Performance of Eosinophil Cationic Protein as a Biomarker in Asthmatic Children

Abstract

Background.

Although blood eosinophils are a frequently utilized marker of Type-2 inflammation in children with asthma, their sensitivity is relatively poor. Additional markers of Type-2 inflammation are needed.

Objective.

We hypothesized that plasma concentrations of eosinophil cationic protein (ECP), a marker of eosinophil activation, would be useful for detection of Type-2 inflammation and would predict poorer asthma outcomes over one year.

Methods.

Children and adolescents 6 through 17 years (N=256) with confirmed asthma completed a baseline visit and a follow-up visit at 12 months. A subset also underwent systemic corticosteroid responsiveness testing with intramuscular triamcinolone. Outcome measures at 12 months included uncontrolled asthma, lung function, and asthma exacerbations treated with systemic corticosteroids.

Results.

Plasma ECP concentrations ranged from 0.03 to 413.61 ng/mL (median, 6.95 ng/mL) and were consistently associated with other markers of T2-inflammation. At baseline, children in the highest ECP tertile had poorer asthma control, more airflow limitation and more exacerbations, but also had greater symptom improvement with intramuscular triamcinolone. At 12 months, associations between the highest ECP tertile and exacerbations, but not lung function or asthma control, persisted after covariate adjustment. However, the sensitivity of ECP was modest and was not markedly different from that of blood eosinophil counts.

Conclusion.

Plasma ECP concentrations may be a useful marker of Type-2 inflammation in children and may help identify those children at highest risk for recurrent exacerbations who could benefit from corticosteroid treatment. However, additional markers may be needed to improve sensitivity for outcome detection.

Introduction

Type-2 inflammation, which involves the recruitment and activation of eosinophils by Th2 cytokines and other pathways, is thought to be a pivotal risk factor for persistent wheezing and poor asthma control in children.¹ Children with Type-2 inflammation, compared to those with non-Type-2 inflammation, often have increased symptoms, impaired lung function measures with increased airway hyperresponsiveness, and more frequent exacerbations.^2–5 This Type-2 “high” asthma phenotype also tends to be more corticosteroid responsive.^3–6 Therefore, identification of children with Type-2 “high” asthma in the clinical setting may be useful for selection of the most appropriate asthma controller medications. However, there is no consensus regarding the “best” biomarker for Type-2 inflammation, or the appropriate cut-points in pediatric populations.^{7, 8}

The blood eosinophil count is currently the most accepted biomarker of Type-2 inflammation in routine clinical practice, but there are also challenges with its measurement. After eosinophils mature in the bone marrow, they are present in the circulation for 8–12 hours before migrating to tissue, where they remain for 7–14 days.⁹ In contrast, eosinophil cationic protein (ECP) is an eosinophil granule protein that is primarily released after the eosinophil has left the circulation,¹⁰ with a turnover time in the blood of approximately 45 minutes.¹¹ ECP is therefore considered a marker of eosinophil activation¹² and, like blood eosinophils, has been associated with several clinical features of atopic asthma.¹³

Although ECP has been of interest in the asthma field for several decades, nearly all the existing studies of ECP in children are cross-sectional in nature and have employed either a case/control design or correlative analyses.¹³ In the few clinical trials that examined ECP as a predictor of treatment response, ECP concentrations were measured at only a single time point.^5,14 Other clinical trials of asthma controller medications or longitudinal studies that measured ECP pre-and post-treatment were limited to very small samples of children (n<35) with mild asthma.^15–18 Phenotyping in these studies was also quite limited. Therefore, major gaps related to the clinical utility of ECP in real world pediatric asthma populations still persist.

We questioned whether plasma ECP concentrations would be a useful biomarker of Type-2 inflammation at multiple timepoints and whether plasma ECP concentrations could discriminate clinical outcomes in children with asthma. We hypothesized that: 1) ECP concentrations in children with asthma would correlate with blood eosinophil counts, the magnitude of aeroallergen sensitization, serum IgE levels, exhaled nitric oxide concentrations, and plasma concentrations of interleukin (IL)-4, IL-5, IL-13 and CC chemokine ligand (CCL)-11 (eotaxin), and that 2) higher ECP concentrations would be associated with poorer asthma control, greater airflow limitation, and a greater occurrence of exacerbations over 12 months of follow-up. We further hypothesized that children with higher ECP concentrations would have a greater response to treatment with systemic corticosteroids.

Methods

Children 6 through 17 years of age with physician-diagnosed atopic asthma across the severity spectrum treated with controller medications for at least 8 weeks were eligible for the study if they had: 1) current or historical evidence of ≥ 12% reversibility in their forced expiratory volume in one second (FEV₁) relative to baseline after bronchodilator administration or 2) airway hyperresponsiveness to methacholine, with a provocative concentration of methacholine causing a 20% drop in FEV₁ (PC₂₀) of ≤16 mg/mL. Atopy was defined as either sensitization to at least one aeroallergen, blood eosinophil counts >300 cells/microliter, or physician-diagnosed allergic rhinitis or atopic dermatitis. Exclusion criteria included premature birth before 35 weeks of gestation and other chronic airway disorders that could mimic asthma such as pulmonary aspiration or vocal cord dysfunction. Permission to proceed with this study was granted by the Emory University Institutional Review Board. Informed written consent was obtained from legal guardians. Verbal assent was obtained from children 6–10 years and written assent was obtained from children and adolescents 11 to 17 years.

Study design and outcome measures.

Participants completed a baseline outpatient research visit and a follow-up outpatient research visit at 12 months. Research visits were scheduled between 9 am and 2 pm on weekdays. These visits were postponed if an asthma exacerbation treated with systemic corticosteroids was reported within the preceding two weeks. After the baseline visit, a subset of children also underwent systemic corticosteroid responsiveness testing with intramuscular triamcinolone acetonide (1 mg/kg, 60 mg maximum dose) administered in the gluteal muscle. Corticosteroid responsiveness was assessed at 14 days (window +7 days) after triamcinolone receipt and was quantified by asthma control and lung function measures. Outcome measures at 12 months included uncontrolled asthma at 12 months, lung function at 12 months, and the occurrence of any asthma exacerbation treated with systemic corticosteroids in the prior 12 months.¹⁹

Procedures.

The schedule of procedures is shown in Figure 1. Participants and their caregivers completed questionnaires pertaining to symptoms, medical history and demographics. Asthma control was assessed with the 6-item Asthma Control Questionnaire (ACQ)²⁰ administered by a trained interviewer.²¹ Participants also completed the Asthma Control Test (ACT)²² or the Childhood Asthma Control Test (CACT).²³ Asthma-related quality of life was assessed with the Pediatric Asthma Quality of Life Questionnaire (PAQLQ)²⁴ with technical assistance recommended for younger children,²⁵ where higher PAQLQ scores reflect better quality of life. Exhaled nitric oxide concentrations were measured according to technical standards²⁶ with a commercial device (NIOX MINO®, Circrassia Pharmaceuticals, Chicago, IL). Aeroallergen sensitization was assessed by specific IgE testing by the ImmunoCAP method by Phadia (Children’s Healthcare of Atlanta, Atlanta, GA) or skin prick testing with 12 extracts: tree mix, grass mix, weed mix, Ambrosia artemisiifolia, Alternaria alternata, Aspergillus fumagatis, Cladosporium herbarum, dog dander, cat dander, Blatella germanica, Dermatophagoides farinae, and Dermatophagoides pteronyssinus (Greer® Laboratories, Lenoir, NC). Venipuncture was also performed for quantification of blood eosinophils (Children’s Healthcare of Atlanta) and total serum IgE (Children’s Healthcare of Atlanta). Spirometry (KoKo® PDS, Ferraris, Louisville, CO) was performed according to technical standards²⁷ with 360 mcg albuterol sulfate for bronchodilator reversibility testing. The best of three forced vital capacity (FVC) maneuvers was interpreted according to Global Lung Function Initiative prediction equations.²⁸ FVC, FEV₁. FEV₁/FVC and the forced expiratory flow at 25–75% of vital capacity (FEF_25–75) were expressed as percentage of predicted values. A subset of participants also underwent bronchoprovocation testing with methacholine concentrations of 0 to 16 mg/mL (Provocholine; Methapharm Inc, Coral Springs, FL) delivered by a Rosenthal dosimeter (Pulmonary Data Service Instrumentation, Louisville, CO). Bronchoprovocation was limited to participants with baseline FEV₁ >70% predicted.

Figure 1.

Inclusion of participants and procedures performed at each visit.

ECP was quantified in plasma using a commercially available kit (TSZ Biological catalog #HU8604, San Francisco, CA). Cytokines and chemokines were quantified in the plasma with multiplex assays for IL-4, IL-5, IL-13, and CCL-11/eotaxin according to technical standards (MilliporeSigma, #HSTCMAG-28SK and #HCYTOMAG-60K, Burlington, MA).

Statistical analyses.

Data were analyzed with SPSS® Statistics (Version 26, IBM, Armonk, NY). Data that were not normally distributed were logarithmically transformed prior to analyses. ECP tertiles were compared with Chi-square tests and ANOVA, with Tukey’s Least significant difference tests for post-hoc comparisons. Linear and logistic regression analyses were performed with covariate adjustment as indicated. Receiver operating characteristic curves were generated to evaluate sensitivity and specificity. A p-value <0.05 was used as the threshold for statistical significance without adjustment for multiple comparisons.

Results

Two-hundred fifty six children were enrolled and completed the baseline characterization visit (Figure 1). Their features are shown in Table 1. Plasma ECP concentrations ranged from 0.03 to 413.61 ng/mL (median, 6.95 ng/mL; 25–75^th percentile, 1.81 – 23.75 ng/mL). ECP concentrations did not differ by age, duration of asthma, ethnicity, obesity, or household tobacco smoke exposure. However, ECP concentrations did differ by sex, race, allergic rhinitis and eczema (Figure E1). ECP concentrations were also significantly higher in children treated with high-dose inhaled corticosteroids (no inhaled corticosteroid treatment: median, 4.29 ng/mL; 25–75^th percentile, 0.94 – 15.34 ng/mL; low-to-medium dose treatment: median, 5.88 ng/mL, 25–75^th percentile, 1.33 – 20.07 ng/mL; high-dose treatment: 11.87 ng/mL; 25–75^th percentile, 3.08 – 26.72 ng/mL; p=0.001). The differences in ECP by sex, race, allergic rhinitis and eczema persisted after adjustment for blood eosinophil counts, inhaled corticosteroid dose, and uncontrolled asthma (Table E1). ECP concentrations were also associated with other markers of T2-inflammation, including blood eosinophils, exhaled nitric oxide, the percentage of positive aeroallergens, and serum IgE (Figure 2A–D). There were no associations between plasma ECP concentrations and concentrations of plasma IL-4 (r=0.078, p=0.236), IL-5 (r=−0.012, p=0.861), IL-13 (r=0.042, p=0.528), or CCL-11/eotaxin (r=−0.027, p=0.683).

Table 1.

Features of the participants. Data represent the median (25^th, 75^th percentile) or the number of participants (%).

Feature		All participants N = 256
Age (years)		10.5 (8.5, 13.83)
Asthma duration (years)		8.0 (5.3, 11.3)
Male		158 (61.7)
Hispanic ethnicity		3 (1.2)
Race	White	28 (10.9)
	Black	214 (83.6)
	More than one race	14 (5.5)
Household education	Refused	14 (5.5)
	Did not complete high school	11 (4.3)
	High school	39 (15.2)
	Some college	89 (34.8)
	College degree	103 (40.2)
Parent with asthma		152 (61.3)
Smoker in household		37 (14.7)
Other diagnoses	Obesity	70 (27.3)
	Allergic rhinitis	221 (87.4)
	Atopic dermatitis	150 (59.3)
Asthma medications	Inhaled corticosteroid (any dose)	213 (83.2)
	High-dose inhaled corticosteroid	122 (47.7)
	Leukotriene receptor antagonist	156 (60.9)
	Long-acting beta-agonist	133 (52.0)

Figure 2.

Associations between eosinophil cationic protein (ECP) concentrations and (A) blood eosinophil counts, (B) exhaled nitric oxide concentrations, (C) the percentage of positive aeroallergens, and (D) serum IgE concentrations.

Associations between ECP and baseline asthma features.

Because there are no established reference intervals of normal for plasma ECP, children were stratified into tertiles according to ECP concentrations as follows: Low ECP, <3.12 ng/mL; Intermediate ECP, 3.12 – 16.22 ng/mL; and High ECP, ≥16.23 ng/mL. Baseline asthma features associated with each ECP tertile are shown in Table 2. Type-2 inflammatory features were elevated in both the Intermediate and High ECP groups, but the High ECP group was distinguished by a greater magnitude of aeroallergen sensitization. The Intermediate and High ECP groups also both had poorer asthma control, more airflow limitation, greater bronchodilator reversibility and airway hyperresponsiveness compared to the low ECP group. However, the high ECP group had significantly more exacerbations in the prior year (Table 2).

Table 2.

Baseline asthma and Type-2 inflammatory features, by eosinophil cationic protein (ECP) tertile. Data represent the mean ± standard error of the mean or the number of participants (%).

	Low ECP (<3.12 ng/ml) N=85	Intermediate ECP (3.12–16.22 ng/mL) N=85	High ECP (≥16.23 ng/mL) N=86
Type-2 inflammatory markers
Blood eosinophils (cells/microliter)	230.1 ± 18.3	307.6 ± 27.4*	375.9 ± 26.8*
Blood eosinophils (%)	3.47 ± 0.31	5.25 ± 0.49*	6.23 ± 0.46*
Exhaled nitric oxide (ppb)	20.8 ± 2.9	36.5 ± 3.6*	36.9 ± 27.2*
Serum IgE (kU/L)	119.9 ± 14.8	426.8 ± 273.5*	1890.6 ± 169.2*^
% positive aeroallergens (of 12)	12.1 ± 1.8	35.0 ± 2.9*	43.5 ± 3.0*^
Exacerbation history (past year)
Any systemic corticosteroid burst	52 (61.2)	48 (56.5)	65 (75.6)*^
Emergency department visit	38 (44.7)	45 (52.9)	52 (60.5)*
Hospitalization	21 (24.7)	19 (22.6)	24 (27.9)
Asthma control
ACQ-6 score	0.96 ± 0.15	1.13 ± 0.14	1.14 ± 0.12
ACT or CACT score	20.5 ± 4.2	20.2 ± 4.0	18.8 ± 4.6
Uncontrolled asthma	32 (38.6)	45 (53.6)*	47 (54.7)*
PAQLQ score	5.57 ± 0.17	5.57 ± 0.16	5.50 ± 0.14
Lung function
FVC (% predicted)	101.3 ± 1.4	102.4 ± 1.9	101.8 ± 1.8
FEV1 (% predicted)	94.1 ± 1.4	90.9 ± 2.1*	88.3 ± 2.0*
FEV1/FVC (% predicted)	92.6 ± 1.1	88.2 ± 1.3*	86.2 ± 1.2*
FEF25–75 (% predicted)	80.7 ± 2.9	71.4 ± 3.2*	66.7 ± 2.9*
Bronchodilator reversibility
FEV1 absolute reversibility (%)1	7.4 ± 1.0	11.8 ± 1.3*	13.0 ± 1.2*
FEV1 relative reversibility (%)2	8.2 ± 1.1	15.4 ± 2.0*	17.0 ± 1.8*
PC20 (mg/mL methacholine)3	14.9 ± 2.8	5.1 ± 1.6*	2.5 ± 1.0*

^*p<0.05 vs. Low ECP

^{^}p<0.05 vs. Intermediate ECP

¹Defined as: % predicted FEV_{1 post-bronchodilator} - % predicted FEV_{1 baseline}

²Defined as: (FEV_{1 post-bronchodilator} - FEV_{1 baseline})/FEV_{1 baseline} * 100

³Low ECP, n=14; Intermediate ECP, n=19; High ECP, n=23

Associations between ECP and corticosteroid responsiveness.

Ninety-one children consented to treatment with intramuscular triamcinolone. Their features were not significantly different from the baseline sample (mean age, 12.2 ± 3.3 years; 66.7% male; 72.5% Black). After triamcinolone treatment, ECP concentrations remained elevated in the High ECP group (Low ECP: 1.4 ± 0.9 ng/mL; Intermediate ECP: 14.9 ± 6.7 ng/mL; High ECP: 56.8 ± 7.3 ng/mL; p<0.001). In contrast, blood eosinophil counts and exhaled nitric oxide concentrations were markedly reduced after triamcinolone irrespective of ECP grouping (Figure 3) and correlations between ECP, blood eosinophil counts and exhaled nitric oxide were no longer present (ECP vs. eosinophil count, r=−.074, p=0.743; ECP vs. exhaled nitric oxide, r=0.268, p=0.068).

Figure 3.

(A) Eosinophil cationic protein (ECP) concentration, (B) blood eosinophil counts, and (C) exhaled nitric oxide concentrations in Low ECP (red), Intermediate ECP (yellow), and High ECP (blue) groups, pre- and post-triamcinolone administration. *p<0.05

Asthma clinical outcomes after receipt of intramuscular triamcinolone are shown in Table 3. Triamcinolone significantly improved (i.e., reduced) ACQ-6 scores in the High ECP group, reflected by a larger mean difference in the ACQ-6 score from baseline, but did not significantly alter asthma quality of life or lung function measures.

Table 3.

Asthma outcomes after intramuscular triamcinolone administration, by baseline eosinophil cationic protein (ECP) tertile. Data represent the mean ± standard error of the mean.

	Low ECP (<3.12 ng/ml)	Intermediate ECP (3.12–16.22 ng/mL)	High ECP (≥16.23 ng/mL)
ACQ-6 score	0.55 ± 0.12	0.95 ± 0.22	0.96 ± 0.14*
Mean difference from baseline1	0.16 ± 0.14	0.29 ± 0.16	0.36 ± 0.14*
PAQLQ score	6.24 ± 0.14	5.77 ± 0.22	5.79 ± 0.16
Mean difference from baseline	−0.60 ± 0.16	−0.48 ± 0.18	−0.48 ± 0.12
FEV1 (% predicted)	100.1 ± 2.7	93.6 ± 3.5	88.4 ± 2.9*
Mean difference from baseline	−0.97 ± 1.78	0.61 ± 2.92	−2.66 ± 2.46
FEV1/FVC (% predicted)	93.9 ± 1.6	89.7 ± 2.4	87.4 ± 1.8*
Mean difference from baseline	0.79 ± 1.63	0.25 ± 1.72	2.10 ± 1.41

¹Pre-triamcinolone value minus post-triamcinolone value

^*p<0.05 vs. Low ECP

Associations between ECP and outcomes at one year.

One hundred ninety-eight participants (80%) provided outcome data at 12 months. Children lost to follow up were more frequently black (98.3%) with higher household educational attainment (52.5% with college degree), less intense medication requirements (18.6% and 22.0% treated with high-dose inhaled corticosteroids and long-acting beta-agonists, respectively), and better asthma control (37.3% with ACQ-6 score <1.50 or ACT or CACT score<20). Therefore, no attempts were made to impute the final dataset.

ECP concentrations, Type-2 biomarkers and asthma outcomes at 12 months are shown in Table 4. At 12 months, ECP concentrations remained elevated in the High ECP group. 85% of children in the Low ECP group had ECP concentrations that remained “low” (i.e., <3.12 ng/mL), whereas 95% of children in the High ECP group had ECP concentrations that remained “high” (i.e., >16.23 ng/mL). The Intermediate ECP group was less stable, with only 50% of children remaining intermediate; 19% had concentrations <3.12 ng/mL and 31% had concentrations >16.23 ng/mL. Similar to observations at baseline, at 12 months of follow-up, ECP concentrations were again correlated with blood eosinophil percentages (r=0.362, p=0.010), blood eosinophil counts (r=0.319, p=0.024), and exhaled nitric oxide values (r=0.294, p=0.022). Children in the High ECP group also had a higher prevalence of uncontrolled asthma, defined as an ACQ-6 score <1.50 or ACT or CACT score<20, greater airflow limitation evidenced by lower FEV1/FVC and FEF_25–75 %predicted values, and a more prevalent exacerbation occurrence during the observation interval (Table 4). The associations between high ECP concentrations and exacerbation occurrence, but not high ECP concentrations and lung function or asthma control, persisted after adjustment for sex, race, household educational attainment and tobacco smoke exposure (Table E2). However, the sensitivity of ECP for detecting exacerbation occurrence was modest and was not markedly different from that of blood eosinophil counts or exhaled nitric oxide concentrations measured at the baseline visit (Figure 4). Inspection of the receiver operating characteristic curve revealed that lower ECP cut-points provide the best sensitivity, with modest specificity (Table E3).

Table 4.

Asthma features at 12 months, by baseline eosinophil cationic protein (ECP) tertile. Data represent the mean ± standard error of the mean or the number of participants (%).

	Low ECP (<3.12 ng/ml)	Intermediate ECP (3.12–16.22 ng/mL)	High ECP (≥16.23 ng/mL)
Type-2 inflammatory markers
ECP (ng/mL)	1.28 ± 0.44	19.50 ± 6.07*	62.29 ± 6.12*
Blood eosinophils (cells/microliter)	244.6 ± 27.2	353.1 ± 43.9*	271.1 ± 31.4
Blood eosinophils (%)	3.60 ± 0.45	6.10 ± 0.84*	4.73 ± 0.47*
Exhaled nitric oxide (ppb)	25.8 ± 5.2	40.7 ± 5.0*	36.0 ± 28.4*
Asthma control1
ACQ-6 score	0.63 ± 0.13	1.18 ± 0.14	1.00 ± 0.14
ACT or CACT score	22.1 ± 0.5	19.1 ± 0.8	20.2 ± 0.7
Uncontrolled asthma	12 (27.3)	21 (51.2)*	23 (46.0)*
PAQLQ score2	6.01 ± 0.14	5.83 ± 0.17	5.83 ± 0.20
Lung function2
FVC (% predicted)	101.9 ± 2.3	103.5 ± 2.8	101.9 ± 1.4
FEV1 (% predicted)	94.4 ± 2.2	92.7 ± 3.1	88.7 ± 2.3
FEV1/FVC (% predicted)	92.5 ± 1.8	89.3 ± 2.1	87.0 ± 1.6*
FEF25–75 (% predicted)	80.2 ± 4.8	73.4 ± 4.8	66.9 ± 3.9*
Exacerbation history, past 12 months3
Any systemic corticosteroid burst	20 (28.2)	23 (36.5)	35 (50.0)*
Any Emergency department visit	19 (26.8)	14 (22.2)	31 (44.3)*
Any hospitalization	3 (4.2)	3 (4.8)	12 (17.2)*

^*p<0.05 vs. low ECP

¹Low ECP, n=44; Intermediate ECP, n =41; High ECP, n=50

²Low ECP, n=45; Intermediate ECP, n =39; High ECP, n=47

³Low ECP, n=71; Intermediate ECP, n=63, High ECP, n=70

Figure 4.

Sensitivity and specificity of baseline (A) eosinophil cationic protein (ECP) concentrations, (B) blood eosinophil counts, and (C) exhaled nitric oxide concentrations for detection of exacerbation occurrence by 12 months.

Discussion

This study addressed whether plasma ECP concentrations could detect Type-2 inflammation and discriminate clinical outcomes in children with atopic asthma. In this cohort, higher ECP concentrations were associated with higher blood eosinophil counts, aeroallergen sensitization, IgE and exhaled nitric oxide concentrations and were further associated with poorer asthma control, greater airflow limitation, and more exacerbations both at baseline and at 12 months of follow-up. Children with higher ECP concentrations also had greater improvement in asthma control with the initiation of systemic corticosteroids. These observations suggest that plasma ECP concentrations may be a useful marker of Type-2 inflammation in children and may help identify those children at highest risk for recurrent exacerbations who could benefit from corticosteroid treatment. The fact that ECP concentrations remained elevated after intramuscular triamcinolone, unlike blood eosinophils or exhaled nitric oxide, also suggests that ECP may be less subject to treatment confounding. However, we also show that, similar to blood eosinophil counts or exhaled nitric oxide, the sensitivity of ECP for predicting exacerbation is relatively poor. Additional biomarkers of Type-2 inflammation, or combinations of existing biomarkers, may ultimately be needed in routine clinical practice.

The results of our study are similar to others that have shown associations between ECP and features of Type-2 inflammation, namely blood eosinophils,²⁹ allergen sensitization^{30, 31} and exhaled nitric oxide,²⁹ in children with asthma. Surprisingly, we did not observe correlations between ECP and the Type-2 cytokines IL-4, IL-5, IL-13 or CCL-11/eotaxin, but these cytokines have not been consistently detected in children with asthma, including those with severe refractory asthma who also have airway tissue eosinophilia.^{32, 33} There is also no gold standard for Type-2 inflammation for the purpose of comparison. The original studies of T2-”high” asthma relied on epithelial gene expression signatures to define immune phenotypes,^{34, 35} but bronchoscopy is not clinically indicated in most patients with asthma and these epithelial-based definitions are therefore not practical in pediatric outpatient settings. While studies of exhaled nitric oxide or blood eosinophils have shown promise, cut-points for these markers also remains debatable³⁶ since these markers may have little concordance in the same patient^{37, 38} and may be subject to significant variability over time and in response to asthma treatments.³⁹

Nonetheless, our study results do have biologic plausibility. Multiple in vitro studies have shown that high levels of ECP are cytoxic and promote fibroblast migration and the interaction between mast cells and eosinophils in some tissue.¹² While the functional impact of physiologic concentrations of ECP in humans is still unclear, others have similarly shown that children with higher ECP concentrations have poorer symptom control⁴⁰ and more airflow obstruction^{29, 41} with greater bronchodilator responsiveness⁴² in small, cross-sectional studies. The present study advances these observations to a large cohort of racially-diverse children who were followed longitudinally for 12 months and demonstrates that the clinical features of Type-2 “high” asthma that are discriminated by high plasma ECP concentrations are relatively stable over time. The present study is also the first to demonstrate that ECP concentrations can predict future exacerbations and systemic corticosteroid treatment responsiveness. The fact that ECP concentrations were highly correlated with total IgE levels is also of interest, since previous studies have shown that blood eosinophils may predict the clinical responses to anti-IgE treatment initiation.^43–45 Whether ECP would be similarly useful in prediction of anti-IgE therapeutic responses or responses to other biologics directed at Type-2 inflammation is unclear and warrants further study.

Strengths of the present study are the large and highly characterized sample which permitted adjustment for key covariates (sex, race, household educational attainment and tobacco smoke exposure) which are known to be associated with asthma outcomes.⁴⁶ The large proportion of enrolled children with an asthma exacerbation treated with systemic corticosteroids in the previous year also ensured a study population with sufficient disease burden at greatest risk for poorer outcomes. We also elected to measure ECP in the plasma, unlike the majority of other pediatric asthma studies,^{5, 14, 29–31, 40–42, 47–49} since the anticoagulant EDTA attenuates eosinophil ECP release and provides a better indication of the in situ ECP concentration.^{10, 12} Serum ECP levels are also highly dependent on sample handling and increase with rising sample temperatures during the clotting process.⁵⁰ Serum ECP levels are therefore considerably higher (i.e., approximately 2-fold) than plasma ECP levels.⁵⁰ Despite the limited number of studies of plasma ECP in children with asthma, one study of plasma ECP concentrations in hospitalized children with acute asthma observed a median ECP concentration of 2.8 ng/mL (with a range up to 15 ng/mL), which decreased after systemic corticosteroid administration.⁵¹ Two other studies of non-atopic healthy adults noted that mean plasma ECP concentrations were ~3.0 ng/mL,^{50, 52} which is similar to the cut-point of 3.12 ng/mL that was used for the lowest ECP tertile in the present study.

However, this study does have limitations. First, this study was limited to children with atopic asthma. The majority of these children had allergic rhinitis and eczema, which also may influence ECP measurements. Therefore, it is unclear whether ECP would perform similarly as a predictor of corticosteroid responsiveness and exacerbations in more heterogeneous populations of children with diverse patterns of inflammation. We were also unable to measure ECP in the airways and it is possible that systemic ECP may not adequately reflect airway tissue eosinophilia, which can persist for 7–14 days after eosinophils exit the bloodstream.⁹ Although a previous study in adults noted correlations between systemic ECP and bronchoalveolar lavage and airway tissue eosinophils, the associations were of a moderate nature (bronchoalveolar lavage eosinophils, r = 0.457, R² = 0.209; tissue eosinophils, r = 0.556, R² = 0.309).⁵³ Airway biomarker measurement is also invasive and difficult to perform in general populations of children. It is also recognized that our selected outcomes of asthma control and exacerbations are multifactorial and could be influenced by unmeasured sources of confounding including adherence to asthma medications, which was not adequately assessed. It is therefore unclear whether the outcomes at 12 months or the triamcinolone response outcomes observed in the present study would have differed with supervised inhaled corticosteroid treatment. Indeed, other highly controlled clinical trials of preschool and school-age children have shown that systemic ECP measures predict symptom and lung function improvement with inhaled corticosteroid initiation.^{5, 14} However, it is also recognized that there are Type-2 “high” refractory asthma phenotypes in children that are poorly responsive to inhaled corticosteroids, which this study did not adequately assess.¹

Other important limitations of the study include the measurement of ECP and Type-2 biomarkers at a limited number of time points, which may not be optimal for assessment of Type-2 inflammation. For example, a previous study in adults found that “same day” measurements of Type-2 biomarkers (including blood eosinophils) identified only half of the eosinophilic group.³⁸ However, there is no gold standard measure of Type-2 inflammation, so it is unclear whether more frequent measurement of ECP would increase the sensitivity of the biomarker. Indeed, other studies have shown that the sensitivity of blood eosinophils is only marginally increased with multiple measures.⁵⁴ Additionally, although we collected blood samples within a 5-hour time window, eosinophils are subject to some diurnal variability.⁵⁵ It is therefore likely that ECP may have similar variation.

In summary, we demonstrate that plasma ECP concentrations may be a useful biomarker of Type-2 inflammation, corticosteroid treatment response and exacerbation outcomes in children with asthma. While independent replication and validation of our results is essential, these findings offer additional insight on the importance and measurement of eosinophil activation in children with asthma. Whether children with elevated plasma ECP concentrations also respond differently to other management strategies for Type-2 inflammation, such as biologics, is unclear and warrants further evaluation.

Highlights Box.

What is already known about this topic?

Blood eosinophils are a frequently utilized marker of Type-2 inflammation in children with asthma, but their sensitivity for predicting outcomes is relatively poor. Alternative markers of Type-2 inflammation may be beneficial in children.

What does this article add to our knowledge?

Plasma eosinophil cationic protein (ECP) concentrations were associated with Type-2 inflammatory markers, poorer asthma control and lung function, and historical exacerbations. Higher plasma ECP also predicted systemic corticosteroid responsiveness and future exacerbations at 12 months.

How does this study impact current management guidelines?

Plasma ECP may be a useful marker of Type-2 inflammation and may help identify those children at highest risk for recurrent exacerbations who could benefit from corticosteroid treatment.

Abbreviations

ACQ

Asthma Control Questionnaire

ACT

Asthma Control Test

CACT

Childhood Asthma Control Test

CCL

CC chemokine ligand

ECP

Eosinophil cationic protein

FEF_25–75

Forced expiratory flow at 25–75% of vital capacity

FEV₁

Forced expiratory volume in one second

FVC

Forced vital capacity

IgE

Immunoglobulin E

IL

Interleukin

PAQLQ

Pediatric Asthma Quality of Life Questionnaire

PC₂₀

Provocative concentration causing a 20% drop in FEV₁