Nineteen RCTs (n=2,208) were included.
All 19 trials were blinded, 10 had adequate allocation concealment, 13 had undertaken an intention-to-treat analysis and 16 had adequate follow-up. The percentage of withdrawals ranged from 3 to 50% across the trials.
Delayed-release mesalamine versus placebo (2 RCTs, n=245): both RCTs showed that higher doses of mesalamine significantly improved response compared with placebo. The first RCT found that 4.8 g/day mesalamine significantly increased complete or partial response compared with placebo (74% versus 18% with placebo, P<0.001; number-needed-to-treat 2) but found less difference between lower dose mesalamine and placebo (27% versus 18%, P not reported). The second trial found that 2.4 and 1.6 g/day mesalamine significantly increased complete or partial response compared with placebo (49% and 43% for higher and lower dose, respectively, versus 23% with placebo, P=0.003 and P=0.03l for drug versus placebo in per protocol analysis). Controlled-release mesalamine versus placebo (1 RCT, n=374): medium and higher (2 and 4 g/day) doses of controlled mesalamine were significantly more beneficial than placebo in terms of remission, complete relief, or a marked improvement in symptoms. There were no significant differences between low-dose (1 g/day) mesalamine and placebo.
Olsalazine versus placebo (4 RCTs, n=216): the results of the trials comparing olsalazine versus placebo were mixed. Two of the RCTs showed no significant differences between treatment with olsalazine and placebo; the other two showed borderline significant differences relative to placebo, but the differences were based on a small number of participants and, therefore, were difficult to interpret.
Balsalazide versus placebo (1 RCT, n=180): there were no significant difference in response rate between treatment with balsalazide (4.5 or 6.75 g/day) and placebo.
Sulfasalazine versus delayed-release mesalazine, controlled-release mesalamine, balsalazide, olsalazine or mesalazine (7 RCTs, n=600): there were no significant differences between treatment with sulfasalazine and any of the comparators in terms of any of the outcomes assessed.
Olsalazine versus mesalazine (1 RCT, n=168): there was significant difference between treatments in remission rates.
Asacol (mesalamine) versus balsalazide (3 RCTs, n=428): two of the studies showed no significant differences between asacol and balsalazide on any of the outcomes assessed; the third study found that balsalazide significantly increased remission rates compared with asacol (62% versus 37%, P<0.05).