UEG Week 2026 was the busiest IBD meeting I can remember. Here is what I have covered so far, all in one place. Eight posts. Long-acting antibodies, diet, obefazimod, VERDICT, STRIDE-III, perianal Crohn's and subcut risankizumab, plus a full Substack piece on what VERDICT and STRIDE-III mean for the clinic. Next up is combination therapy. DUET, TARGET-CD, and what I actually do in clinic. Links to every post below.

FUZION, guselkumab in perianal fistulising Crohn's, presented to week 48 by Ailsa Hart as a late-breaker at UEG. It is the first large phase 3 trial dedicated to perianal fistulas to show a systemic drug works since ACCENT II with infliximab in 2004. The local stem-cell option, darvadstrocel, worked in ADMIRE-CD in 2016, but failed its confirmatory trial and has left the EU market. 286 adults with at least one draining perianal fistula, confirmed on blinded central MRI, were randomised 2:2:1 to guselkumab 200 mg IV at weeks 0, 4 and 8, then 100 mg subcutaneous every 8 weeks or 200 mg every 4 weeks, or placebo. Setons came out between weeks 4 and 12. Most were luminally quiet (78% had a CDAI of 220 or less) and nearly two-thirds had already had a biologic. At week 24, 28% on 100 mg and 27% on 200 mg reached combined fistula remission, against 10% on placebo: every external opening closed, no drainage even on gentle compression, and no collection over 2 cm on MRI. Response, at least 50% fewer draining fistulas, was 33% and 36% against 14%. At week 48, combined remission was 35% on 100 mg and 26% on 200 mg. That isn't a dose effect. 49% of the 100 mg group (55 of 113) hadn't responded by week 24 and were escalated, blind, to 200 mg every 4 weeks, but stayed in the 100 mg group, while non-responders on 200 mg had a sham escalation. So the 35% reflects stepping up the people who hadn't responded. There is no placebo comparison at week 48, because placebo non-responders switched to guselkumab. Safety: serious adverse events 16% on 100 mg, 8% on 200 mg and 14% on placebo. Serious infection (12%) and anal abscess (12%) were higher on 100 mg, in small numbers. Two cancers on 200 mg were judged unrelated. Guselkumab has an excellent safety profile across many studies now, so I have zero concerns here. My read: about 1 in 4 heal against 1 in 10 on placebo, and it holds to week 48. No filing for perianal disease has been announced yet, but surely will follow. It is fantastic to have controlled data to support p19 therapies for fistulising Crohn's - such a hard phenotype to treat especially after TNF failure. Most of us will still use infliximab plus azathioprine first in these patients. The MRI part of the primary endpoint only rules out a collection over 2 cm, and so few patients had one that combined and clinical remission came out identical. Guselkumab improved MRI fistula activity more than placebo at week 24 (MAGNIFI-CD change −1.4 and −1.7 v −0.4, nominal p 0.034 and 0.004), but we haven't yet seen MRI response, remission or week 48 scans.

The paper, Lancet Gastroenterology & Hepatology, 19 Oct 2026: doi.org/10.1016/S2468-1253(2… Edinburgh IBD: writing the next decade of IBD care: edinburghibd.com/

First thoughts on VERDICT ... pending publication of the full manuscript by @vipuljairath and colleagues. 672 pts with mod-sev UC randomised to 1 of 3 targets on a vedo-based algorithm (stepped up to q4w if needed), with central reading of endoscopy and histology: • G1: symptoms • G2: + endoscopic remission (MES 0–1) • G3: + histological remission (Geboes <2B.0) The primary outcome was time to a UC complication over up to 80 weeks from reaching the target. Mean age was 41, with eight years of disease, a Mayo score of 8.7, nearly two-thirds Mayo 3, almost half on steroids and a fifth after an advanced therapy. By week 48, 74% (G1), 56% (G2) and 52% (G3) had reached their targets. Treating to endoscopy made a huge difference: 22% of G1 had a UC complication, against 7% in G2 and 9% in G3. The primary comparison, histology (G3) against symptoms (G1), gave a hazard ratio of 0.33 (95% CI 0.17 to 0.67), and endoscopy against symptoms (G2 vs G1) gave 0.30 (0.14 to 0.64). In the Q&A, Vip said the effect was greater still at Mayo 0 (data not shown). About half the complications were rescue therapy for a flare; only one was an admission and none a colectomy. We don't know what happened to the patients who did not meet their targets, across all groups. But there was no clear additive benefit from histology: G3 vs G2 had a HR of 1.10 (0.45 to 2.69). Read that with one more number. 79% of group 2 (99 of 125) were also in histological remission when they reached their endoscopic target, so the groups were clinically similar and the comparison rested on about one in five. And endoscopy was read centrally. When it is, Mayo 0 is usually free of neutrophils: 84% in LUCENT and over 90% in QUASAR. Local reading is another matter. In OCTAVE, central readers confirmed only 38 to 46% of local Mayo 0 calls. We do not have the equivalent data on paired centrally read endoscopy plus histology in UC patients not on advanced therapy (as far as I know) - that might matter in the real world. We know histological inflammation carries long-term risk, including cancer. VERDICT covers only one to two years. Most of that long-term evidence comes from scoring old H&E slides, recut from paraffin blocks if needed. You can't go back and re-read an old endoscopy the same way. So ... with endoscopy read well, histology added nothing we could detect. In the clinic, where we read our own scopes, it may still catch what endoscopy misses. FINALLY ... in the debate on histology vs endoscopy we must not lose sight of just how important VERDICT is for treat-to-target in UC. It is very clear that treating to endoscopic healing makes a huge difference to important patient outcomes. Of course, I still want to know if calprotectin is a good enough proxy for the real world ... ok let's await the full paper!