BACKGROUNDCobitolimod is a topically administered, DNA-based oligonucleotide that activates Toll-like receptor 9 (TLR9), and previous research has shown clinical efficacy in patients with moderate-to-severe ulcerative colitis. Here we assessed the efficacy and safety of different dose regimens of cobitolimod for induction therapy in patients with moderate-to-severe, left-sided ulcerative colitis.METHODSCONDUCT was a randomised, double-blind, five-arm, placebo-controlled, dose-ranging phase 2b study that recruited patients with moderate-to-severe, left-sided ulcerative colitis, with inadequate response to conventional or biological therapies, from 91 hospitals or outpatient clinics in 12 European countries. Eligible patients had a Mayo score of 6-12 with a centrally read endoscopic subscore (modified to exclude friability from grade 1) of 2 or higher and no individual subscore of less than 1, and confirmation of left-sided disease. Patients were randomised (1:1:1:1:1; block size of ten) via a computer-generated schedule and centralised interactive voice and web response system to receive rectal enemas of cobitolimod at 31 mg, 125 mg, or 250 mg at weeks 0 and 3 (2 × 31 mg, 2 × 125 mg, and 2 × 250 mg groups), cobitolimod at 125 mg at weeks 0, 1, 2, and 3 (4 × 125 mg group), or placebo. Randomisation was stratified by current glucocorticosteroid and previous tumour necrosis factor inhibitor treatment. Patients and all study personnel were masked to treatment allocation. The primary endpoint was the proportion of patients achieving clinical remission (Mayo subscores for rectal bleeding of 0, for stool frequency of 0 or 1 [with ≥1-point decrease from baseline], and for endoscopy of 0 or 1 [excluding friability]) at week 6. The primary analysis (based on intention to treat) and safety analysis were done in all randomly assigned patients who received at least one dose of active study drug or placebo. In this exploratory study, statistical tests were one-sided; p values of less than 0·10 were regarded as statistically significant, with no adjustment for multiplicity. This study is registered with ClinicalTrials.gov, NCT03178669, and is completed; the results here represent the final analysis.FINDINGS213 patients were randomly assigned between June 30, 2017, and June 26, 2019. Of these, 211 patients received study treatment: 40 in the cobitolimod 2 × 31 mg group, 43 in the 2 × 125 mg group, 42 in the 4 × 125 mg group, 42 in the 2 × 250 mg group, and 44 in the placebo group. A greater proportion of patients were in clinical remission at week 6 in the cobitolimod 2 × 250 mg group than in the placebo group (nine [21%] of 42 patients vs three [7%] of 44; odds ratio [OR] 3·8 [80% CI 1·5-9·5]; one-sided p=0·025). We identified no significant difference in the proportion of patients with clinical remission in the cobitolimod 2 × 31 mg group (five [13%] of 40 patients; OR 2·0 [80% CI 0·7-5·5], p=0·18), 2 × 125 mg group (two [5%] of 43; 0·7 [0·2-2·2], p=0·66), or 4 × 125 mg (four [10%] of 42; 1·4 [0·5-3·9], p=0·33) compared with the placebo group. Treatment-emergent adverse events occurred in 21 (48%) patients in the placebo group, ten (25%) patients in the cobitolimod 2 × 31 mg group, 17 (40%) patients in the 2 × 125 mg group, 15 (36%) patients in the 4 × 125 mg group, and 18 (43%) patients in the 2 × 250 mg group. Severe adverse events occurred in eight (4%) of 211 patients (worsening of ulcerative colitis [seven patients] and abdominal hernia and wound dehiscence [one patient]). Ten patients (two [5%] in the placebo group, two [5%] in the cobitolimod 2 × 31 mg group, two [5%] in the 4 × 125 mg, and four [10%] in the 2 × 250 mg group) had a total of 13 serious adverse events; these were worsening of ulcerative colitis (eight events) and pruritus, rash, abdominal hernia, fascia dehiscence, and deep vein thrombosis (one event each). One patient in the placebo group died from total organ failure after receiving a colectomy for a serious adverse event of disease worsening.INTERPRETATIONTwo topical administrations of cobitolimod 250 mg were well tolerated and more effective than placebo in inducing clinical remission 6 weeks after the start of treatment. TLR9 activation is a promising novel therapeutic target in ulcerative colitis and warrants further testing, with phase 3 trials of cobitolimod planned.FUNDINGInDex Pharmaceuticals.