Proteomic characterization of atopic dermatitis blood from infancy to adulthood

Abstract

Background:

Patients with atopic dermatitis (AD) have systemic biomarker dysregulation that differs by age group; however, the proteomic characteristics of these age-based changes are unknown.

Objective:

To profile blood proteins of patients with AD across different age groups versus age-appropriate controls.

Methods:

Using the Olink high-throughput proteomic platform, we profiled 375 serum proteins of 20 infants (age, 0–5 years), 39 children (age, 6–11 years), 21 adolescents (age, 12–17 years), and 20 adults (age, ≥18 years) with moderate-to-severe AD and 83 age-appropriate controls.

Results:

Each group presented a distinct systemic proteomic signature. Th2-related proteins were increased in infant AD and further intensified with age through adolescence and adulthood (interleukin 4/CCL13/CCL17). In contrast, Th1 axis down-regulation was detected in infants with AD and gradually reversed to increased Th1 products (interferon γ/CXCL9/CXCL10/CCL2) in patients with AD from childhood to adulthood. Despite their short disease duration, infants already had evidence of systemic inflammation, with significant upregulation of innate immunity (interleukin 17C/ interleukin-1RN), T-cell activation/migration (CCL19), Th2 (CCL13/CCL17), and Th17 (PI3) proteins. Adults with AD present unique upregulation of cardiovascular proteins related to coagulation and diabetes.

Limitations:

Cross-sectional observational study with a single time point.

Conclusion:

Systemic immune signatures of AD are age-specific beyond the shared Th2 immune activation. These data advocate for precision medicine approaches based on age-specific AD profiles.

INTRODUCTION

Atopic dermatitis (AD) is the most common inflammatory skin disease, affecting up to 20% of children worldwide.¹ Although approximately 60% of patients are diagnosed within the first year of life, chronic persistence of AD into adulthood is reported in up to 50% of school-aged patients.^2–6

AD is heterogeneous and changes with age, as was recently shown in the skin and lymphocytes (using flow cytometry) of patients with AD as they mature from infancy through adulthood. All AD age groups present Th2/Th22 skewing. Infants with AD also display strong Th17-related signaling, which decreases during childhood and adolescence, whereas Th1 activation develops during adult AD as a sign of chronicity.⁷ Flow cytometry blood analysis also established a Th2 axis–related upregulation in AD across all age groups, along with Th17/Th22 skewing from infancy and an increase in interferon γ (Th1) signaling in adults.⁸ Nevertheless, the evolution of circulating proteins across consecutive AD age groups is still unknown.

AD has been associated with several systemic comorbidities potentially because of the long-standing inflammation characterizing the disease.^9,10 However, data on the dysregulation of circulating proteins in AD are largely limited to adults with years of disease activity^11–13 and a few reports exploring early-onset infant AD, which show systemic signs of inflammation within months of disease onset.^9,13–15 Characterization of the proteomic changes associated with AD at different age groups may shed light on the chronic inflammatory process of AD using a minimally invasive approach. Proteomic analysis may also detect age-specific biomarkers for clinical trials, longitudinal studies, and, in the future, better-tailored clinical management. Moreover, as the treatment of moderate-to-severe pediatric AD is still challenging,^16,17 proteomic characterization of AD may ultimately guide age-specific therapeutic development.

Using 375 analytes of the Olink platform, we profiled circulating proteins in patients with AD in different age groups (infants/toddlers, aged 0–5 years; children, aged 5–12 years; adolescents, aged 12–18 years; and adults, aged ≥18 years) in comparison to blood proteins from age-appropriate controls. Our results portray the evolution of the systemic molecular characteristics of AD across ages, revealing unique AD proteomic signatures across different age groups.

METHODS

Patient characteristics

One-hundred patients with AD (20 infants [aged 0–5 years; mean, 2.39 years], 39 children [aged 6–11 years; mean, 8.24 years], 21 adolescents [aged 12–17 years; mean, 15.0 years], and 20 adults [aged ≥18 years; mean, 40.9 years]) with moderate-to-severe AD and 83 healthy age-appropriate controls were enrolled in the study (demographics in Table I). Before study inclusion, parents (and patients if aged ≥12 years) signed institutional review board–approved consent. Immunoglobulin E (IgE) and clinical indices were collected, including the SCORing of Atopic Dermatitis (SCORAD) severity score,¹⁸ as well as history of atopic comorbidities (Table I). The exclusion criteria included recent use (<4 weeks) of systemic immunosuppressive treatment before this study. Subjects enrolled in the control groups had a negative personal and family history of atopy and a negative personal history of cardiovascular/immune/neoplastic diseases, hypertension, and hyperlipidaemia.

Table I.

Demographic characteristics

	Infant AD (n = 20)	Infant N (n = 20)	P value	Child AD (n = 39)	Child N (n = 27)	P value	Adolescent AD (n = 21)	Adolescent N (n = 16)	P value	Adult AD (n = 20)	Adult N (n = 20)	P value
Age (y), mean (SD)	2.5 (1.8)	2.5 (1.3)	0.97	8.3 (2)	8.2 (2)	0.87	15 (2)	14.5 (1.6)	0.46	42.3 (18)	37.9 (15.6)	0.41
Female sex, n (%)	9 (45)	8 (40)	1	25 (64.1)	17 (63)	1	10 (47.6)	8 (50)	1	10 (50)	11 (55)	1
Total serum IgE (IU/mL), mean (SD)	1072.1 (1838.2)	26.69 (42.5)	0.18	2026.6 (2401)	45.99 (49.5)	1.00E-05	3841 (3860)	236 (533)	0.002	2710 (2334)	97 (92)	0.002
SCORAD mean (SD)	50.5 (18.1)	NA		54 (15.3)	NA		57.6 (22.1)	NA		58.4 (12.1)	NA
Ethnicity (%)
Caucasian	10 (50)	17 (85)	0.04	26 (66.7)	21 (77.8)	0.41	14 (66.7)	13 (81.2)	0.46	15 (75)	15 (75)	1
African American	4 (20)	1 (5)	0.34	4 (10.3)	4 (14.8)	0.71	2 (9.5)	1 (6.3)	1	2 (10)	2 (10)	1
Asian	6 (30)	2 (10)	0.23	6 (17.9)	2 (7.4)	0.45	3 (14.3)	0 (0)	0.24	3 (15)	3 (15)	1
Mixed/Unknown	0 (0)	0 (0)	NA	3 (5.1)	0 (0)	NA	2 (9.5)	2(12.5)	NA	0 (0)	0 (0)	NA
Atopic comorbidities *	10/20	None	NA	37/39	None	NA	19/21	None	NA	11/20	None	NA

AD, Atopic dermatitis; IgE, immunoglobulin E; NA, not applicable; SCORAD, SCORing of Atopic Dermatitis.

^*Atopic comorbidities included asthma, food allergy, allergic rhinitis, allergic conjunctivitis, and seasonal allergies.

Blood samples

Serum aliquots from subjects with AD and healthy controls were analyzed using a proteomic Olink Proseek multiplex assay. Specifically, we used a comprehensive panel detecting inflammation-related proteins, cardiovascular panels II and III, and the Olink neurology panel. This method quantified protein levels from 10-μL serum for quantification by a real-time polymerase chain reaction platform. IgE was assessed in the sera of all patients with AD.

Statistical analysis

Analyses were performed using the R language (R Core Team and the R Foundation for Statistical Computing; R-project.org). Fold changes (FCHs) for comparisons were estimated with an empirical Bayesian method, and hypothesis testing was conducted using contrasts under the general framework for linear models in the limma package. The results were normalized to healthy controls to differentiate pathologic changes from physiological maturation. A protein with an FCH of >1.2 and a P value of <.05 or a false discovery rate (FDR) of <0.05 was considered differentially expressed. |FCH| of >1.2 was chosen as an arbitrary threshold found appropriate for serum data in which skin-related biomarkers are harder to detect. The sample size was determined on the basis of a power calculation from a previous Olink study in an AD cohort, in which 20 or more AD samples and 9 or more controls had greater than 90% power (P = .05) to detect differences in key Th2 (CCL13 and CCL17) markers.¹⁹

The Benjamini-Hochberg procedure was used to adjust P values for multiple hypotheses by controlling the FDR (values in brackets and Supplementary Table I, available via Mendeley at https://data.mendeley.com/datasets/rvybyf8zzd). For correlation studies, unsupervised hierarchical clustering of variables or samples/patients was performed using the correlation coefficient as a distance metric with the average agglomeration algorithm. Pathway analysis was performed with XGR software using canonical/KEGG/Reactome/BioCarta pathways.²⁰

RESULTS

To evaluate circulating AD-related versus growth-related changes across pediatric and adult age groups, we profiled 375 serum proteins from 100 patients with AD and 83 healthy age-appropriate controls using the Olink high-throughput proteomic platform (see demographic characteristics in Table I).

Across the panels, we were able to identify up to 228 differentially expressed proteins (DEPs) using criteria of FCH > |1.2| and P < .05 or FDR < 0.05 as depicted in Supplementary Table I. Because the cohort across age groups was composed primarily of Caucasian patients, we also conducted a sensitivity analysis to confirm that similar results were obtained in subjects of all ethnicities (Supplementary Table II, available via Mendeley at https://data.mendeley.com/datasets/rvybyf8zzd).

Differentially expressed immune-related proteins change across age groups

To characterize the immune phenotype of the different AD age groups, we investigated a panel of immune DEPs that were previously reported as AD-related,^13,21–23 as depicted in an unsupervised hierarchical clustering heat map (Fig 1). A common AD signature was shared across all age groups and included key Th2 markers (ie, CCL17/TARC and CCL13; FDR < 0.05) along with the adhesion marker CDH3 (FDR < 0.05).²⁴

Fig 1.

Comparison of differentially expressed immune-related proteins in the blood of patients with atopic dermatitis of different age groups. Heatmap of immune signaling proteins that are differentially expressed in the blood of different AD age groups, using criteria of fold change (FCH) ≥ |1.2| and P < .05. Proteins are ordered by unsupervised hierarchical clustering. The columns of the table on the right show FCH by AD age groups vs age-matched controls, and FCH across normalized groups. +P < .10, ∗P < .05, ∗∗P < .01. AD, atopic dermatitis; Adol, adolescent; N, normal.

In addition to these shared DEPs, each AD age group presented specific trends. The infant AD already presented increased systemic activation compared with that in age-appropriate controls, with significant upregulation of some innate immunity (with interleukin (IL) 17C being uniquely upregulated only in infant AD), cellular activation/migration (CD8A, CD84, CCL19), and Th17 (KYNU, PI3) markers, all presenting the highest expression levels in infants or a unique upregulation in this age group (FDR < 0.05). Moreover, key Th1 markers, such as CXCL10, were uniquely downregulated in the infant AD group (FDR < 0.05; Fig 1 and Supplementary Table II).

In comparison to infant AD and age-appropriate controls, childhood AD presented an increase in innate immunity markers (ie, IL-8), Th2-related markers (CCL7 and transforming growth factor α), and key Th1/Iinterferon γ-associated proteins (CXCL10; FDR < 0.05 for all). On the other hand, children with AD presented a decrease in immune markers related to T-cell activation/recruitment (CD40LG and IL-16) when compared to other age groups (Fig 1 and Supplementary Table I).

Adolescent AD presented intensification of Th2 (CCL13 and CCL17) and Th1/interferon γ (CXCL10/11) markers in comparison to patients of younger age with AD.

Adults with AD displayed a further increase in Th17/Th22-related (S100A12 and IL-20) and Th2-related (CCL13, CCL17, OSM, IL-1RL1/IL-33R, IL-4, and IL-4R) (FDR < 0.05) markers versus younger AD age groups, whereas IL-4R and IL-4 were among the proteins showing upregulation versus controls exclusively in the adult AD age group. Adults also displayed an intensification of Th1 markers as compared with younger age groups (IL-1RN, IL-16, and CCL2; Fig 1). Conversely, LAT, essential for T-cell activation,²⁵ was significantly downregulated in adult AD versus controls as well as versus other AD age groups.

Cardiovascular/atherosclerosis- and neurology-related proteins differ by age

Our analysis also included hallmark proteins of systemic inflammation that may be associated with other conditions or comorbidities, specifically cardiovascular/atherosclerosis,^13,26 as depicted in Fig 2, and neurology-related processes²⁷ (Supplementary Table I).

Fig 2.

Comparison of differentially expressed cardiovascular/atherosclerosis-related proteins in the blood of patients with atopic dermatitis of different age groups. Heatmap of cardiovascular/atherosclerosis signaling proteins that are differentially expressed in the blood of different AD age groups, using criteria of fold change (FCH) ≥ |1.2| and P < .05. Proteins are ordered by unsupervised hierarchical clustering. The columns of the table on the right show FCH by AD age groups vs age-matched controls, and FCH across normalized groups. +P < .10, ∗ P < .05, ∗∗ P < .01. AD, atopic dermatitis; Adol, adolescent; N, normal.

Overall, adult patients with AD presented the greatest upregulation of cardiovascular/atherosclerosis-related proteins, as shown in the pink box in Fig 2. Among these, proteins associated with platelet aggregation and thrombus formation (eg, PLAUR, MPO, ADAMTS13, SERPINE1, SELE; FDR < 0.01; Fig 2 and Supplementary Table I).

The infant AD showed a unique and significant upregulation of other circulating proteins linked with vascular growth factors (green box, Fig 2). For example, vascular endothelial growth factor D/FIGF, ADM, tissue inhibitor of metalloproteinase/TIMP4, PLAU, and HB-EGF^13,15,19 were all uniquely upregulated in infants, along with proteins associated with increased susceptibility to obesity (ie, LPL and AGRP^28,29) (FDR < 0.05 for all). Of note, in comparison to the infant AD, children with AD already displayed upregulation of some atherosclerosis-related markers^30,31 (ie, CEACAM8, PTX3, OLR1, AZU1, and MMP9; FDR < 0.05).

Using the Olink neurology panel, we also detected significant upregulation of factors associated with neural development, which were upregulated only in infant AD compared with age-matched controls. These included glial cell line–derived neurotrophic factor receptor α/GFRA3, galectin/4LGALS4, dorsal inhibitory axon guidance protein/DRAXIN, and BCAN (an integral component of the neuro synapses;³² FDR < 0.05 for all; Supplementary Table I).

Proteins related to T-cell activation correlate with AD severity across age groups

The Pearson correlation of protein levels with AD severity scores (assessed by SCORAD), patient’s age, and total serum IgE yielded strong correlations in all AD age groups (Table II). Markers that were correlated with SCORAD across all age groups included the T-cell activation marker IL-16 and the Th17-related marker PI3 (r > 0.4; P < .01). Within specific age groups, infants displayed correlations of AD severity (SCORAD) with proteins implicated in wound recovery and response to infections (TNFRSF13B, ITGB2, SELPG, and FCGR2B)^33–36 along with a regulatory protein of keratinocyte proliferation (IGFBP7)³⁷ (r > 0.45, P < .05). MMP12 was correlated with SCORAD in children and adolescents, and children and adults presented the strongest correlations with LGALS9, related to T-regulatory response against Th2. Similar trends were observed for Eczema Area and Severity Index and blood biomarkers, with disease severity correlating with wound healing (PECAM1) in infants, markers of general inflammation in children and adolescents (FCGR2B and TNFRSF10A), and T-cell response markers in adults (IL2R and TNFRSF1B), as depicted in Table II.

Table II.

Spearman correlations of selected serum proteins with Eczema Area and Severity Index, SCORing of Atopic Dermatitis, immunoglobulin E, and age (|r|>0.4, P < .05)

SCORAD												Age
Infant			Child			Adolescent			Adult			All age groups
Marker	r	P value	Marker	r	P value	Marker	r	P value	Marker	r	P value	Marker	r	P value
IL1RL1	0.61	0.005	PI3	0.54	<.001	IL18	0.62	<.001	IL16	0.71	0.001	PRELP	0.7	<.001
OSCAR	0.58	0.008	IL16	0.47	<.001	XCL1	0.6	<.001	OSCAR	0.57	0.009	TIMP4	0.64	<.001
SELE	0.56	0.011	MMP12	0.47	<.001	MMP12	0.55	0.011	TNFRSF1B	0.56	0.011	LEP	0.61	<.001
TNFRSF13B	0.55	0.014	ADM	−0.41	0.01	TGM2	0.54	0.012	LGALS9	0.54	0.016	BMP6	0.59	<.001
ITGB2	0.53	0.018	FCGR2B	−0.42	0.007	CHIT1	0.54	0.013	CSTB	0.53	0.017	ADM	0.58	<.001
PI3	0.53	0.018	MB	−0.43	0.006	CPB1	0.5	0.021	PI3	0.51	0.023	AMBP	0.55	<.001
CCL15-CCL14	0.52	0.019	MEPE	−0.52	<.001	CTRC	0.48	0.031	TNFRSF10B	0.51	0.024	MMP3	0.55	<.001
FABP4	0.52	0.021				PI3	0.45	0.044	ADM	0.49	0.031	PLAT	0.52	<.001
IL2RA	0.51	0.022				TNFRSF10A	0.44	0.048	IL2RA	0.48	0.032	CCL17	0.52	<.001
CA5A	0.48	0.034				MEPE	−0.45	0.044	GDF15	0.47	0.036	PAPPA	0.49	<.001
IGFBP7	0.47	0.038				CNTN1	−0.45	0.042	TNFRSF1A	0.46	0.041	HBEGF	0.48	<.001
F3	0.46	0.041				MB	−0.46	0.036	SPON1	0.46	0.044	CPA1	0.48	<.001
IL1RL2	0.46	0.044				SFTPD	−0.56	0.01				PDGFB	0.47	<.001
ACP5	0.46	0.045										CPB1	0.47	<.001
IL1R2	0.45	0.047										NPPB	0.46	<.001
SELPLG	0.45	0.047										PI3	0.46	<.001
EASI												IL16	0.43	<.001
Infant			Child			Adolescent			Adult			FGF21	0.42	<.001
Marker	r	P value	Marker	r	P value	Marker	r	P value	Marker	r	P value	IL27	0.41	<.001
IL1RL2	0.56	0.012	LGALS9	0.49	<.001	IL18	0.68	<.001	PI3	0.62	0.037	BLMH	−0.42	<.001
FCGR2B	0.5	0.028	AZU1	0.46	<.001	PI3	0.55	0.009	REN	0.62	0.037	GP6	−0.43	<.001
FABP6	−0.47	0.041	IL6	0.43	0.006	TNFRSF10A	0.53	0.013	TNFRSF1B	0.6	0.043	SELP	−0.44	<.001
DECR1	−0.48	0.04	MPO	0.42	0.007	CHIT1	0.48	0.026	IL2RA	0.59	0.046	IGFBP2	−0.44	<.001
PECAM1	−0.48	0.038	MMP9	0.42	0.008	XCL1	0.47	0.03	ACP5	0.59	0.049	LTBR	−0.44	<.001
XCL1	−0.50	0.03	PRTN3	0.41	0.01	SFTPD	−0.45	0.038	CCL3	−0.78	0.005	TNFRSF14	−0.45	<.001
			MB	−0.58	<.001	CNTN1	−0.59	<.001				IL17RA	−0.45	<.001
			FCGR2B	−0.66	<.001							BOC	−0.46	<.001
												PARP1	−0.49	<.001
IgE
Infant			Child			Adolescent			Adult			All age groups
Marker	r	P value	Marker	r	P value	Marker	r	P value	Marker	r	P value	Marker	r	P value
MMP9	0.89	0.012	IL2RA	0.42	0.008	TGM2	0.52	0.033	CCL3	0.78	0.039	CCL17	0.52	<.001
COL1A1	−0.79	0.048	LGALS3	0.41	0.012	SLAMF7	0.51	0.036	CXCL16	−0.78	0.039
CA5A	−0.79	0.048	MMP7	−0.48	<.001				PIGR	−0.85	0.015
TFPI	−0.82	0.034							VWF	−0.85	0.015
CXCL1	−0.82	0.034							TNFRSF10C	−0.89	0.007
HAO1	−0.86	0.024

EASI, Eczema Area and Severity Index; IgE, immunoglobulin E; SCORAD, SCORing of Atopic Dermatitis.S

Multiple correlations were found between age and cardiovascular/hypertension (eg, LEP, ADM, HB-EGF, PDGFB, NPPB). The AD-related biomarker, CCL17/TARC, was also positively correlated with age, as were PI3 and IL-16 (r > 0.43, P < .001). Although all age groups showed correlations between various markers and serum IgE levels, only CCL17/TARC was correlated with IgE across all age groups.

Pathway analysis further distinguishes age-specific proteomic profiles

For an unbiased, broad evaluation of age-specific AD-related changes in proteomic expression, all DEPs in AD versus controls by age groups were analyzed using function-based pathway databases (canonical/KEGG/Reactome/BioCarta pathways)²⁰, as depicted in Fig 3. Top enriched pathways across age groups largely overlapped and included overexpression of cytokine- and inflammation-related signatures along with an ensemble of extracellular matrix-associated proteins (FDR < 0.05). Only adolescents and adults presented enrichment of pathways related to Th1 and Th2 differentiation. JAK-STAT (Janus kinase–signal transducer and activator of transcription) pathway enrichment was detected starting from childhood. The infant AD uniquely presented enrichment of pathways related to TRAIL signaling, β1/2/3 integrin cell surface interactions, and immunoregulatory interactions between lymphoid and nonlymphoid cells. Adult AD uniquely presented significant enrichment of pathways related to complement, coagulation, and diabetes.

Fig 3.

Pathway analysis. Common/most enriched pathways in atopic dermatitis age groups versus age-appropriate controls using differentially expressed proteins in atopic dermatitis blood versus healthy control serum by age groups analyzed with publicly available function-based pathway databases (using canonical/KEGG/Reactome/BioCarta pathways), performed with XGR software. The dashed line denotes the significance threshold (FDR < 0.05).

DISCUSSION

AD often manifests during infancy and persists in adulthood in many patients.³⁸ Recent investigations of the molecular characteristics of AD in adults identified novel therapeutic targets, which transformed the clinical management of AD.^22,39,40 Ideally, a better understanding of disease mechanisms will guide age-specific therapeutic development to improve clinical management across all ages, including pediatric populations. Proteomic studies in blood using the Olink proteomic platform, which requires only 10 μL of blood, have been instrumental in identifying biomarkers linked to disease severity and response to various treatments.^41–46 We have previously described the blood proteomic profile of early-onset AD in infants.¹⁵ However, proteomic studies in the serum of children and adolescents with AD (5–18 years of age) are unavailable. Thus, we performed an analysis of serum from consecutive pediatric AD age groups (infants/toddlers aged 0–5 years, children aged 5–12 years, adolescents aged 12–17 years, and adults with AD aged ≥18years) relative to age-matched healthy controls. This approach allowed us to portray differences in proteomic expressions across age groups and to highlight specific AD signatures throughout life.

A robust Th2 (CCL13 and CCL17) upregulation was detected in all ages and intensified with increasing age. We also found differences in systemic immune polarization across age groups: Th17-related proteins showed the highest increases in infants (KYNU and PI3), while Th1-skewing was most prominent in adults (IL-1RN, CXCL16, and CXCL10), consistent with the fact that overexpression of Th1-related markers is associated with disease chronicity in long-standing AD.⁴⁷ Correlations between circulating proteins and age revealed that the expression of Th17-related markers significantly decreased with age, potentially reflecting the great burden of infections in AD during early life.^48,49

We also evaluated age-related changes that are associated with various biological processes and found a gradual increase in cardiovascular-related proteins. Of note, although adolescents with AD already show upregulation of many cardiovascular-related proteins, such as thrombomodulin/THBD, serine protease 27/PRSS27, CX3CL1, and FABP4, a pathway enrichment analysis shows significant upregulation in cardiovascular/atherosclerosis-associated pathways only in adults. These data suggest an increase in the systemic burden of AD with age and may support the need for early systemic treatment in patients with AD with moderate-to-severe disease. Tracking the cardiovascular burden should be a focus of future longitudinal studies in adolescents and younger children.

Using the Olink neurology/neuroinflammation panel, we also explored the potential changes in proteins associated with neural growth and development because previous studies linked altered neurotrophic factors with AD.⁵⁰ For example, GFRA3 mediates hypersensitivity to warm sensation, mimicking warmth-provoked pruritus in AD,⁵¹ and key AD markers, such as IL-4 and IL-31, can induce the growth of sensory nerves in vitro.^52–56 We found that only infants display enrichment of neural development pathways and associated proteins, such as GFRA3 and CADM3, along with unique upregulation of NCAN. Some of these markers have been linked to attention-deficit/hyperactivity disorder and autism,^57–60 and further investigation is needed to establish their potential role in the higher risk of developing attention-deficit/hyperactivity disorder in younger populations with AD.^61–64 Overall, these findings may be relevant to early disease initiation in AD, particularly in light of recent studies showing a link between sensory neurons, inflammatory pathways, and itch.^65–67

This study has a few limitations. Although all age groups from infants to adults were represented in our large cohort, this is not a longitudinal study, and thus, our results reflect a cross-sectional observation at a single time point. In addition, our cohort included a majority of Caucasian participants, potentially limiting the generalizability of our results to other ethnicities.

In summary, our results portray a continuum of the inflammatory blood proteome of AD across ages, filling an important gap in understanding the systemic alterations in AD from infancy to adulthood. Our proteomic data show consistent and progressive Th2 activation across all age groups, as also suggested by the efficacy of Th2-targeting agents, such as dupilumab, in all ages.^{16,22,68–72} However, because less than 50% of patients achieve disease clearance with Th2-targeting alone across all age groups,^73–75 these results potentially suggest the need for additional, age-specific, therapeutic approaches that will improve efficacy and may prevent disease chronicity and the development of associated comorbidities.

CAPSULE SUMMARY.

• Patients with atopic dermatitis have systemic biomarker dysregulation that differs by age; however, the evolution of circulating proteins across consecutive atopic dermatitis age groups is still unknown.

• The results from this study advocate for age-specific therapeutic approaches that can improve clinical outcomes in patients with atopic dermatitis of different ages.

Abbreviations used

AD

atopic dermatitis

DEP

differentially expressed protein

FCH

fold change

FDR

false discovery rate

IgE

immunoglobulin E

IL

interleukin

SCORAD

SCORing of Atopic Dermatitis