Mirikizumab held Crohn's responses through a second year, without a control

Patients with Crohn's disease who kept taking mirikizumab for two years mostly held on to the responses they had at one year, according to the first results from the VIVID-2 open-label extension [s1]. At week 104, using the conservative accounting that counts missing data as failure, 60·5% of patients had an endoscopic response and 73·4% were in symptomatic remission [s1] — reassuring durability numbers, with the important caveat that this extension had no placebo group, so it can describe what happened over time but cannot separate the drug's effect from the disease's natural ups and downs.

Mirikizumab is a monoclonal antibody that blocks interleukin-23p19, part of an inflammatory pathway central to inflammatory bowel disease. It was already established in ulcerative colitis, and its pivotal Crohn's trial, VIVID-1, tested it as induction and maintenance treatment over 52 weeks in patients who had failed conventional or biologic therapy [s2]. VIVID-2 is what happened next: the long-term extension in which patients from VIVID-1 continued on the drug and were followed to two years [s1].

What the controlled trial had shown

VIVID-1 is the source of the placebo-controlled evidence, and it is worth stating plainly because the extension cannot. In that phase 3 trial, both co-primary endpoints were met: a composite of patient-reported clinical response at week 12 and endoscopic response at week 52 was reached by 220 (38·0%) of 579 patients on mirikizumab versus 18 (9·0%) of 199 on placebo, and a composite with CDAI clinical remission at week 52 by 263 (45·4%) of 579 versus 39 (19·6%) of 199 (both p<0·0001) [s2]. That is the comparison that shows the drug works; everything in the extension builds on it.

VIVID-1 was also unusually rigorous in its design: it was a "treat-through" study that additionally included an active comparator arm, randomising 1150 treated patients 6:3:2 to mirikizumab, ustekinumab (about 6 mg/kg intravenously then 90 mg subcutaneously every 8 weeks), or placebo across 324 sites in 33 countries [s2]. Mirikizumab maintenance was 300 mg subcutaneously every 4 weeks after intravenous induction [s2]. Having a placebo and an active drug in the same trial is what lets the 9·0% placebo figure anchor the mirikizumab result; the extension, by design, keeps only the mirikizumab arm and drops that anchor.

What the extension found

Of the patients randomised to mirikizumab in VIVID-1 who rolled into VIVID-2, 465 were in the safety population and 430 in the efficacy population for this interim analysis [s1]. That group split into 251 who had an endoscopic response at week 52 of VIVID-1 and 179 who did not and were reinduced with 900 mg intravenously before returning to subcutaneous dosing [s1].

At week 104, in the full efficacy population, 60·5% achieved an endoscopic response, 37·9% endoscopic remission and 73·4% CDAI clinical remission under modified nonresponder imputation; the numbers were higher — 67·1%, 41·7% and 80·9% — when only observed cases were counted [s1]. Among patients who had reached their week-52 endpoint, most kept it: 81·8% maintained an endoscopic response and 86·9% maintained CDAI remission at week 104 (mNRI) [s1].

The more interesting figure is for the patients the drug had not yet worked in. Among week-52 endoscopic nonresponders, 30·9% achieved an endoscopic response at week 104 after reinduction [s1] — evidence that a year of treatment is not always long enough to declare failure, though without a control arm some of that gain could reflect regression to the mean rather than the extra dose.

It is worth stressing that these are interim results from an ongoing extension, not a final report [s1]. Interim analyses can shift as more patients reach the timepoint and as the sicker or less-responsive patients drop out, which tends to flatter the numbers that remain — another reason uncontrolled durability data should be read as a trend to confirm rather than a settled figure. The gap between the modified-nonresponder-imputation and observed-case results — for instance 60·5% versus 67·1% for endoscopic response — is a direct measure of how much the accounting choice moves the headline [s1].

Safety

During the second year, 64·7% of patients had a treatment-emergent adverse event and 7·7% had a serious one [s1]; the authors describe the profile as consistent with what is already known about the drug [s1]. For context, in the controlled VIVID-1 trial serious adverse events were actually more common on placebo (17·1%) than on mirikizumab (10·3%), a reminder that active Crohn's disease itself carries risk [s2]. Both trials were funded by Eli Lilly and Company [s1][s2]; the company's framing of "durable" two-year benefit rests on uncontrolled data and should be read as such.

What it means

The takeaway is narrow but real: patients who respond to mirikizumab and stay on it tend to keep responding into a second year, and some initial nonresponders catch up. What the extension cannot tell you is how many would have stayed well anyway, because there was no comparison group — the standard limitation of open-label extensions. The placebo-controlled proof lives in VIVID-1; VIVID-2 adds duration, not certainty. Anyone weighing long-term IL-23 blockade for Crohn's should keep those two things separate.

Sources

• [s1] Mirikizumab Long-Term Efficacy and Safety in Patients With Crohn's Disease: Results From the VIVID-2 Open-Label Extension Trial. Clinical Gastroenterology and Hepatology, 5 March 2026. https://doi.org/10.1016/j.cgh.2026.01.049
• [s2] Efficacy and safety of mirikizumab in patients with moderately-to-severely active Crohn's disease (VIVID-1). The Lancet, 21 November 2024. https://doi.org/10.1016/S0140-6736(24)01762-8