adverse event difficult to prove.
In summary, the 5-ASA preparations mesalazine, olsalazine and balsalazide appear to be safe. Adverse event rates in mesalazine-treated patients were similar to those seen with placebo and lower than those seen with sulfasalazine. Although olsalazine resulted in watery diarrhoea in some patients, the overall rate of adverse events or study withdrawals was not significantly different from that seen with placebo or sulfasalazine. Balsalazide appears to have fewer adverse events than sulfasalazine.
This work was supported by an unrestricted grant from Procter and Gamble Pharmaceuticals. Drs Loftus and Kane have received research support from Procter and Gamble Pharmaceuticals.