Skip to content

Biomarkers

TL;DR — No single biomarker diagnoses asthma, measures control, predicts every attack and selects every drug. Blood eosinophils and FeNO are the most accessible type-2 markers: they provide partly independent information, are prognostic for attacks, and enrich response to some biologics, but vary with time, corticosteroids, atopy, smoking and infection. FeNO diagnostic meta-analysis found sensitivity 0.65 and specificity 0.82; it is better for ruling in type-2-supported asthma than ruling it out (Karrasch 2017, PMID 27388487). Biomarker-guided care can reduce exacerbations—adult FeNO-guided treatment OR 0.61 (95% CI 0.44–0.83) for at least one attack—but algorithms, thresholds and steroid exposure differ (Korevaar 2023, PMID 37293870). Always state the decision, timing, threshold and confounders before ordering a marker.

Start with the clinical decision

Decision Helpful marker What it cannot do alone
Support diagnosis FeNO, eosinophils, sensitization Prove variable airflow or exclude non-type-2 asthma
Estimate attack risk Blood eosinophils + FeNO + prior attacks Predict the date or trigger of an attack
Test adherence/ICS response FeNO suppression in selected high-FeNO disease Measure every dose or inhaler technique
Choose biologic Eosinophils, FeNO, allergen sensitization/IgE, comorbidity Rank all biologics head-to-head for an individual
Monitor response Attacks, OCS, control, lung function plus relevant marker Replace clinical outcomes
Identify type-2-low disease Repeatedly low markers off confounding systemic steroids Define one stable non-type-2 mechanism

A biomarker becomes useful only when its result changes an action. Pre-analytic context is part of the result.

Type-2 biology in one table

Marker Main biological source/signal Practical strengths Major limitations
Blood eosinophils Systemic IL-5-associated eosinophilia Cheap, available, validated thresholds in trials Variable; suppressed by steroids; elevated in other diseases
Sputum eosinophils Airway inflammatory cell proportion Closest routine measure of airway eosinophilia Specialist induction/processing; sampling failure
FeNO Epithelial iNOS signal driven partly by IL-4/IL-13 Noninvasive, immediate, repeatable Atopy, height, smoking, infection, nitrate and ICS affect it
Total IgE Systemic IgE burden Required with weight for omalizumab dosing ranges Poor specificity; level does not equal allergen causality
Specific IgE/skin testing Sensitization Connects exposure and allergic phenotype Sensitization is not necessarily clinically relevant allergy
Serum periostin IL-13-associated matrix protein Mechanistic/clinical-trial interest Bone and other tissues contribute; assay/threshold issues

Type-2 markers correlate but are not redundant. Eosinophils reflect circulating effector cells; FeNO more directly reflects airway epithelial IL-4/IL-13 signaling; IgE/sensitization describes allergic recognition.

Blood eosinophils

Blood eosinophil count is reported as absolute cells/µL or ×10⁹/L. Trial thresholds commonly include 150, 300 or 400 cells/µL, but they are treatment- and population-specific, not universal biological borders.

Interpretation requires:

  • current and recent oral/injected corticosteroids;
  • ICS exposure and adherence;
  • time since attack or infection;
  • diurnal and within-person variability;
  • parasitic exposure, drug reaction, hematologic disease and eosinophilic systemic disease;
  • whether the value is current or the historical maximum.

Repeated measurement is often more informative than a single low result. A count obtained after systemic corticosteroids cannot safely establish “type-2-low” disease.

Blood eosinophils predict attack risk at the population level and enrich response to anti-IL-5/5R and dupilumab, but the relationship differs by drug and outcome. In children receiving dupilumab, higher baseline eosinophils and FeNO independently predicted greater exacerbation and lung-function benefit (Bacharier 2024, PMID 38272375).

Unexpected marked eosinophilia is a diagnostic signal, not merely a biologic eligibility number. Consider eosinophilic granulomatosis with polyangiitis, allergic bronchopulmonary aspergillosis, parasites, drug reactions and hypereosinophilic syndromes in the right context.

Sputum eosinophils and neutrophils

Induced sputum differential counts can identify eosinophilic, neutrophilic, mixed or paucigranulocytic patterns. They sample airway inflammation more directly than blood, but reproducibility depends on collection, processing and corticosteroid timing.

In pooled randomized evidence, sputum-eosinophil-guided treatment reduced the proportion experiencing at least one exacerbation from 82 to 62 per 100, OR 0.36 (95% CI 0.21–0.62), without consistent improvement in daily symptoms or lung function (Petsky 2018, PMID 29858277).

Neutrophilic sputum is not a single endotype. Infection, smoking, pollution, recent corticosteroid exposure and severe disease can contribute. A high neutrophil percentage does not by itself justify antibiotics or macrolides.

FeNO

FeNO is measured at a standardized expiratory flow and reported in parts per billion. It is most useful as a continuous, contextual measure rather than a binary test.

Diagnostic performance

Across 26 studies and 4,518 participants, FeNO had pooled sensitivity 0.65 (95% CI 0.58–0.72), specificity 0.82 (0.76–0.86), diagnostic OR 9.23 (6.55–13.01) and AUC 0.80 (0.77–0.85). Specificity increased with higher thresholds by OR 1.46 per 10-ppb increase (Karrasch 2017, PMID 27388487).

Thus:

  • high FeNO can support type-2 airway inflammation in a compatible case;
  • low FeNO does not exclude asthma;
  • a positive result does not demonstrate variable airflow;
  • the test is harder to interpret after ICS has started.

ATS interpretive guidance uses low/intermediate/high ranges that differ for adults and children and urges cautious interpretation in the intermediate range (Dweik 2011, PMID 21885636). Local devices and guidelines should supply the operational cutoffs.

Confounders

Raises FeNO Lowers FeNO or masks signal
Atopy/allergen exposure ICS and systemic corticosteroids
Eosinophilic airway inflammation Cigarette smoking
Some viral/inflammatory states Poor exhalation technique/low production states
Height/age-related normative differences Recent effective biologic targeting type-2 pathway

Smoking asthma has lower FeNO on average, but FeNO can still be informative; dedicated review found limited data and a preserved response signal in some smokers (Ahovuo-Saloranta 2019, PMID 30861509).

FeNO-guided treatment

An updated adult meta-analysis included 12 RCTs and 2,116 patients. FeNO-guided care reduced the odds of at least one exacerbation, OR 0.61 (95% CI 0.44–0.83), and exacerbation rate, RR 0.67 (0.54–0.82), with moderate certainty; all trials had high or unclear risk of bias in at least one domain (Korevaar 2023, PMID 37293870).

In pediatric meta-analysis of 23 RCTs, FeNO guidance reduced the number of children with exacerbations, RR 0.73 (95% CI 0.63–0.84), but increased daily ICS dose by a weighted mean 64.17 μg (53.59–74.75) (Wang 2020, PMID 32525611).

Earlier pediatric evidence found lower odds of more than one exacerbation but no clear FEV1 or steroid-use difference (Lu 2015, PMID 25634163). These differences show that “FeNO-guided” is not one intervention: thresholds, escalation rules and control arms matter.

FeNO suppression

In a patient with high FeNO despite prescribed ICS, directly observed or electronically monitored ICS followed by repeat FeNO can test whether the signal suppresses. Suppression supports steroid-responsive biology and may reveal nonadherence, but lack of suppression can reflect severe biology, continuing exposure or incorrect assumptions (Butler 2021, PMID 33369570).

IgE and sensitization

Total IgE is neither an asthma diagnostic test nor a severity scale. Its main current treatment role is confirming that a patient fits product-specific omalizumab dosing/eligibility rules alongside body weight and clinically relevant perennial allergen sensitization.

Specific IgE and skin-prick testing demonstrate sensitization. Clinical relevance requires concordant exposure and symptoms. Broad panels create incidental positives and can prompt burdensome avoidance without benefit.

IgE genetics and ancestry are complex; a multi-ancestry GWAS meta-analysis identified shared and ancestry-specific loci, warning against treating one population’s reference distribution as universal (Levin 2013, PMID 23146381).

Periostin

Periostin is induced by IL-13 and participates in extracellular-matrix biology. It was developed as a candidate companion marker for type-2 targeted therapy and remodeling (Matsumoto 2014, PMID 24759559).

Diagnostic meta-analyses found moderate accuracy. One pooled nine studies/1,448 asthma cases: sensitivity 0.68 (95% CI 0.42–0.86), specificity 0.87 (0.76–0.93), DOR 13.78 (4.64–40.93) (Jia 2021, PMID 31738608). Another nine-study analysis found sensitivity 0.58 and specificity 0.86 with heterogeneous methods (Yang 2020, PMID 32605695).

Serum periostin is not a routine stand-alone diagnostic or universal biologic-selection test. Assay standardization, age/bone turnover and overlapping distributions limit portability.

Combining biomarkers for attack risk

Prior severe attack remains one of the strongest clinical predictors. Biomarkers add information; they do not replace history.

ORACLE2 pooled control-group individual data from 22 RCTs to estimate incremental risk from blood eosinophils and FeNO alongside clinical variables. Its central contribution is joint modeling of two treatable type-2 signals rather than choosing one marker (Meulmeester 2025, PMID 40215991).

A practical risk matrix:

Eosinophils FeNO Interpretation to test
High High Concordant type-2 activity; check adherence/exposure and attack risk
High Low IL-5/eosinophil-predominant signal, smoking or treated airway IL-13 signal
Low High Airway IL-4/13 signal, atopy or recent systemic-steroid suppression of blood cells
Low Low Treated type-2 disease, true type-2-low biology or mistimed measurement

Never infer stability from one quadrant without treatment timing and repeated values.

Selecting and monitoring biologics

Target Biomarker/trait that enriches response Caveat
IgE Sensitization + total IgE/weight within dosing rules IgE level alone poorly ranks benefit
IL-5/IL-5R Recurrent attacks + blood eosinophils Steroids can hide eligibility signal
IL-4Rα Eosinophils and/or FeNO; atopic comorbidity Eosinophilia can transiently rise
TSLP Broad severe-asthma eligibility; benefit across marker strata Greater absolute benefit often occurs at higher baseline risk/type-2 signal

Tezepelumab reduced exacerbations across individual baseline biomarker strata in PATHWAY while lowering eosinophils, FeNO, IgE and other type-2 markers; subgroup breadth does not mean biomarkers are irrelevant to absolute benefit (Corren 2022, PMID 34913186).

In 3,751 real-world biologic initiators, higher pre-biologic eosinophils and FeNO were associated with greater FEV1 improvement for anti-IgE and anti-IL-5/5R, but baseline eosinophils were not associated with exacerbation change for every biologic class (Porsbjerg 2024, PMID 38711495).

Small cohorts can overfit thresholds: a 24-patient benralizumab study reported baseline eosinophil and FeNO cutoffs predicting response, but validation is required before bedside adoption (Watanabe 2022, PMID 34348060).

After biologic initiation, measure meaningful outcomes at a prespecified interval: severe attacks, OCS exposure, validated control, lung function, adverse effects and patient priorities. Biomarker depletion without clinical benefit is not success.

Biomarkers not ready for routine selection

Research candidates include volatile organic compounds, nasal transcriptomics, urinary eicosanoids, microRNAs, proteomics, metabolomics, mucus-plug imaging and microbiome signatures. Severe-asthma cohort reviews describe promise but limited external validation, assay standardization and clinical utility (Lee 2021, PMID 33733634; Narendra 2019, PMID 31735516).

Before adoption, require:

  1. analytic validity;
  2. reproducible clinical association;
  3. incremental value beyond simple variables;
  4. a defined action threshold;
  5. prospective evidence that acting improves outcomes;
  6. equity and cost assessment.

Reporting template

Field Example
Decision Assess type-2 activity before biologic selection
Result Blood eosinophils 420 cells/µL; FeNO 48 ppb
Timing Stable visit, 6 weeks after last OCS
Treatment context High-dose ICS/LABA; refill/monitor data summarized
Confounders Atopy present; nonsmoker; no current infection
Interpretation Concordant type-2-high signal, not proof of adherence or diagnosis
Action Confirm eligibility and choose with comorbidity/access/preferences

Open questions

  • What repeated-measure schedule best distinguishes stable biology from treatment/exposure variation?
  • Do composite eosinophil–FeNO algorithms improve outcomes beyond prior attacks and adherence data? (Meulmeester 2025, PMID 40215991)
  • Which markers can select effective therapy for type-2-low asthma?
  • Can biomarker-guided step-down reduce steroid exposure without increasing attacks?
  • How can assays and thresholds be validated across ancestry, age, smoking and low-resource settings?

References

  1. Karrasch S, et al. Accuracy of FeNO for diagnosing asthma: systematic review. Thorax. 2017. PMID 27388487
  2. Petsky HL, et al. Tailoring treatment on eosinophilic markers: systematic review and meta-analysis. Thorax. 2018. PMID 29858277
  3. Korevaar DA, et al. FeNO-guided treatment in adults: systematic review and meta-analysis. Clin Exp Allergy. 2023. PMID 37293870
  4. Wang X, et al. FeNO-guided management in children: systematic review and meta-analysis. Pediatr Pulmonol. 2020. PMID 32525611
  5. Lu M, et al. FeNO and asthma treatment in children: systematic review and meta-analysis. Medicine. 2015. PMID 25634163
  6. Dweik RA, et al. ATS guideline: interpretation of FeNO. Am J Respir Crit Care Med. 2011. PMID 21885636
  7. Ahovuo-Saloranta A, et al. FeNO in smoking asthmatics: systematic review. J Breath Res. 2019. PMID 30861509
  8. Butler CA, et al. Fractional exhaled nitric oxide and asthma treatment adherence. Curr Opin Pulm Med. 2021. PMID 33369570
  9. Bacharier LB, et al. Eosinophils and FeNO as prognostic and predictive biomarkers in childhood asthma. J Allergy Clin Immunol Pract. 2024. PMID 38272375
  10. Levin AM, et al. GWAS meta-analysis for total IgE in diverse populations. J Allergy Clin Immunol. 2013. PMID 23146381
  11. Matsumoto H. Serum periostin as an asthma biomarker. Allergol Int. 2014. PMID 24759559
  12. Jia X, et al. Diagnostic accuracy of periostin: systematic review and meta-analysis. J Asthma. 2021. PMID 31738608
  13. Yang L, et al. Serum periostin in asthma diagnosis: meta-analysis. Allergy Asthma Proc. 2020. PMID 32605695
  14. Meulmeester FL, et al. Inflammatory and clinical risk factors for asthma attacks: ORACLE2. Lancet Respir Med. 2025. PMID 40215991
  15. Corren J, et al. Baseline type-2 biomarkers and response to tezepelumab. Allergy. 2022. PMID 34913186
  16. Porsbjerg CM, et al. Pre-biologic type-2 biomarker combinations and real-world response. Eur Respir J. 2024. PMID 38711495
  17. Watanabe H, et al. Eosinophils and FeNO predicting benralizumab effectiveness. J Asthma Allergy. 2022. PMID 34348060
  18. Lee Y, et al. Biomarkers for severe asthma: longitudinal cohorts. Allergy Asthma Immunol Res. 2021. PMID 33733634
  19. Narendra D, et al. Immunological biomarkers in severe asthma. Semin Immunol. 2019. PMID 31735516
  20. Gvalani A, et al. Biomarkers in severe asthma: identifying the treatable trait. Lung India. 2023. PMID 36695260
  21. Essat M, et al. FeNO for management of asthma in adults: systematic review. Eur Respir J. 2016. PMID 26846832
  22. Jartti T, et al. Childhood asthma management guided by repeated FeNO. Pediatr Respir Rev. 2012. PMID 22726875
  23. Petsky HL, et al. Tailoring asthma treatment on eosinophilic markers. Thorax. 2012. PMID 20937641
  24. Fahy JV. Type 2 inflammation in asthma. Nat Rev Immunol. 2015. PMID 25534623
  25. Niessen NM, et al. Type-2-low asthma: past, present and future biologic therapies. Eur Respir Rev. 2022. PMID 35487388
  26. Malinovschi A. Limited use of biomarker-guided therapy in mild asthma. Lancet Respir Med. 2020. PMID 32171066