TUSCANY-2UC · Induction and maintenance
Anti-TL1A antibody, afimkibart, in moderately-to-severely active ulcerative colitis (TUSCANY-2): a multicentre, double-blind, treat-through, multi-dose, randomised, placebo-controlled, phase 2b trial
Danese S, et al.
Lancet Gastroenterol Hepatol 2025 · PMID 40706613 (PubMed, opens in a new tab) · doi 10.1016/S2468-1253(25)00129-3 (full paper, opens in a new tab)
| Outcome | Afimkibart | Placebo | Estimate (confidence interval) | P |
|---|---|---|---|---|
| Total Mayo score · primary endpoint | ||||
| 50 mg (primary) | 12/47 (26%) | 5/43 (12%) | +13.9 (90% CI −0.2 to 27.7) | P=0.0545 |
| 150 mg (primary) | 14/60 (23%) | 5/43 (12%) | +11.7 (90% CI −1.7 to 24.1) | P=0.0823 |
| 450 mg (primary) | 21/88 (24%) | 5/43 (12%) | +12.2 (90% CI −0.6 to 22.9) | P=0.0642 |
| Modified Mayo score · secondary | ||||
| 50 mg | 14/47 (30%) | 5/43 (12%) | +18.2 (90% CI 3.3–32.2) | |
| 150 mg | 21/60 (35%) | 5/43 (12%) | +23.4 (90% CI 6.2–36.3) | |
| 450 mg | 28/88 (32%) | 5/43 (12%) | +20.2 (90% CI 3.2–31.3) | |
Changing the endpoint definition moved every interval clear of zero, and the doses were not formally compared with each other.
Source: Danese 2025, abstract. All comparisons share one placebo group (5/43). The modified Mayo analysis was reported in line with updated FDA guidance.
Structured abstract
BackgroundTL1A, a TNF-family cytokine, is a new therapeutic target in IBD. TUSCANY-2 was a dose-ranging phase 2b trial of the anti-TL1A antibody afimkibart in moderately to severely active UC.
MethodsPhase 2b, multicentre, double-blind, treat-through, multi-dose, randomised, placebo-controlled trial (114 centres, 23 countries). 246 patients. Primary endpoint: Clinical remission by total Mayo score (≤2, no subscore >1) at week 14.
ResultsPrimary result. Not met: 26% (50 mg), 23% (150 mg) and 24% (450 mg) vs 12% with placebo; risk differences 13.9% (90% CI −0.2 to 27.7; P=0.0545), 11.7% (90% CI −1.7 to 24.1; P=0.0823) and 12.2% (90% CI −0.6 to 22.9; P=0.0642)
Take-homeAfimkibart missed its phase 2b primary endpoint (total Mayo remission), although remission by the modified Mayo score favoured it; anti-TL1A therapy still needs phase 3 proof.
| Patients | 246 |
|---|---|
| Population | Adults aged 18–75 with moderately to severely active UC (total Mayo score 6–12, endoscopic subscore ≥2) |
| Comparison | Afimkibart SC 50 mg, 150 mg or 450 mg every 4 weeks vs placebo for 12 weeks; then afimkibart treat-through maintenance to week 56 |
| Drugs | afimkibart |
| Classes | Anti-TL1A |
| Published | Danese S, et al. Lancet Gastroenterol Hepatol 2025;10:882–95. |
Key points
- By the modified Mayo score (reported in line with updated FDA guidance), week-14 remission was 30%, 35% and 32% vs 12%; risk differences 18.2%, 23.4% and 20.2%, with 90% CIs above zero.
- Treatment-emergent adverse events during induction were similar with afimkibart and placebo.
- Induction: six serious adverse events in the active groups and four with placebo. Maintenance: 12 of 224 patients (5%) had 13 serious adverse events. No deaths.
Limitations
- The primary endpoint was not met; the positive modified-Mayo results are secondary.
- Small placebo group (43 patients) and 90% confidence intervals.
- No dose response across 50, 150 and 450 mg.
Commentary
Teaching point. TUSCANY-2 shows how the choice of endpoint definition matters: the same patients gave a non-significant result by total Mayo score and a favourable one by modified Mayo score. Read the endpoint definition before comparing remission rates across trials.
Related. Other anti-TL1A trials in UC: tulisokibart (ARTEMIS-UC, PMID 39321363) and duvakitug (RELIEVE UCCD, PMID 42462750).
Journal club
Design
Phase 2b, multicentre, double-blind, treat-through, multi-dose, randomised, placebo-controlled trial (114 centres, 23 countries). 246 patients.
Primary result. Not met: 26% (50 mg), 23% (150 mg) and 24% (450 mg) vs 12% with placebo; risk differences 13.9% (90% CI −0.2 to 27.7; P=0.0545), 11.7% (90% CI −1.7 to 24.1; P=0.0823) and 12.2% (90% CI −0.6 to 22.9; P=0.0642)
Validity and applicability
- The primary endpoint was not met; the positive modified-Mayo results are secondary.
- Small placebo group (43 patients) and 90% confidence intervals.
- No dose response across 50, 150 and 450 mg.
Where it sits now
Afimkibart missed its phase 2b primary endpoint (total Mayo remission), although remission by the modified Mayo score favoured it; anti-TL1A therapy still needs phase 3 proof.
- ASTRO (2026): Guselkumab can be started with subcutaneous 400 mg injections instead of IV infusions, with clear benefit over placebo by week 12.
- RELIEVE UCCD (UC) (2026): Duvakitug, an anti-TL1A antibody, showed a dose-related remission signal in UC in phase 2b; phase 3 trials will decide its place.
- QUASAR (2025): Guselkumab induces and maintains remission in UC, and the follow-up ASTRO trial showed induction can be given subcutaneously.
- ARTEMIS-UC (2024): Anti-TL1A therapy is a new mechanism in UC; tulisokibart produced a large placebo-adjusted remission rate in a short phase 2 trial, with phase 3 results still awaited.
Questions for discussion
- The primary endpoint missed by the total Mayo score, but the modified Mayo score favoured afimkibart. When is it legitimate to lean on a secondary endpoint?
- The trial used 90% confidence intervals and a 43-patient placebo group. What can a phase 2b result like this tell you?
- There was no clear dose response across a nine-fold dose range. What could explain that, and what would you want phase 3 to show?