In giant cell arteritis, stopping upadacitinib at a year let disease roar back

Patients with giant cell arteritis whose disease was in remission on the oral drug upadacitinib relapsed far more often when they were switched to placebo than when they stayed on the drug through a second year — flaring in 59.5% of cases versus 7.4% — according to the extension phase of the SELECT-GCA trial [s1]. The finding answers a practical question the original trial left open: once this drug controls the disease, can it be stopped? On this evidence, mostly not — continuing treatment kept far more patients in steroid-free remission [s1].

Giant cell arteritis is an inflammation of medium and large arteries, most often in people over 50, that can cause headache, jaw pain, and — if it strikes the arteries to the eye — sudden, permanent blindness. The mainstay of treatment has long been high-dose glucocorticoids (steroids), which work but carry heavy cumulative harms: bone loss, diabetes, infection. The search in this field is for drugs that control the disease while sparing steroids. The interleukin-6 blocker tocilizumab was the first proven steroid-sparing option; a rival, secukinumab, recently fell short in its own phase 3 trial. Upadacitinib takes yet another route — it is a Janus kinase (JAK) inhibitor, a pill that blocks the internal signalling of several inflammatory cytokines, including interleukin-6.

What the earlier result established

SELECT-GCA is a phase 3 trial in patients aged 50 or older with new-onset or relapsing giant cell arteritis, who were randomly assigned in a 2:1:1 ratio to upadacitinib 15 mg or 7.5 mg once daily with a 26-week steroid taper, or placebo with a 52-week taper [s2]. Its primary result, reported in 2025, was that the 15 mg dose beat placebo on sustained remission at week 52 — achieved by 46.4% of patients versus 29.0% — while the 7.5 mg dose, at 41.1%, was not significantly better than placebo [s2]. The 209 patients on 15 mg, 107 on 7.5 mg and 112 on placebo also showed no excess of major cardiovascular events over the year, a specific concern for this drug class [s2].

That established the drug works for a year. It did not say what happens next — whether a patient in remission needs to keep taking it.

What the extension asked

In a blinded second period, patients who had reached at least 24 weeks of continuous remission by week 52 were re-randomised 2:1 either to continue upadacitinib 15 mg or to switch to placebo, and were followed from week 52 to week 104 [s1]. Of the 103 patients on the 15 mg dose who qualified, 68 continued and 35 switched [s1].

The gap that opened between them was wide. Over that second year, patients who continued upadacitinib had far fewer disease flares than those switched to placebo (7.4% versus 59.5%), were far more likely to maintain glucocorticoid-free remission (71.7% versus 31.4%), and took less steroid overall — a median cumulative dose of 0 mg versus 1048 mg [s1]. Safety over the two years was generally similar between the groups, with no new safety concerns identified [s1].

How to read it

The message is coherent and clinically useful: upadacitinib maintains remission, and stopping it at a year lets the disease return in most patients [s1]. The near-total relapse in the withdrawal arm also underlines that a year of control is not a cure. But the caveats are real. This was a small subgroup — barely a hundred patients, split unevenly — carried into the extension only if they had already done well, so it speaks to maintenance in responders, not to everyone who starts the drug [s1]. The authors note that all the P values in this analysis are nominal, meaning the comparisons were not adjusted for testing several outcomes at once, so the individual figures should be read as descriptive rather than as fresh confirmatory proof [s1]. And because the trial withdrew to placebo rather than testing a slow taper, it cannot say whether some patients could come off the drug gradually [s1]. The trial was funded by upadacitinib's manufacturer [s2].

Where the drug fits also depends on cost and on comparisons it did not run: SELECT-GCA did not test upadacitinib against tocilizumab, the established steroid-sparing standard, so which patients should get which drug remains unsettled [s1][s2].

What to watch

The open questions are durability beyond two years, whether any patients can safely stop, and head-to-head data against tocilizumab [s1][s2]. This article describes research and is not medical advice; treatment of giant cell arteritis is a decision for treating clinicians.

Sources

• Clinical outcomes of upadacitinib continuation vs withdrawal in giant cell arteritis through 2 years — Annals of the Rheumatic Diseases, 28 July 2026
• A Phase 3 Trial of Upadacitinib for Giant-Cell Arteritis — New England Journal of Medicine, 2 April 2025