Gilead Posts Phase 2 Emvistegrast Win In Ulcerative Colitis, Setting Up A 2027 Phase 3 Monotherapy Study

Gilead Sciences has delivered a mid-stage ulcerative colitis readout that supports taking emvistegrast into late-stage testing, adding a new oral entrant to one of biopharma’s most competitive inflammatory disease markets. In the ongoing phase 2 Swift trial, 66% of patients receiving the 200-mg dose achieved a clinical response at week 12, compared with 37% of placebo recipients, and clinical remission was achieved by 31% of treated patients versus 8% on placebo.

The result matters because emvistegrast is not just another UC asset chasing an established mechanism. Gilead is developing the drug as a once-daily oral inhibitor of α4β7 integrin, a pathway already validated by Takeda’s Entyvio but currently served by infused or injected therapy. If larger trials hold up the efficacy and tolerability profile, the commercial argument becomes straightforward: Gilead would have a pill against a target physicians already know.

The Data

Swift enrolled 238 adults with moderately to severely active ulcerative colitis and evaluated 25 mg, 75 mg and 200 mg doses of emvistegrast in a randomized, placebo-controlled design. The findings were presented at United European Gastroenterology Week 2026 in Barcelona, Spain.

Beyond the primary endpoint, the 200-mg dose also improved multiple secondary measures. At week 12, 25% of patients achieved histological endoscopic mucosal improvement, compared with 10% of placebo recipients. Gilead also reported endoscopic improvement in 39% of patients on the 200-mg dose versus 12% on placebo.

The safety profile in the release was relatively even against placebo. Across the three emvistegrast dose groups, mostly grade 2 or lower adverse events were reported in 36% of patients, compared with 37% of placebo recipients. Serious adverse events occurred in 2% of emvistegrast recipients, with none considered related to treatment, and 1% of all trial participants discontinued because of adverse events.

There are still limits on how far the readout can be stretched. Gilead said 78% of participants were either new to advanced therapies or had not previously experienced treatment failure with one, while 22% had failed one or two advanced therapies. That leaves open the question of how durable the effect will be and how the drug will perform in tougher refractory populations.

The Commercial Picture

Analysts at Evercore ISI called the dataset “the strongest datasets ever for a monotherapy in UC,” and said Gilead is “sitting on a foundational asset in UC.” That is a strong endorsement, but the commercial burden now shifts to replication in larger studies and proving the asset can compete in a market already filled with drugs aimed at different inflammatory pathways.

The approved competition cited in the source spans biologics and oral drugs. AbbVie’s Skyrizi and Johnson & Johnson’s Tremfya target IL-23, while AbbVie’s Rinvoq is an oral JAK inhibitor. The pipeline is also active: Spyre Therapeutics recently reported a phase 2 win for SPY003, and Merck & Co. reported positive phase 3 results in June for tulisokibart, its anti-TL1A antibody acquired through the 2023 purchase of Prometheus Biosciences.

Gilead’s strategic angle is narrower and potentially cleaner than trying to outflank every mechanism at once. Swati Tole, senior vice president of clinical development for inflammation at Gilead, said the company plans to initiate a Phase 3 monotherapy study in 2027 and sees emvistegrast as a potential backbone for future combination approaches. That framing suggests Gilead is treating the current readout as both a standalone opportunity and a platform around which it could build broader inflammatory bowel disease combinations if the maintenance data support durability.

For now, the main signal is that oral delivery against a proven gut-homing pathway has moved from theoretical differentiation to a data-backed late-stage plan. In UC, that is enough to make emvistegrast more than a mechanism echo of Entyvio; it becomes a test of whether convenience can be added to pathway validation without giving up efficacy or tolerability.