Background: Ulcerative colitis has been suggested to be caused by infection and there is circumstantial evidence linking Escherichia coli with the condition. Our aim was to find out whether the administration of a non-pathogenic strain of E. coli (Nissle 1917) was as effective as mesalazine in preventing relapse of ulcerative colitis. We also examined whether the addition of E. coli to standard medical therapy increased the chance of remission of active ulcerative colitis.
Methods: This was a single-centre, randomised, double-dummy study in which 120 patients with active ulcerative colitis were invited to take part. 116 patients accepted; 59 were randomised to mesalazine and 57 to E. coli. All patients also received standard medical therapy together with a 1-week course of oral gentamicin. After remission, patients were maintained on either mesalazine or E. coli and followed up for a maximum of 12 months. A two-stage, conditional, intention-to-treat analysis was done.
Findings: 44 (75%) patients in the mesalazine group attained remission compared with 39 (68%) in the E. coli group. Mean time to remission was 44 days (median 42) in the mesalazine group and 42 days (median 37) for those treated with E. coli. In the mesalazine group, 32 (73%) patients relapsed compared with 26 (67%) in the E. coli group. Mean duration of remission was 206 days in the mesalazine group (median 175) and 221 days (median 185) in the E. coli group.
Interpretation: Our results suggest that treatment with a non-pathogenic E. coli has an equivalent effect to mesalazine in maintaining remission of ulcerative colitis. The beneficial effect of live E. coli may provide clues to the cause of ulcerative colitis.