GARDENIAUC · Head-to-head
Etrolizumab versus infliximab for the treatment of moderately to severely active ulcerative colitis (GARDENIA): a randomised, double-blind, double-dummy, phase 3 study
Danese S, et al.
Lancet Gastroenterol Hepatol 2022 · PMID 34798038 (PubMed, opens in a new tab) · doi 10.1016/S2468-1253(21)00294-6 (full paper, opens in a new tab)
| Outcome | Etrolizumab | Infliximab | Estimate (confidence interval) | P |
|---|---|---|---|---|
| Response and remission, Primary endpoint (primary) | 18.6% (37/199) | 19.7% (39/198) | −0.9 (95% CI −8.7 to 6.8) | P=0.81 |
A superiority trial that misses is not proof of equivalence: the interval allows infliximab to be up to 8.7 points better.
Source: Danese 2022, abstract (modified intention to treat). Designed to show superiority; 48% (etrolizumab) and 52% (infliximab) completed week 54.
Structured abstract
BackgroundEtrolizumab, a gut-targeted anti-β7 integrin antibody, had improved remission over placebo in phase 2; GARDENIA tested whether it was superior to infliximab in anti-TNF-naive UC.
MethodsPhase 3, randomised (1:1), double-blind, double-dummy, parallel-group superiority trial (114 centres), 54 weeks. 397 patients. Primary endpoint: Clinical response at week 10 and clinical remission (Mayo score ≤2, no subscore >1) at week 54.
ResultsPrimary result. 18.6% (37/199) vs 19.7% (39/198); adjusted difference −0.9% (95% CI −8.7 to 6.8); P=0.81
Take-homeEtrolizumab, an anti-β7 integrin antibody, did not beat infliximab in anti-TNF-naive UC, and it never became a treatment option.
| Patients | 397 |
|---|---|
| Population | Adults aged 18–80 with moderately to severely active UC (Mayo score 6–12, endoscopic subscore ≥2, disease extending ≥20 cm from the anal verge) who were naive to TNF inhibitors |
| Comparison | Etrolizumab 105 mg SC every 4 weeks vs infliximab 5 mg/kg IV at weeks 0, 2 and 6 and then every 8 weeks |
| Drugs | etrolizumab, infliximab |
| Classes | Anti-integrin, Anti-TNF |
| Published | Danese S, et al. Lancet Gastroenterol Hepatol 2022;7:118–27. |
Key points
- Etrolizumab was not superior to infliximab; the authors judged the two clinically similar.
- Serious adverse events (including serious infections) were more frequent with etrolizumab: 16% vs 10%.
- Only 48% (etrolizumab) and 52% (infliximab) completed the study through week 54.
- The first phase 3 maintenance study in UC to use infliximab as the active comparator.
Limitations
- Designed for superiority, so the result cannot be read as proof of equivalence.
- High dropout before week 54.
- Anti-TNF-naive patients only.
Commentary
Context. Etrolizumab blocks the β7 integrin subunit and was expected to improve on vedolizumab’s gut-selective approach. Its phase 3 programme instead gave mixed results. In the HIBISCUS I and II induction trials, etrolizumab beat placebo for week-10 remission in HIBISCUS I (19.4% vs 6.9%; difference 12.3%, 95% CI 1.6–20.6; P=0.017) but not in HIBISCUS II (18.2% vs 11.1%; P=0.17), and the pooled analysis showed no superiority over adalimumab (Rubin DT, et al. Lancet Gastroenterol Hepatol 2022;7:17–27; PMID 34798036).
Why it is worth knowing. GARDENIA and HIBISCUS are among the few active-comparator trials in UC. Together with VARSITY, they show that a plausible mechanism is not enough: new agents need direct comparisons with established drugs. Etrolizumab is not approved in Canada, and the manufacturer stopped developing it for IBD.
Journal club
Design
Phase 3, randomised (1:1), double-blind, double-dummy, parallel-group superiority trial (114 centres), 54 weeks. 397 patients.
Primary result. 18.6% (37/199) vs 19.7% (39/198); adjusted difference −0.9% (95% CI −8.7 to 6.8); P=0.81
Validity and applicability
- Designed for superiority, so the result cannot be read as proof of equivalence.
- High dropout before week 54.
- Anti-TNF-naive patients only.
Where it sits now
Etrolizumab, an anti-β7 integrin antibody, did not beat infliximab in anti-TNF-naive UC, and it never became a treatment option.
- VARSITY (2019): In the first head-to-head biologic trial in IBD, vedolizumab beat adalimumab for clinical remission and endoscopic improvement at 52 weeks in UC.
Questions for discussion
- The CI runs from −8.7 to +6.8 points. Could you call the drugs equivalent? What would a non-inferiority design have needed?
- Only about half of patients completed week 54. How does heavy dropout affect a composite response-plus-remission endpoint?
- Serious adverse events were 16% vs 10%. How should a numerically worse safety signal weigh against similar efficacy for a new drug?