Mirikizumab improved histologic outcomes in patients with moderately to severely active Crohn's disease, and biopsy findings identified inflammation that endoscopy did not detect, according to an analysis of 1,065 patients in the primary analysis set of the phase 3 VIVID-1 trial, published in the Journal of Crohn's and Colitis.
At week 12, histologic response was reported in 53% of mirikizumab-treated patients vs. 31% of placebo-treated patients (P < .0001), and histologic remission in 22% vs. 12% (P = .0013). At week 52, histologic response was higher with mirikizumab than with the active comparator ustekinumab, at 58% vs. 49% (P = .0075), although histologic remission rates were nearly identical, at 30% and 29% (P = .69).
VIVID-1 was “the first phase 3 trial in [Crohn's disease] in which intestinal biopsies were systematically collected and reported from five segments at multiple timepoints,” the authors wrote. The response and remission definitions used in the trial were aligned with the position statement of the European Crohn's and Colitis Organisation and with expert consensus recommendations, they noted.
Mirikizumab, a humanized IgG4 monoclonal antibody that binds the p19 subunit of interleukin-23, is approved for Crohn's disease and ulcerative colitis. The authors said their data imply “the potential benefit of IL23p19 over IL23p40 inhibition, especially among patients with previous biologic failure.”
Methods
VIVID-1 (NCT03926130) was a multicenter, randomized, double-blind, double-dummy, parallel-group, placebo- and active-controlled trial comparing mirikizumab with placebo and ustekinumab. It enrolled patients with demonstrated intolerance, inadequate response, or loss of response to conventional and/or biologic therapies.
Investigators collected two biopsies from each of five intestinal segments — the terminal ileum and four colonic segments — at screening and at weeks 12 and 52, or at an early termination visit. Biopsies were obtained from the edge of ulcers or erosions when present, or from the most macroscopically involved mucosa. Histologic response and remission were assessed using prespecified criteria based on the Robarts Histopathology Index and the Global Histologic Disease Activity Score, and the criteria had to be met in all five segments — a stringency that helps explain the comparatively low remission rates. All histologic scoring was performed by a single central reader blinded to treatment assignment and visit sequence.
Histologic response and remission at weeks 12 and 52 in the primary analysis set were the prespecified endpoints. Combined histologic-endoscopic outcomes, segmental analyses, subgroup analyses by prior biologic failure, and the analysis linking week 12 findings to week 52 outcomes were all post hoc. Unless otherwise noted, reported P values are nominal and were not adjusted for multiplicity.
Results
At baseline, 90% of patients in the primary analysis set had active histologic disease. Histology and endoscopy identified the same disease location in 59% of patients. Among 422 patients classified as having isolated colonic disease on endoscopy, 26% had additional histologic inflammation in the terminal ileum. Nearly half of that group had no endoscopic activity in the ileum, and about a third had ileal segments recorded as missing by all three central endoscopy readers — cases in which the authors said biopsies were likely taken at the ileocecal valve and read as colonic. Central histologic evaluation nonetheless confirmed active microscopic inflammation in the ileal mucosa across the group.
Early histologic outcomes were similar between the two active treatments, at 53% vs. 51% for response (P = .58) and 22% vs. 20% for remission (P = .44) at week 12. The week 52 histologic response advantage for mirikizumab was more pronounced among patients with previous biologic failure (P = .0064), a post hoc subgroup analysis using an unadjusted test.
Among patients with active histologic disease at baseline, combined histologic-endoscopic response at week 52 was numerically higher with mirikizumab than ustekinumab in the overall population, at 41% vs. 34% (P = .06). In patients with prior biologic failure, the difference was statistically significant, at 40% vs. 28% (P = .024), on a nominal, unadjusted P value from a post hoc analysis. Combined histologic-endoscopic remission rates were comparable between treatment arms.
Segmental analyses showed statistically significant reductions in histologic inflammation with mirikizumab compared with placebo across all five intestinal segments. Compared with ustekinumab, mirikizumab showed greater reductions in Robarts Histopathology Index score in the transverse colon (P = .004), the left colon (P = .025), and across all four colonic segments combined (P = .003).
Agreement between the two assessment methods was incomplete. Histologic and endoscopic response were congruent in 62.1% of patients at week 12, which the authors characterized as fair agreement (κ = 0.27), and there was no agreement between histologic response and clinical outcomes. The authors described histologic and endoscopic findings as “consistently discordant” across all segments and anatomical locations, while elsewhere characterizing the relationship as moderate concordance.
Early histologic response appeared informative only in a specific group. Among mirikizumab-treated patients with active histologic disease at baseline who achieved an endoscopic response at week 12, histologic response did not add prognostic information: week 52 outcomes did not differ significantly between histologic responders and nonresponders. Among week 12 endoscopic nonresponders, however, those with a histologic response had higher week 52 rates of histologic response (68.9% vs. 39.2%, P < .0001), histologic remission (34.4% vs. 12.8%, P < .0001), endoscopic response (43.1% vs. 31.4%, P = .026), and endoscopic remission (20.5% vs. 12.3%, P = .0395), along with greater reductions in fecal calprotectin.
The authors suggested that microscopic healing at the ulcer edge may occur before endoscopic ulcer closure, meaning histologic improvement may be detectable before macroscopic healing is visible. In that context, they wrote, “Therefore, biopsies and histologic evaluation may guide treatment decisions for endoscopic non-responders.”
Optimal fecal calprotectin cutoffs at week 52 were 346 µg/g for histologic response, 158 µg/g for histologic remission, 269 µg/g for combined histologic-endoscopic response, and 258 µg/g for combined remission.
Limitations
Several limitations should be considered. Analyses in patients with active histologic disease at baseline were post hoc, although the authors noted that results for the prespecified and post hoc populations were nearly identical. The histologic scoring systems used in the trial have not been validated for Crohn's disease, and optimal biopsy sampling remains uncertain because inflammation is patchy and may differ by segment. Obtaining biopsy specimens precisely during endoscopy is technically challenging, and the protocol did not include specific instructions for patients with prior bowel resections. The study did not evaluate relationships between mucosal and transmural inflammation or longer-term outcomes such as surgery and hospitalization.
The authors said longer-term follow-up will be needed to determine whether histologic endpoints predict clinically meaningful outcomes. They concluded that VIVID-1 supports the idea that histologic evaluation helps identify small bowel inflammation not detected by endoscopy and may provide additional information relevant to treatment decisions.
The study was funded by Eli Lilly and Company. Several authors reported consulting, advisory, speaking, research, or employment relationships with pharmaceutical companies, including Eli Lilly and Company.