Astegolimab cut COPD flare-ups in one trial but not in its twin
Two large trials of the anti-ST2 antibody gave discordant results: a modest signal in the phase 2b study, and a missed primary endpoint for the same dose in the confirmatory phase 3.
| Group | Value (value) |
|---|---|
| ALIENTO — every 2 weeks | 0.85 (0.72 to 1) |
| ALIENTO — every 4 weeks | 0.93 (0.79 to 1.1) |
| ARNASA — every 2 weeks | 0.85 (0.72 to 1.01) |
| ARNASA — every 4 weeks | 0.82 (0.7 to 0.98) |
Astegolimab, an antibody that blocks the ST2 receptor for interleukin-33, reduced flare-ups of chronic obstructive pulmonary disease in one large trial but failed to hit its target in a second, larger one testing the same doses [s1]. The two trials, ALIENTO and ARNASA, were published together in The Lancet, and their split verdict is the story: a treatment that looked promising in the phase 2b study did not clearly confirm that promise when the confirmatory phase 3 study repeated the experiment [s1].
COPD is a progressive lung disease in which recurrent exacerbations — episodes of worsening breathlessness, cough and sputum — drive hospital admissions and decline. Most patients are managed with inhaled bronchodilators and steroids, but a subset keep exacerbating despite optimised inhalers, and it is that group biologic drugs are trying to reach. Astegolimab targets the interleukin-33/ST2 pathway, which is implicated in both the neutrophilic and eosinophilic inflammation seen during COPD exacerbations [s1]. Crucially, the trials enrolled patients irrespective of their baseline blood eosinophil count, unlike some earlier biologics aimed only at eosinophil-high disease [s1].
What the two trials did
Both were randomised, double-blind and placebo-controlled, and both gave subcutaneous astegolimab 476 mg either every 2 weeks or every 4 weeks, on top of optimised inhaled maintenance therapy, over 52 weeks [s1]. Participants were current or former smokers with COPD and a history of frequent exacerbations. The primary endpoint in each was the annualised rate of moderate or severe exacerbations [s1].
The phase 2b trial, ALIENTO, randomised 1301 participants (433 to the every-2-week dose, 437 to the every-4-week dose and 431 to placebo) between Oct 5, 2021, and Feb 19, 2024 [s1]. The phase 3 trial, ARNASA, randomised 1375 participants (459, 459 and 457 respectively) between Jan 9, 2023, and June 25, 2024 [s1].
The discordant result
In ALIENTO, the every-2-week dose lowered the exacerbation rate versus placebo with an adjusted rate ratio of 0·85 (95% CI 0·72–1·00; p=0·049) — a result that scraped past the conventional significance threshold [s1]. The every-4-week dose in ALIENTO did not separate from placebo (rate ratio 0·93, 0·79–1·10; p=0·38) [s1].
In ARNASA, the pattern inverted. The every-2-week dose — the one that had worked in ALIENTO — returned a near-identical point estimate but missed significance (rate ratio 0·85, 0·72–1·01; p=0·068), while the every-4-week dose, which had done nothing in ALIENTO, reached significance (rate ratio 0·82, 0·70–0·98; p=0·024) [s1]. A drug that reduces exacerbations should not depend on which trial and which dosing interval you look at, and the inconsistency is why the phase 3 study cannot be read as a clean confirmation.
The reason this ordering matters is that phase 3 exists precisely to confirm phase 2b. A phase 2b study like ALIENTO is designed to find a signal worth pursuing; the larger phase 3 study, ARNASA, is the one meant to nail it down before regulators act. When the confirmatory trial does not reproduce the phase 2b result at the same dose, the usual scientific reading is caution, because early-phase findings that fail to replicate are common and the replication is the more reliable of the two [s1]. That a different dose reached significance in ARNASA does not rescue the every-2-week story; it more plausibly reflects the play of chance across multiple comparisons than a real dose reversal [s1].
The authors' own interpretation is measured: ALIENTO showed the every-2-week dose lowering the annual exacerbation rate, and in ARNASA "these findings did not meet statistical significance," though together they "suggest a role" for targeting the ST2/IL-33 pathway [s1]. An accompanying Lancet commentary frames astegolimab within an emerging wave of COPD biologics while acknowledging the field's need for options in patients who keep exacerbating despite inhalers [s2].
Safety
Adverse events were common but balanced across groups: 1093 (84·0%) of 1301 participants in ALIENTO and 1176 (85·5%) of 1375 in ARNASA reported at least one [s1]. Deaths occurred in 40 (3·1%) of 1301 participants in ALIENTO and 44 (3·2%) of 1375 in ARNASA, distributed evenly across treatment groups; across both trials, three deaths (0·1%) were judged by investigators to be treatment-related [s1]. The trials were funded by Genentech, a member of the Roche Group, and F. Hoffmann-La Roche [s1]; any characterisation of the drug's benefit that leans only on the significant arms is a sponsor's reading, not the trials' combined one.
What it means
For now, astegolimab is a candidate with a signal, not a settled treatment. The honest summary is that one dose worked in one trial, a different dose worked in the other, and the dose that looked best in phase 2b did not confirm in phase 3 — the exact situation in which enthusiasm outruns the data. Confidence intervals that touch or cross 1·00 in three of the four comparisons mean the true effect could be small or, in some arms, nil [s1]. Whether regulators or guideline writers act on this will hinge on how they weigh a consistent-but-modest point estimate against inconsistent statistical significance, and possibly on subgroup analyses not covered here. Readers should treat claims that the ST2 pathway is now a proven COPD target as premature.
Sources
- [s1] Safety and efficacy of astegolimab for COPD with frequent exacerbations regardless of baseline blood eosinophil counts (ALIENTO and ARNASA). The Lancet, 18 May 2026. https://doi.org/10.1016/S0140-6736(26)00637-9
- [s2] Astegolimab and the new era of biologics in COPD care. The Lancet, 18 May 2026. https://doi.org/10.1016/S0140-6736(26)00924-4
Sources
- Safety and efficacy of astegolimab for COPD with frequent exacerbations regardless of baseline blood eosinophil counts (ALIENTO and ARNASA): randomised, double-blind, placebo-controlled, phase 2b and 3 trials — The Lancet , May 18, 2026
- Astegolimab and the new era of biologics in COPD care — The Lancet , May 18, 2026
More on
An antibody against a single inflammatory signal cut COPD flare-ups in two trials
Tozorakimab, which blocks interleukin-33, lowered the rate of moderate or severe COPD exacerbations by about a third against placebo, without an eosinophil requirement to qualify.
Triple inhaler therapy cuts COPD flare-ups more than two-drug inhalers
Two large trials, ETHOS and IMPACT, found adding an inhaled steroid to a two-bronchodilator inhaler lowered exacerbation rates in COPD — at the cost of more pneumonia.
A beta-blocker did not improve outcomes in COPD patients without heart disease
A 1,695-patient trial tested whether metoprolol helps people with COPD and no cardiovascular indication. It missed its primary endpoint — but, unlike an earlier trial, showed no signal of harm.
A biologic helped polymyalgia patients taper steroids in a phase 3 trial
REPLENISH randomised 381 patients with relapsed polymyalgia rheumatica. Sustained remission at 52 weeks reached about 41% on secukinumab versus 20.4% on placebo. Novartis funded the trial.