ARO-MUC5AC was safe, quiet, and dropped: an inhaled RNAi program with no target engagement readout

What was tried

Arrowhead Pharmaceuticals ran AROMUC5AC-1001 (NCT05292950), a Phase 1/2a randomized, triple blind, sequential study of ARO-MUC5AC inhalation solution in healthy volunteers and in patients with moderate to severe asthma or COPD. Enrollment was 78. It ran from 27 June 2022 to 12 November 2024, carries status TERMINATED with the posted reason "Sponsor business decision", and has posted results.

Fourteen groups were run. Single ascending dose cohorts in healthy volunteers covered 24, 56, 108 and 232 mg against pooled placebo. Multiple ascending dose cohorts gave three doses on Days 1, 15 and 29 to healthy volunteers at 24, 56 and 108 mg, asthma patients at 56 and 108 mg, and COPD patients at 56 mg. Patient cohorts were 7 and 6 for asthma and 6 for COPD, against 4 and 2 placebo.

Entry required a source verifiable diagnosis, stable controller therapy, and an induced sputum sample at screening. ARO-MUC5AC is an inhaled RNA interference therapeutic on Arrowhead's pulmonary TRiM platform, designed to reduce production of the mucin MUC5AC. It has no ChEMBL record. The target gene is MUC5AC (ENSG00000215182).

The biological hypothesis

Airway mucus is dominated by two gel forming mucins with different jobs. MUC5B is required for mucociliary clearance and airway defence, shown in mice where Muc5b deficiency caused impaired clearance, bacterial accumulation and inflammation. MUC5AC is the inducible one, driven by type 2 inflammation and implicated in occluding plugs.

The clinical case is strong. Mucus plugs on CT occur in 58 percent of asthma patients against 4.5 percent of controls, persist in the same segments for years, and a high plug score appears in 67 percent of patients with FEV1 below 60 percent predicted against 6 percent of those above 80 percent. In COPD, sputum MUC5AC concentration tracks with initiation, progression and exacerbation risk, rising more steeply with severity than MUC5B. Arrowhead's 2022 rationale described MUC5AC as highly upregulated in the asthmatic airway and implicated in mucus plugging and exacerbations.

The genetic support splits by indication in a way the design did not respect. Open Targets scores MUC5AC against asthma at 0.3815, with a GWAS credible sets contribution of 0.5814 and a literature score of 0.9231. Against COPD it scores 0.1115, from literature alone, with no genetic association evidence. The program dosed both.

What actually happened

The safety result was clean. Across all 78 participants the posted results record zero serious adverse events and zero deaths. Treatment emergent events were expected inhalation and background terms: cough in 7 participants including 1 of 8 on placebo, upper respiratory tract infection in 6, headache in 9, oropharyngeal pain in 4. Three asthma exacerbation terms appear, one in the placebo arm.

The efficacy picture is uninformative and was always going to be. Change from baseline in FEV1 at Day 85 was 0.048 litres on asthma placebo, minus 0.082 litres at 56 mg and plus 0.107 litres at 108 mg, with standard deviations of 0.398 and 0.355 litres against cohorts of six. COPD showed 0.060 litres on placebo, from two participants, against 0.088 litres at 56 mg. Dispersion exceeds the point estimates several times over.

The decisive observation is what is absent. Nineteen outcome measures were posted: treatment emergent adverse events, FEV1, FVC, and fifteen pharmacokinetic parameters. No sputum MUC5AC measurement, no mucus plug imaging, no target engagement endpoint of any kind appears, despite an entry criterion requiring induced sputum. Arrowhead's current pipeline lists ARO-RAGE at Phase 2 and ARO-MMP7 at Phase 1, and ARO-MUC5AC does not appear.

Failure mechanism, best guess

No safety signal stopped this and no efficacy hypothesis was tested. The posted reason is a business decision, and nothing contradicts it.

The mechanism worth naming is a design one. A first in human study of an inhaled oligonucleotide against a secreted airway mucin has one question that comes before all others: does inhaled dosing reduce MUC5AC in the airway. The posted outcome set cannot answer it. Spirometry in six patients over 85 days is no proxy for mucin knockdown, and plasma pharmacokinetics measures systemic spillover rather than lung target engagement. When the readout ended, the program held a safe molecule with an unknown pharmacodynamic effect and no basis for prioritizing it over assets that had one. That is a reprioritization made easy by absent evidence rather than by its content.

A second contributor is indication scope. The COPD cohorts consumed capacity in a disease where the target has no genetic support, at n of 6 and 2, enough for tolerability and nothing else.

How to prevent this next time

Endpoint level data exist here, but with no pharmacodynamic endpoint to model and no powered comparison a Bayesian update is not supportable. The levers are design levers.

Make target engagement the gating endpoint of a first in human study for any novel mechanism, and power for it rather than for tolerability alone. Sputum was already collected at screening, and sputum MUC5AC is an established measure in the COPD cohort literature.

Use the validated disease biomarker that already exists. Mucus plug scoring on CT has a published method, reproducible persistence, and demonstrated movement under effective therapy: tezepelumab reduced mucus plug score by 1.7 points against 0.0 on placebo. A mucus directed agent that never measures mucus plugs gives up its strongest readout.

Enrich on the mechanism. A high mucus score and eosinophilic asthma identify patients in whom MUC5AC driven obstruction is present, and both are measurable before randomization.

Match indication to evidence class. Asthma carries GWAS support here and COPD does not, so the COPD cohorts should have followed a positive asthma pharmacodynamic result rather than run alongside it.

The single highest leverage change would have been to make sputum MUC5AC knockdown the primary pharmacodynamic outcome of the multiple ascending dose phase, so the program ended with a knockdown number rather than only a tolerability profile.

What this means for similar programs

Open Targets lists three clinical stage programs against MUC5AC, all oncology antibodies treating the mucin as a tumour antigen: ensituximab at Phase 2, and clivatuzumab tetraxetan and its yttrium-90 conjugate at Phase 3. No respiratory program appears. ARO-MUC5AC was the only clinical attempt to reduce MUC5AC production in the airway, and it ended without a public answer on whether that is achievable.

The Claidex Mechanistic Risk Score for MUC5AC is 28 of 100, band yellow. Genetic deficit contributes 9.28 of 15 and saturation 8.26 of 15, with phase burden only 5.13 of 30 because the single failure sits at Phase 1/2. That low phase burden is the honest reading: this target has barely been tested. The biology thesis is intact and the gap is specific. What is missing is one knockdown number in human airway.

Open questions

Did ARO-MUC5AC reduce airway MUC5AC at any dose? No knockdown figure appears in the posted record.

Was the limiting factor delivery rather than target choice? An inhaled RNAi agent must reach secretory cells beneath a mucus layer the target itself thickens, a self limiting delivery problem that a pharmacodynamic endpoint would have separated from a target problem.

Does reducing MUC5AC while sparing MUC5B preserve mucociliary defence in humans? The mouse work establishes MUC5B is required, and no human data address the selectivity margin.

Sources

1. • ClinicalTrials.gov, NCT05292950, AROMUC5AC-1001 record with posted results, accessed 17 September 2026. https://clinicaltrials.gov/study/NCT05292950 - Arrowhead Pharmaceuticals, "Arrowhead Files for Regulatory Clearance to Initiate Phase 1/2a Study of ARO-MUC5AC for Treatment of Muco-Obstructive Lung Diseases", 24 March 2022. https://ir.arrowheadpharma.com/news-releases/news-release-details/arrowhead-files-regulatory-clearance-initiate-phase-12a-study-0 - Arrowhead Pharmaceuticals pipeline, accessed 17 September 2026. https://arrowheadpharma.com/en-us/pipeline - Roy MG, et al. Muc5b is required for airway defence. Nature, 2014. https://- Dunican EM, et al. Mucus plugs in patients with asthma linked to eosinophilia and airflow obstruction. J Clin Invest, 2018. https://- Radicioni G, et al. Airway mucin MUC5AC and MUC5B concentrations and the initiation and progression of chronic obstructive pulmonary disease. Lancet Respir Med, 2021. https://- Nordenmark LH, et al. Tezepelumab and Mucus Plugs in Patients with Moderate-to-Severe Asthma. NEJM Evidence, 2023. https://- Liegeois MA, et al. Cellular and molecular features of asthma mucus plugs provide clues about their formation and persistence. J Clin Invest, 2025. https://- Open Targets Platform, MUC5AC (ENSG00000215182) association and clinical candidate data, accessed 17 September 2026. https://platform.opentargets.org/target/ENSG00000215182 - openFDA drug adverse event API, queried 17 September 2026. https://api.fda.gov/drug/event.json.