Open questions — asthma¶
Last curated: 2026-10-29 (OQ-6, -16 and -21 sharpened; OQ-17 registry status rechecked). Stable identifiers (OQ-n) preserve question history. New citations were returned live by PubMed E-utilities in this sweep.
How to read priority¶
Tier 1 questions could plausibly change practice, prevention, or the allocation of major resources and have a testable design now. Tier 2 questions remain important but depend on measurement tools, cohorts, implementation infrastructure, or a Tier-1 result. Priority is curator judgment, not a formal consensus exercise.
Dots not yet connected¶
The table joins evidence streams that already exist in this knowledge base but are rarely studied together. “Missing junction” means the build did not retrieve a definitive connecting study; it is not a proof that none exists.
| # | Dot A | Dot B | Missing junction | Powers |
|---|---|---|---|---|
| D1 | About one-third of adults with a physician diagnosis could not have current asthma confirmed in structured reassessment (PMID 28114551) | Primary-care barriers to objective testing are well characterized (PMID 34740591) | A scalable diagnostic-reverification pathway with long-term clinical and economic outcomes | OQ-1 |
| D2 | Type-2 biomarkers are prognostic and sometimes predictive (PMID 40215991) | Molecular phenotypes vary over time (PMID 33008937) | Repeated-measure decision rules that outperform one pre-treatment blood draw | OQ-2, OQ-4 |
| D3 | Biologics can produce clinical remission in some severe asthma (PMID 39549709) | Remission definitions remain consensus constructs (PMID 31866436) | A validated, patient-valued remission endpoint linked to treatment withdrawal and durability | OQ-3 |
| D4 | Tezepelumab reduces exacerbations across a broad severe-asthma population (PMID 33979488) | Non-type-2 asthma remains mechanistically heterogeneous (PMID 33160187) | Prospective markers that identify genuine T2-low responders rather than average subgroup effects | OQ-4 |
| D5 | Asthma genetics now benefits from multi-ancestry discovery (PMID 36778051) | Pollution, tobacco, wildfire and occupational exposures are modifiable (PMID 38311978, 38783343) | Intervention studies stratified by genetic susceptibility without converting risk into biological determinism | OQ-5, OQ-6 |
| D6 | Infant RSV infection is prospectively linked with later asthma (PMID 37086744) | Preschool anti-inflammatory therapy controls symptoms without clearly changing natural history (PMID 16687711) | A prevention trial that targets a mechanistically enriched early-life trajectory and follows durable asthma incidence | OQ-5 |
| D7 | Attack risk rises when inflammatory and clinical risk factors co-occur (PMID 40215991) | Digital inhaler monitoring improves measurement more consistently than outcomes (PMID 36986513) | A prospective system that turns prediction into timely, equitable, beneficial action | OQ-7 |
| D8 | Oral-steroid bursts carry acute harms at population level (PMID 28404617) | Patients experience steroid trade-offs as a major quality-of-life burden (PMID 37643678) | A steroid-toxicity endpoint used routinely in trials, commissioning and shared decisions | OQ-8 |
| D9 | Shared decision-making has a growing qualitative evidence base (PMID 38613765) | Adherence is shaped by beliefs, routines and treatment burden (PMID 30053965, 32210541) | A scalable consultation intervention that improves hard outcomes without increasing inequity | OQ-9 |
| D10 | Area-level social determinants predict paediatric utilization (PMID 37455665) | Controller access and specialist pathways remain geographically unequal (PMID 36220057) | Trials that treat housing, pharmacy access, transport and school capacity as intervention components | OQ-10 |
| D11 | Parents make high-stakes attack decisions under uncertainty (PMID 35967097) | School nurses and caregivers identify different post-discharge barriers (PMID 30287588) | A child-centred emergency pathway tested across home, school and healthcare settings | OQ-11 |
| D12 | Patients define severe-asthma goals in functional terms (PMID 33577946) | Trials and payer continuation rules privilege standardized scores and exacerbation counts | A patient-weighted benefit framework that can be audited without becoming arbitrary | OQ-12 |
| D13 | Obesity is a treatable trait with multiple plausible mechanisms (PMID 29627041) | Weight-loss trials suggest benefit but remain small and heterogeneous (PMID 30605347) | A mechanism-stratified comparison of pharmacologic, surgical and behavioural weight loss in asthma | OQ-13 |
| D14 | Climate change increases wildfire smoke, heat and pollen hazards (PMID 37739070, 36917187) | Action plans usually focus on medicines and symptoms | Pragmatic trials of exposure forecasts, filtration, workplace protection and medication adaptation | OQ-14 |
When a junction study appears, connect it here and update the question it powers.
Tier 1 — highest-value directions¶
OQ-1. How should an asthma diagnosis be verified—and periodically re-verified—at scale?¶
Objective confirmation is central because variable symptoms alone are non-specific, yet normal spirometry on one day does not exclude asthma. Structured adult reassessment found a substantial fraction of prior diagnoses could not be confirmed, while implementation research identifies time, equipment, training and interpretation barriers in primary care (PMID 28114551, 29756989, 34740591). The missing study is a pragmatic pathway comparing staged spirometry, peak-flow variability, FeNO and challenge testing against usual care, with misdiagnosis, missed disease, exacerbations, medication withdrawal, cost and equity followed for years. Sharpened 2026-10-15: a meta-analysis restricted to normal-spirometry presentations proposed adult FeNO rule-in (≥65 ppb) and rule-out (<20 ppb) thresholds and showed blood eosinophils add nothing, validated in one single-centre primary-care-referral cohort; the paediatric validation is ongoing (PMID 42743935). The pathway trial can now prespecify FeNO as the triage step, and the open question is whether that triage improves long-term outcomes and equity rather than test accuracy. → diagnosis and objective testing
OQ-2. Can longitudinal biomarkers guide treatment better than threshold snapshots?¶
Blood eosinophils and FeNO provide probabilistic information, and combinations improve attack-risk stratification, but values vary with exposure, infection, corticosteroid use and adherence. Molecular phenotypes can also shift over time (PMID 40215991, 33008937). A useful test would compare repeated biomarker trajectories and treatment-response updates with current single-threshold algorithms, prespecifying calibration, subgroup performance and net benefit. → biomarkers
OQ-3. Can asthma remission become a durable, patient-valued treatment target?¶
Consensus frameworks distinguish clinical remission from complete biological remission, and real-world biologic cohorts show that some patients meet clinical definitions (PMID 31866436, 39549709). Unknowns include durability, airway-remodelling consequences, acceptable residual treatment, and whether stopping a biologic preserves benefit. Trials need a common core that includes attacks, oral-steroid exposure, lung function, symptoms, participation and treatment burden, followed beyond the usual one-year window. Sharpened 2026-10-15: a modified Delphi now proposes paired remission and minimal clinical disease activity definitions for severe asthma, explicitly pending validation (PMID 42463041). The pragmatic PERFORM RCT reported remission as a secondary outcome with inhaled triple versus ICS/LABA therapy (29.7% vs 17.0% at week 52; PMID 42474701). Remission is therefore no longer biologic-specific, and the gap has narrowed from "no definition" to "no validated, durability-tested definition". → phenotypes, endotypes and treatable traits, severe asthma and biologics
OQ-4. What works for genuinely non-type-2 asthma?¶
The field's most reproducible drug advances cluster around type-2 biology. Non-type-2 disease includes neutrophilic, paucigranulocytic, obesity-related, exposure-driven and infection-associated states rather than one coherent endotype (PMID 25534623, 33160187, 35487388). Broad upstream efficacy does not remove the need to identify who benefits and through which mechanism. Enriched adaptive trials should require repeated evidence of T2-low status and measure infection, airway structure and non-inflammatory traits. → airway immunobiology, biomarkers
OQ-5. Which early-life intervention prevents persistent asthma rather than postponing symptoms?¶
Viral lower-respiratory illness, microbiome development, allergy, tobacco exposure and inherited susceptibility form plausible early trajectories, but preventive nutrition and controller strategies have not yielded a universal intervention (PMID 21535176, 26947981, 35215404). The INSPIRE cohort strengthens the RSV–asthma temporal link, yet causality and preventability still require intervention evidence (PMID 37086744). A decisive programme would enrich infants by mechanistic risk, intervene before recurrent wheeze is established, and use objective school-age asthma rather than transient preschool symptoms as the endpoint. Updated 2026-10-22: DIVA did not establish a reduction in steroid-treated attacks with vitamin D3 in high-morbidity preschool asthma; recruitment shortfall left a wide confidence interval (Ducharme 2026, PMID 42871976; full text, Table 2). Its established-asthma population does not answer primary prevention of asthma onset. → genetics, environment and prevention, asthma in children
OQ-6. Which environmental interventions prevent attacks and new disease at realistic exposure reductions?¶
Traffic pollution, tobacco smoke, occupational agents, wildfire smoke and extreme weather are consistently associated with asthma outcomes (PMID 27881237, 38783343, 36821481, 37739070). Association does not specify the health gain from a particular filter, clean-air shelter, workplace substitution, housing repair or emissions policy. Cluster-randomized and natural-experiment designs should measure actual exposure change, asthma incidence or severe attacks, distributional effects and displacement of risk. Sharpened 2026-10-29: in COPSAC 2000/2010 (n=846), an IQR increment of 2.2 μg/m³ postnatal PM2.5 was associated with 2.70% lower FEV1 (95% CI 4.27% to 1.10% lower) and 7.30% higher specific airway resistance at age 6, with slower growth thereafter (Pedersen 2026, PMID 42933820; full text via PMC13711545). The authors could not exclude residual confounding by co-varying greenness, microbial diversity and farm proximity. This is a lung-function trajectory association in a relatively low-pollution European setting; it does not estimate the health gain from a specified exposure-reduction intervention. → genetics, environment and prevention
OQ-7. Can attack prediction improve outcomes without increasing surveillance burden or inequity?¶
ORACLE2 combines inflammatory and clinical risks from 22 trial control groups, defining a strong basis for prediction but not yet an implementation pathway (PMID 40215991). Electronic monitoring can detect use patterns, yet outcome effects are variable and infrastructure-heavy (PMID 36986513). A prospective trial must link risk states to predefined actions, compare against simpler rules, audit alert fatigue and false reassurance, and include patients with limited connectivity or medicine access. Sharpened 2026-10-15: in the same ORACLE2 data, symptom burden (ACQ-5) added little prognostic information (aRR 1.09 per 0.5 points; ΔR² 0.02) and did not modify mepolizumab benefit, whereas combined high eosinophils and FeNO did (PMID 42624803). Any risk-to-action system that escalates on symptoms alone is therefore testing the weaker signal. Sharpened 2026-10-22: an adult prediction-model review rated six of eight reports at high risk of bias (Li 2026, PMID 42827866; full text, Table 7). The reviewed tools need clinical-impact testing as well as discrimination and calibration; this is a finding within that review’s scope, not a universal absence claim. → exacerbations and acute care, inhaler technique, adherence and self-management
OQ-8. Can cumulative systemic-corticosteroid harm become a preventable asthma outcome?¶
Short courses are associated with sepsis, venous thromboembolism and fracture in population data, while patients report that oral and inhaled corticosteroid adverse effects shape asthma-specific quality of life (PMID 28404617, 37643678). Research should define a validated cumulative toxicity measure, determine which bursts are avoidable, and test steroid-sparing pathways across primary, emergency and specialist care. The endpoint belongs alongside exacerbations—not hidden in safety appendices. Sharpened 2026-10-15: an EAACI task-force meta-analysis across upper and lower airway disease reports that cumulative exposure of 500 mg–1 g prednisolone-equivalent raises the risk of most adverse effects, and proposes 500 mg as the exposure to avoid (PMID 42011838; abstract-level). The open part is now validation and implementation of that threshold, not whether a dose–harm gradient exists. → red flags and safety, severe asthma and biologics
OQ-9. What is the smallest scalable intervention that improves adherence through partnership?¶
Adherence is produced by beliefs, routines, device usability, affordability and relationships rather than knowledge alone (PMID 30053965, 32210541, 34902272). Shared decision-making is acceptable and conceptually aligned, but implementation is inconsistent and outcome evidence remains limited (PMID 38613765, 28972652). Factorial pragmatic trials could separate technique coaching, regimen simplification, electronic feedback, pharmacist support and collaborative goal-setting, with adherence and severe attacks both measured. → inhaler technique, adherence and self-management
OQ-10. Can asthma trials intervene on structural constraints rather than merely adjust for them?¶
Income and employment are associated with asthma outcomes; racial inequities and area-level social determinants shape care and paediatric utilization (PMID 40688041, 36220057, 37455665). Most efficacy trials exclude or statistically adjust these realities. Pragmatic trials should test formulary navigation, home remediation, transport, community health workers, school nursing and same-day medication access as components of asthma treatment, with absolute effects by baseline constraint. → epidemiology and global burden, patient experience and advocacy
Tier 2 — important enabling questions¶
Measurement and mechanisms¶
- OQ-11. What constitutes a safe, usable child attack-decision system? Parent decision-making, school barriers and post-hospital transitions have each been studied, but not as one pathway across settings (PMID 35967097, 30287588). The system must work for different literacy levels and distinguish urgent escalation from routine adjustment. → asthma in children
- OQ-12. Do patient-weighted endpoints reorder severe-asthma treatment choices? Patients select goals involving function, confidence and steroid burden; the Severe Asthma Questionnaire was built to capture disease-specific quality of life (PMID 33577946, 29794132). Re-analysis and prospective head-to-head studies should test whether patient weighting changes rankings. → patient experience and advocacy
- OQ-13. Which obesity–asthma mechanisms are reversed by which weight-loss strategy? Obesity-associated asthma is heterogeneous, and available weight-loss trials are small (PMID 29627041, 30605347). Trials should distinguish mechanical, metabolic, inflammatory and behavioural pathways rather than treating body mass as the mechanism. Sharpened 2026-10-15: a systematic review found observational GLP-1 receptor agonist exposure consistently associated with fewer exacerbations, but airway-intrinsic effects remain mostly preclinical (PMID 42167405). Two placebo-controlled tests are registered. GATA-3 (NCT05254314) is a single-centre US semaglutide trial (~100 adults, ACQ-7 and serum periostin primary) that reached primary completion on 2026-09-28 with no results posted. WRAPT (ISRCTN91904759) plans 138 adults on tirzepatide versus saline for 52 weeks, ACQ-7 primary with mechanistic secondaries, recruiting from 2026-11-09. Both use symptom scores as primary outcome, so neither alone will separate weight-dependent from airway-intrinsic effects. → phenotypes, endotypes and treatable traits
- OQ-14. Can climate-adaptive asthma plans prevent harm? Wildfire smoke, pollen and extreme temperatures are increasing hazards (PMID 37739070, 36917187, 38311978). Forecast alerts, portable filtration, clean-air rooms, workplace protections and pre-agreed medication steps need comparative effectiveness testing. Sharpened 2026-10-15: 2026 EAACI GRADE guidance on pollen conditionally recommends exposure reduction, real-time pollen monitoring and combined weather–pollen–pollutant warnings, on moderate to very low certainty evidence (PMID 42359506). A guideline recommending these interventions on low-certainty evidence is itself the case for the comparative trials. → genetics, environment and prevention
- OQ-15. Does suppressing inflammation early prevent irreversible airflow limitation? Early budesonide improves control, but the relation between early treatment, remodelling and decades-later lung-function trajectory remains unresolved (PMID 12672309, 22386510). Long follow-up with imaging or physiological remodelling markers is needed. → mild and moderate asthma, airway immunobiology
Treatment strategy¶
- OQ-16. How should clinicians choose among overlapping biologic eligibilities? Network meta-analysis supports class efficacy, but a live PubMed and ClinicalTrials.gov re-query on 2026-08-30 confirmed that direct comparative evidence remains limited and indirect rankings inherit population differences (PMID 31395084). A platform trial with common eligibility, biomarker trajectories, comorbidity outcomes and switch rules would answer a question currently handled by inference. Sharpened 2026-10-15: a randomized switch trial its authors describe as the first in severe asthma (NIMBLE, NCT04718389; PMID 41738176) compared twice-yearly depemokimab with continued mepolizumab or benralizumab in responders and missed its non-inferiority margin (rate ratio 1.16, 95% CI 0.98–1.38). An indirect comparison found that depemokimab's apparent advantage disappeared after full covariate adjustment (PMID 42173441). Direct evidence now exists for convenience switching within one pathway, but still not for choosing between pathways. Sharpened 2026-10-22: a Medicare mepolizumab–dupilumab comparison adds evidence from an older population but remains observational, with lung function and type-2 biomarkers unavailable in claims (Germain 2026, PMID 42886433; full text, Limitations). A comparative trial should prespecify age and comorbidity strata rather than assume one universal ranking. Sharpened 2026-10-29: CHRONICLE self-controlled case-series data show that smoking history, prior exacerbations, FeNO and blood eosinophils modify the observed biologic–exacerbation association, whereas spirometry-based RCT exclusions (FEV1 % predicted, FEV1/FVC, reversibility) and COPD/bronchiectasis did not (Boakye-Dankwa 2026, PMID 42242757; full text, Table 2). A platform trial can therefore justify relaxing lung-function exclusions more readily than dropping type-2 or smoking criteria; the analysis pooled biologics and remains observational. → severe asthma and biologics
- OQ-17. When, if ever, should a successful biologic be reduced or stopped? Remission is increasingly attainable but loss-of-response risk, treatment burden, cost and patient preference have not been joined in a robust withdrawal framework (PMID 39549709, 33758515). A live ClinicalTrials.gov re-query on 2026-08-30 identified a phase 4 dupilumab withdrawal-strategy study (NCT06818019); it remained recruiting with no results posted on the 2026-10-29 API re-query. A 24-person uncontrolled series reported that dupilumab early responders could extend the dosing interval (median 4 weeks, range 2–6) while maintaining control (Greig 2026, PMID 42918302; abstract-level; bibliography only). The evidence gap is therefore outcome evidence from a controlled stopping or interval-extension trial, not the absence of any observation. → severe asthma and biologics
- OQ-18. Can macrolide-responsive asthma be identified without accelerating avoidable antimicrobial harm? Azithromycin reduces exacerbations in some persistent uncontrolled asthma, while later reviews emphasize population and regimen heterogeneity (PMID 28687413, 38296770). Biomarker enrichment, microbiology and resistance outcomes should be mandatory. Sharpened 2026-10-15: a GRADE meta-analysis of 19 RCTs (n=1,825) found macrolides likely reduce severe exacerbations (IRR 0.75, 0.57–0.98) with similar relative effects in type-2-high and type-2-low disease (PMID 41654262; abstract-level). If that holds, the question shifts from "which T2-low patients" to "which patients of either endotype justify the resistance cost", and it bears on OQ-4: macrolides are not a T2-low-specific answer. → severe asthma and biologics
- OQ-19. How should biologics be used across conception, pregnancy and lactation? Observational safety data are expanding and an international consensus exists, but comparative fetal, maternal and asthma-control evidence remains sparse (PMID 39216499, 36411004). A live ClinicalTrials.gov re-query on 2026-08-30 found completed pregnancy-exposure studies for omalizumab (NCT00373061) and dupilumab (NCT04173442), plus a terminated benralizumab study (NCT03794999). The remaining evidence gap is harmonized, adequately adjusted comparative evidence with exposure timing and childhood follow-up, not the absence of prospective registry activity. → pregnancy and reproductive health
Systems and lived experience¶
- OQ-20. What closes the gap between action-plan possession and usable emergency knowledge? Supported self-management is effective at system level, yet attack interviews continue to show uncertainty about recognition and response (PMID 28302126, 39773014). Teach-back, rehearsal and just-in-time support should be tested against document delivery alone. → inhaler technique, adherence and self-management
- OQ-21. How should severe-asthma services measure diagnostic harm? The path to severe-asthma diagnosis can be prolonged, while both overdiagnosis and underdiagnosis remain documented (PMID 39184910, 29756989). Time-to-correct-diagnosis, inappropriate steroid exposure and missed mimic should become service outcomes. Sharpened 2026-10-29: SHARP shows that European severe-asthma registries already capture type-2 and pulmonary comorbidity more consistently than steroid-related harm, with 41% of surveyed specialists lacking multidisciplinary-team access (Rink 2026, PMID 42936732; full text). Service metrics for diagnostic and steroid harm will not be comparable until recording of those comorbidities is standardised. → diagnosis and objective testing, comorbidity and mimics
- OQ-22. Can school policy reduce both stigma and medical risk? Young people actively manage disclosure and inhaler visibility, while caregivers and school nurses report practical barriers after hospitalization (PMID 30461359, 30287588). Policy trials should measure access, confidence, absenteeism, attacks and unintended stigma. → asthma in children
- OQ-23. Which components of digital asthma care create benefit? Digital adherence interventions and monitoring devices are heterogeneous packages (PMID 35691614, 36986513). Dismantling studies should separate reminders, feedback, clinician response, technique sensing and relationship effects—and report the digital divide. → inhaler technique, adherence and self-management
- OQ-24. How can global mortality be reduced where inhaled corticosteroids and acute care are least accessible? GBD 2021 documents persistent burden and unequal mortality (PMID 40147466). Implementation studies must test procurement, workforce, diagnosis, emergency referral and affordable ICS-containing care as a connected system rather than isolated education. Sharpened 2026-10-15: CHEST-Africa (4,477 school children and adolescents, four countries) found that over half of those with self-reported asthma had severe symptoms, about one in five had a diagnosis, and fewer than 5% used ICS. Objective testing confirmed only 41.6–61.0% of self-reported cases (PMID 42744587). Diagnosis and ICS access, not prevalence ascertainment, look like the binding constraint there. Latin American evidence could not be checked because SciELO/LILACS are unreachable (HTTP 403), so no claim is made about how far this transports beyond the four countries. → epidemiology and global burden
Closed¶
(None yet. Move a question here only when a study or convergent evidence directly resolves it; record the closing date and citation.)