Abstract
Background
Non-lesional skin in atopic dermatitis (AD) is abnormal, but the pathobiology of lesional and non-lesional skin and the definition of endotypes are poorly understood.
Objective
To define lesional and non-lesional endotypes of AD by building the first US-based early life prospective cohort of children with AD, the Mechanisms of Progression from AD to Asthma in Children (MPAACH) cohort.
Methods
We assessed lesional and non-lesional skin TEWL, filaggrin (FLG) and alarmin (S100A8, S100A9) expression, staphylococcal colonization, and patterns of aeroallergen and food sensitization to define non-lesional and lesional phenotypes and endotypes.
Results
Pathophysiologic changes were present in lesional and non-lesional skin and were associated with SCORAD. Non-lesional skin had features characteristic of diseased skin including low FLG and high alarmin expression, and increased colonization with S. aureus. In a multivariate model, non-lesional, but not lesional, FLG expression was associated with the development of co-sensitization and moderate-severe AD. Lesional skin was characterized by further deficits in FLG expression (p<0.001), but alarmin expression was the same as observed in non-lesional skin.
Conclusions
This study reveals that events in the non-lesional, not the lesional skin, promote the subsequent development of AD severity and co-sensitization, which is a key risk factor for allergic co-morbidities. Collectively, these data suggest the presence of a subclinical eczema endotype that may predispose to the development of allergic disease in the absence of overt eczema. This may represent a new definition of the atopic march that starts with skin barrier dysfunction rather than eczema.
Keywords: atopic dermatitis, atopic march, children, allergic disease progression, cohort study, filaggrin expression, alarmin expression, S. aureus colonization, co-sensitization
INTRODUCTION
Atopic dermatitis (AD) is a common chronic, inflammatory skin disorder with a complex etiology and heterogeneous presentation affecting up to 20% of children worldwide(1). AD has been highlighted as the first step in the “atopic march”, whereby AD typically predates the development of other allergic disorders. It has been estimated that one-third to half of AD patients will develop asthma(2), although the mechanisms that promote disease progression remain unclear. Atopic sensitization and food allergy (FA) have been reported to be major determinants of AD progression to respiratory allergy; however, recent data indicate that the mechanisms are far more complex(3). To evaluate the current status of knowledge in this area, the National Institute of Allergy and Infectious Diseases recently convened a workshop titled “Atopic dermatitis and the atopic march: mechanisms and interventions”(1). They concluded that only about 3% of children follow the complete course of what has been conventionally referred to as the atopic march, although early-life AD remains a major risk factor for the development of any atopic disease. They recommended that a new, large, prospective cohort study incorporating evaluations of skin, gut, airway and peripheral blood and the use of multiparametric approaches is required to better define phenotypic/endotypic subgroups of AD and to predict AD outcomes and development of allergic comorbidities.
The Mechanisms of Progression of Atopic Dermatitis to Asthma in CHildren (MPAACH) cohort was designed to meet this need. MPAACH is the first US-based early life cohort of children with AD. The goals of the cohort are to define AD phenotypes and endotypes, dissect the mechanisms that contribute to the progression of AD to other allergic disorders, and to identify novel biomarkers that identify children at high risk for the development of asthma and wheezing phenotypes. There have been advances in AD research to identify AD biomarkers using tape strips, a minimally invasive method that samples the keratinocytes. However, most of these studies have been in adults with chronic AD(4–14), children with moderate to severe AD(15, 16) or older children(17, 18). MPAACH represents mild, moderate and severe AD which allows us to evaluate factors that contribute to the whole spectrum of AD severity over time.
MPAACH is uniquely designed to delineate immunologic, skin, biome, physiologic, genetic, genomic, epigenetic, and environmental factors that promote the development of allergic comorbidities in over 550 infants and toddlers with AD. To enable mechanistic studies, extensive biospecimens are collected from both lesional and non-lesional skin. Unlike previous tape strip studies, our novel skin tape sampling methodology simultaneously samples the surface skin biome as well as the underlying human DNA and RNA, which can be used to characterize host responses to environmental cutaneous exposures. Herein, we focus on lesional and non-lesional skin TEWL, host keratinocyte filaggrin (FLG) expression, skin staphylococcal colonization, and host keratinocyte S100A8 and S100A9 (alarmin) expression, as well as patterns of aeroallergen and food sensitization to define AD phenotypes and endotypes. While a phenotype is any observable characteristic or trait of a disease without any implication of mechanism, an endotype is a subtype of a disease defined by distinct functional or pathobiological mechanisms. The purpose of this paper was to identify phenotypes and endotypes of moderate-severe AD in the first 400 MPAACH children.
METHODS
Study Design and Inclusion/Exclusion Criteria
MPAACH is a prospective early life cohort of children with AD who will be followed for 5 years including annual visits (Fig1) with extensive biospecimen collection (Table 1).
Fig 1.
Overview of MPAACH study design, visit structure and biospecimens collected.
Table 1.
Samples and specimens collected at each annual study visit from MPAACH participants.
| Source/Assessment | Samples Collected | Planned/Completed Assays |
|---|---|---|
| Saliva | DNA | Genotyping |
| Skin Assessment | SCORAD | |
| Lesional TEWL | ||
| Non-lesional TEWL | ||
| Blood | PBMCs | Flow cytometry, CyTOF |
| Serum | vitamin D, cotinine, Total IgE, specific IgE, cytokines and chemokines | |
| Plasma | ||
| Whole blood | DNA and RNA | |
| Sensitization Assessment | 11 aeroallergens | |
| 13 foods | ||
| Keratinocytes | Lesional tape strip DNA | DNA methylation |
| Non-lesional taps strip DNA | DNA methylation | |
| Lesional tape strip RNA | Gene expression | |
| Non-lesional tape strip RNA | Gene expression | |
| Nasal Samples | Biome swab | Microbiome analysis |
| Nasal brushing DNA* | DNA methylation | |
| Nasal brushing RNA* | Gene expression | |
| Gut | Stool | Microbiome analysis |
| Skin biome | Lesional contact plate – colonies | S. aureus and S. epidermidis presence and biofilm propensity (crystal violet assay) |
| Non-lesional contact plate – colonies | S. aureus and S. epidermidis presence and biofilm propensity (crystal violet assay) | |
| Lesional tape strip – bacterial DNA | Microbiome analysis | |
| Non-lesional tape strip – bacterial DNA | Microbiome analysis | |
| Home Dust | Bedroom | Allergen levels, microbiome |
| Primary activity room | Allergen levels, microbiome |
Sample collected at age 6–7 years study visit only.
Inclusion criteria were: 1) aged 1–2 years upon enrollment; 2) gestation of ≥36 weeks AND 3) a diagnosis of AD (based on the Hanifin and Rajka Criteria for Atopic Dermatitis(19)), OR the parent(s)/legal authorized representative (LAR) indicates a positive response to each of the 3 questions from the Children’s Eczema Questionnaire (CEQ) (20). Exclusions include: 1) a co-morbid lung condition including cystic fibrosis, congenital anomaly, or bronchopulmonary dysplasia; 2) dependence on immunosuppression or oral steroids for a medical condition other than asthma; 3) condition that precludes sampling of the proposed biologic samples or completion of spirometry; 4) a bleeding diathesis.
For the analyses presented herein, data was available for the first visit on the first 400 participants recruited into MPAACH from November 11, 2016 through September 5, 2018. Details on the recruitment strategy and outcomes, questionnaires administered, biologic specimens collected, planned assays and co-morbid condition and exposure definitions can be found in the Online Repository Text and Fig E1.
Data Analysis
Descriptive statistics were used to characterize the study population. Frequency (percentage) was reported for categorical variables while median (IQR) was reported for continuous variables because some measures exhibited non-normal distributions. Continuous variables (age, serum cotinine, IgE and vitamin D, TEWL, SCORAD and FLG expression) were assessed for normality through the Shapiro-Wilk test. The maximum value across the two tapes assayed for RNA expression were included in the final analysis for each gene. For ease of interpretation and display of the data, the gene expression levels were all multiplied by 1000. Details of the association testing can be found in the Online Repository Text.
RESULTS
Characteristics of MPAACH participants: Demographics, Exposures, and Comorbidities
Among the first 400 MPAACH participants, 61.5% identified as African-American, 65.0% had public insurance and 51.3% were male (Table 2). The median age was 2.3 years. Over a third (34.5%) owned a dog and 11.8% owned a cat in the child’s first year of life. Approximately half (49.5%) of the parent(s) reported exposures to secondhand smoke (SHS) and 8.5% reported exposure to vaping or e-cigarettes. The median serum cotinine was 0.37 ng/mL and was significantly higher in children whose parents reported SHS exposure (0.46 ± 4.22 versus 0.20 ± 1.51, p=0.006). The median ECAT level was 0.37 μg/m3. Parents reported a doctor’s diagnosis of FA in 19.6% of the cohort. This may be underestimated since many of the children are under 2 years of age and have not yet been exposed to common food allergens such as egg and peanut. One-third (32.6%) reported at least one wheezing episode in the last 12 months, 24% reported recurrent wheeze, 17.1% reported wheeze without a cold and 17.3% had allergic rhinitis.
Table 2.
Characteristics of the MPAACH Cohort
| N = 400 | |
|---|---|
| Demographics | |
| African-American Race | 246 (61.5%) |
| Public Insurance | 260 (65.0%) |
| Male Sex | 205 (51.3%) |
| Age (years), median (IQR) | 2.3 (1.7 – 2.5) |
| Birth Season | |
| Winter | 109 (27.3%) |
| Spring | 125 (31.3%) |
| Summer | 87 (21.8%) |
| Fall | 79 (19.8%) |
| Visit 1 Season | |
| Winter | 81 (20.3%) |
| Spring | 103 (25.8%) |
| Summer | 135 (33.8%) |
| Fall | 81 (20.3%) |
| Environmental Exposures | |
| Owned a dog in the 1st year of life | 138 (34.5%) |
| Currently live with a dog | 102 (25.5%) |
| Owned a cat in the 1st year of life | 47 (11.8%) |
| Currently live with a cat | 40 (10.0%) |
| Parental reported SHS exposure (cigarettes/cigars) | 198 (49.5%) |
| Parental reported exposure to Vaping/E-cigarettes | 34 (8.5%) |
| Serum cotinine (ng/mL), median (IQR)1 | 0.37 (0.00 – 0.99) |
| ECAT (μg/m3), median, (IQR) | 0.37 (0.30 – 0.49) |
| AD Began Age 0–3 Months | 160 (40.2%) |
| AD Severity | |
| SCORAD (score) | 18.7 (11.4 – 29.8) |
| SCORAD (severity group) | |
| Mild (SCORAD <25) | 263 (65.9%) |
| Moderate (SCORAD ≥25) | 118 (29.6%) |
| Severe (SCORAD ≥50) | 18 (4.5%) |
| Skin Barrier Function (Median, IQR) | |
| Lesional Skin | |
| TEWL (g/m2/h) | 11.7 (8.2 – 19.8) |
| Keratinocyte FLG expression2 | 0.92 (0.21 – 2.69) |
| Keratinocyte S100A8 expression2 | 0.15 (0.04 – 0.36) |
| Keratinocyte S100A9 expression2 | 0.25 (0.08 – 0.64) |
| Non-lesional Skin | |
| TEWL (g/m2/h) | 9.4 (7.1 – 13.6) |
| Keratinocyte FLG expression2 | 1.61 (0.51 – 4.06) |
| Keratinocyte S100A8 expression2 | 0.13 (0.04 – 0.40) |
| Keratinocyte S100A9 expression2 | 0.23 (0.05 – 0.71) |
| Skin Bacterial Presence | |
| S. aureus | 106 (26.6%) |
| S. epidermidis | 285 (71.4%) |
| Nutritional Status | |
| Serum Vitamin D median (IQR)1 | 30.0 (24.2 – 34.5) |
| Sensitization | |
| Total Serum IgE, median (IQR)4 | 37 (14 – 113) |
| SPT to Aeroallergens | |
| Negative | 262 (65.8%) |
| Positive to 1 aeroallergen | 66 (16.6%) |
| Positive to 2+ aeroallergens | 70 (17.6%) |
| SPT to Food Allergens | |
| Negative | 279 (70.1%) |
| Positive to 1 food allergen | 46 (11.6%) |
| Positive to 2+ food allergens | 73 (18.3%) |
| SPT Profile | |
| SPT Negative | 214 (53.8%) |
| Monosensitized | 63 (15.9%) |
| Aeroallergen only | 21 (33.3%) |
| Food allergen only | 42 (66.7%) |
| Polysensitized5 | 49 (12.3%) |
| Aeroallergens only | 23 (47.0%) |
| Food allergens only | 26 (53.1%) |
| Co-sensitized6 | 71 (17.8%) |
| Co-morbid Conditions | |
| Food allergy | 78 (19.6%) |
| Wheezing in the last 12 months | 130 (32.5%) |
| Recurrent wheeze | 96 (24.0%) |
| Wheeze without a cold | 68 (17.1%) |
| Allergic rhinitis | 69 (17.3%) |
n=293.
n=354, normalized to 18S and multiplied by 1000 for ease of interpretation.
n=281.
N=284.
defined as SPT+ to ≥2 aeroallergens OR ≥2 food allergens BUT not sensitized to food and aeroallergens.
defined as SPT+ to ≥1 aeroallergen AND ≥1 food allergen.
Sensitization Patterns and Total IgE
Overall, 46.2% of the cohort was sensitized to at least one allergen. Almost 16% of the cohort was monosensitized (sensitized to only 1 allergen), 12.3% of the cohort was polysensitized (sensitized to ≥2 aeroallergens (33.3%) or ≥2 food allergens (66.7%) but NOT both), and 17.8% were co-sensitized (sensitized to ≥1 aeroallergen AND ≥1 food allergen, Fig 2A). The most prevalent aeroallergen sensitizations were dog (13.3%) and trees (11.6%), while the most prevalent food allergens were egg (22.1%) and peanut (18.1%, Fig 2B). Among those who were food negative, only 23.3% were sensitized to an aeroallergen. In contrast, 60.2% of the food sensitized children were co-sensitized to aeroallergen(s) (p<0.001). The median total serum IgE was 37 IU/mL (Table 2). Participants who were SPT+ had higher total IgE levels (55 IU/mL) compared SPT negative children (26 IU/mL, p<0.0001), and total serum IgE increased significantly with increasing number of positive SPTs (r=0.33, p<0.0001).
Fig 2.
(A) Percentage of participants in MPAACH in each sensitization category. (B) Prevalence of sensitization to each individual aeroallergen and food allergen.
Factors related to Clinical Severity of AD
The median SCORAD of the cohort was 18.7. Over one-third of MPAACH children (34.1%) had moderate to severe AD, defined as SCORAD ≥25 (Table 2). Female participants had lower SCORAD scores (p=0.003).
Children who were co-sensitized to at least one aeroallergen and at least one food allergen had elevated SCORAD scores (median 28.3 g/m2/h) compared children who were mono or polysensitized (18.5 g/m2/h) and those that were SPT negative (16.6 g/m2/h, p<0.0001). Compared to polysensitized children, co-sensitized children had double the median total number of sensitizations (4 IQR 3–6 versus 2 IQR 2–3, p<0.0001) and a higher median number of aeroallergen (2 IQR 1–3 versus 0 IQR 0–2, p<0.0001) and food sensitizations (2 IQR 1–2 versus 2 IQR 0–2, p=0.0003). To determine if the increased SCORAD observed in the co-sensitized children was due to increased sensitization load, we performed multiple linear regression adjusting for the number of sensitizations. Even after adjustment for sensitization load, co-sensitized children had significantly increased SCORAD (p=0.026) compared to polysensitized children, supporting that co-sensitized children have increased AD severity independent of sensitization load.
Skin colonization with S. aureus was significantly associated with AD severity, evident by an increase in SCORAD (p=0.047 Fig E2A), but not with lesional or non-lesional TEWL (Fig E2(C) and 3E). Colonization of the skin with S. epidermidis was not associated with AD severity or TEWL (Fig E2(B, D, F)).
Fig 3.
Associations of demographics, comorbidities, sensitization and AD severity with factors in (A) non-lesional and (B) lesional skin in the MPAACH participants. Numbers shown are Spearman correlation coefficients if p<0.05.
Children with higher SCORAD scores had increased serum cotinine levels (r2=0.18, p=0.003, Fig 3A), even though there was no association between SCORAD and parental reported SHS exposure (p=0.87). Thus, children with similar levels of parental-reported environmental exposure to SHS had higher biologic levels of serum cotinine if their AD was more severe, suggesting that nicotine is absorbed through the skin at significant levels, at least in children with AD. Children with allergic rhinitis also had significantly higher SCORAD scores (21.8 vs. 18.4, p=0.02). There was no association between SCORAD and any wheezing in the last 12 months or ECAT.
Molecular Endotypes of AD
We next examined the pathophysiology of non-lesional and lesional skin and determined their relationships to severe AD in order to define endotypes. Median TEWL was higher for lesional skin (11.7 g/m2/h) compared to non-lesional skin (9.4 g/m2/h, p<0.0001, Table 1). Expression of FLG was significantly decreased in lesional skin compared to non-lesional skin (0.92 vs. 1.61, p<0.0001). There were no differences in S100A8 or S100A9 expression overall, but S100A9 expression was nominally increased in the lesional skin of children with moderate-severe AD compared to (p=0.077).
Lesional Skin
We determined associations of lesional TEWL, FLG expression, alarmin expression, and skin staphylococcal colonization with sensitization patterns and SCORAD (Fig 3A). Higher SCORAD was associated with higher lesional TEWL (p<0.001), aeroallergen (p<0.0001), food allergen (p<0.0001) and co-sensitization (p<0.0001), increased alarmin expression (p<0.0001), and a marked decrease in lesional FLG expression (p<0.0001). Lesional TEWL and lesional FLG expression were also inversely associated with each other (p=0.01). SCORAD was associated with an increase in S100A8 (p<0.001) and S100A9 (p<0.001) expression and a decrease in FLG expression (p<0.001) in lesional skin. Increased TEWL and FLG expression were associated with increased S100A8 (p=0.02 and p=0.0003, respectively) and S100A9 expression levels (p=0.009 and p=0.03, respectively) in lesional skin.
In the lesional skin, both presence and load of food and aeroallergen sensitization were associated with decreased FLG expression and increased TEWL (all p<0.01). TEWL was significantly increased in children who were co-sensitized (p=0.0006). Even after adjustment for sensitization load, co-sensitized children had significantly increased lesional TEWL compared to polysensitized children. Children with FA had significantly higher (14.8 vs. 11.1, p=0.002) lesional TEWL and decreased lesional FLG expression (0.43 versus 1.07, p=0.001).
Colonization with S. aureus was associated with increased SCORAD (p=0.047, Fig 3A). and decreased FLG expression in lesional skin (0.55 vs. 1.09, p=0.001, Fig 4). Staphylococcal presence (S. aureus or S. epidermidis) was not associated with lesional S100A8 or S100A9 expression. A higher level of serum cotinine was associated with higher expression of S100A8 in lesional skin (r2=0.16, p=0.01) suggesting that cutaneous SHS exposures trigger skin host defense responses. There was a borderline significant increase in S100A9 expression as well (r2=0.12, p=0.053).
Fig 4.
Association of the presence of S. aureus colonization on the skin with non-lesional and lesional filaggrin gene expression.
Taken together, these results illustrate that a molecular endotype of severe AD in the lesional skin is characterized by high lesional TEWL, low lesional FLG expression, high lesional S100A8/A9 expression, S. aureus colonization, co-sensitization, and elevated serum cotinine.
Non-lesional Skin
As in the lesional skin, higher SCORAD scores were associated with a decrease in non-lesional FLG expression (p<0.0001, Fig 3B) and increased non-lesional TEWL (p=0.001) supporting that pathologic features characteristic of AD are present even in the non-lesional skin.
In contrast to lesional skin, presence and load of food sensitization, but not aeroallergen sensitization, was significantly associated with increased non-lesional TEWL (p=0.04). Similar to our findings in lesional skin, children who were co-sensitized had significantly decreased FLG expression in their non-lesional skin (p=0.0009). In addition, children with FA also had significantly higher non-lesional TEWL (11.8 vs. 9.2, p=0.01) and decreased non-lesional FLG expression (0.97 versus 1.77, p=0.002). Both SCORAD and non-lesional FLG expression were associated with increased S100A8 (p=0.01, p<0.0001) and S100A9 expression in non-lesional skin (p=0.02, p>0.0001). S. aureus colonization was associated with a significant decrease in FLG expression in non-lesional skin (1.05 vs. 1.94, p=0.004, Fig 4).
Collectively, these results illustrate that a molecular endotype of severe AD can be defined in non-lesional skin and it is characterized by high non-lesional TEWL, low non-lesional FLG expression, high non-lesional S100A8 expression, S. aureus colonization of non-lesional skin, polysensitization, food sensitization and FA. These results confirm that even children without active AD have a global epithelial defect that is not limited to active or typically active lesional locations.
We next utilized multivariate logistic modeling to determine what combination of factors in the non-lesional skin drive co-sensitization. Children with co-sensitization were more likely to have decreased FLG expression in their non-lesional skin (OR 0.78, 95%CI 0.63–0.92, p=0.007). In a similar model, a decrease in non-lesional FLG expression was associated presence of S. aureus on the skin (OR 0.86, 95%CI 0.76–0.96, p=0.005). These data demonstrate that the non-lesional (normal appearing) skin in children with AD has skin barrier dysfunction as evidenced by decreased FLG. Further, this visually normal skin may directly contribute to co-sensitization and colonization with more persistent S. aureus strains, and thus, promote the development of inflammation and clinically apparent AD.
Multivariate Model of Moderate-Severe AD
The presence of S. aureus, sex, lesional and non-lesional TEWL and lesional and nonlesional FLG, and S100A8 and S100A9 expression were considered for entry into the model. The final model revealed that moderate-severe AD is characterized by a significant decrease in non-lesional FLG expression (OR 0.89, 95%CI 0.81–0.97, p=0.004) and significant increases in lesional TEWL (OR 1.03, 95%CI 1.01–1.04, p=0.0002) and lesional S100A8 expression (OR 1.25, 95%CI 1.10–1.43, p=0.0003). The results support that expression level of FLG in nonlesional skin was significantly associated with co-sensitization and the development of moderate-severe AD, while expression of FLG in lesional skin fell out of the multivariate model, indicating that events in the non-lesional skin promote the subsequent development of lesions.
DISCUSSION
MPAACH is the first early life prospective cohort of children with AD in the US incorporating extensive evaluations of the skin, gut, airway and peripheral blood, as well as the use of multiparameter approaches to define phenotypic and endotypic subgroups of AD. The rigorous phenotyping and comprehensive biospecimen collections make MPAACH uniquely poised to identify endotypes that predict AD outcomes including the development of other atopic conditions (FA, asthma, and allergic rhinitis) and enable mechanistic studies. In the first 400 MPAACH participants, we have observed distinct endotypes of AD in lesional and non-lesional skin that provide novel insights into disease pathophysiology (Fig 5). A common pathologic feature of both lesional and non-lesional skin in AD was decreased FLG expression. This supports that in patients with known AD, barrier dysfunction is global beyond actively inflamed areas and that events in the non-lesional skin are integral to the development of atopic disorders. Further, in our multivariate model, the expression level of FLG in non-lesional skin was associated with co-sensitization and the development of moderate-severe AD, while expression of FLG in the lesional skin was not, indicating that events in the non-lesional skin promote the subsequent development of lesions. Collectively, our data provide novel insights into disease pathophysiology, suggesting that the events taking place in the non-lesional skin may contribute to the development of active lesions by promoting inflammation, a feed-forward subsequent decrease in FLG expression, and increased skin permeability.
Fig 5.
Hypothesized model of the factors that contribute to AD in non-lesional and lesional skin.
The decrease in FLG expression in non-lesional skin is likely driven by a combination of polygenetic and environmental factors. Strikingly, we found that non-lesional FLG expression was strongly associated with skin colonization by S. aureus and co-sensitization. Thus, while non-lesional skin may appear normal, it exhibits characteristics typical of diseased skin. Indeed, recent findings by Leung et al. demonstrated that FLG content in the non-lesional skin of children with AD and FA were substantially lower than nonatopic controls(18). Guttman-Yassky et al. also reported that FLG expression was decreased in the lesional and non-lesional skin of children <5 years of age compared to healthy controls(16). However, their study was restricted to children with moderate-severe eczema, which may explain why they did not see an association between SCORAD and lesional or non-lesional FLG expression(16). Our findings provide additional evidence that this decrease in FLG is not simply an altered barrier but leads to skin dysbiosis and host alarmin expression by keratinocytes and likely contributes to AD symptoms and severity, as well as disease progression, i.e. the development of allergic comorbidities. This non-lesional endotype predicted increased AD severity and co-sensitization, which are known risk factors for asthma and allergic rhinitis development(21, 22). Thus, it is possible that biomarkers in the non-lesional skin in early life could predict subsequent allergic disease development. Indeed, biomarkers in non-lesional skin were strongly predictive of co-sensitization, and co-sensitization is a strong risk factor for asthma. We will be able to determine this in the MPAACH cohort as the children age.
MPAACH children had a subclinical form of eczema in their non-lesional skin. This raises the possibility that infants and toddlers without a diagnosis of eczema may actually have subclinical eczema with normal appearing skin, but low skin FLG expression, skin dysbiosis, and increased skin alarmin expression. These physiologic changes may be sufficient to confer risk for the development of allergic disease without overt eczema. We previously identified two high-risk early life groups who were at high risk for asthma development, one with early eczema and one without early eczema(23). It is possible that some of the high-risk group without eczema actually had subclinical disease.
The sensitization rates we observed in MPAACH children were higher than those previously reported in the general population as well as in high risk birth cohorts(24, 25). Thus, the presence of eczema confers higher risk for sensitization than parental atopy. Our data clearly demonstrate that co-sensitization is associated with AD severity over and above polysensitization, and this was independent of the total number of sensitizations. Taken with our finding that children with higher SCORAD scores have barrier defects evidenced by high TEWL and low FLG expression, this suggests both food and aeroallergen sensitization occur through the skin. While it has been reported that children with AD may develop peanut and egg sensitizations through epidermal allergen exposure,(26, 27) our findings suggest not only cutaneous sensitization of food allergens, but also aeroallergens.
Our data demonstrate that children with higher SCORAD scores (r2=0.17, p=0.003) had significantly increased levels of serum cotinine compared to children with lower SCORAD scores, despite equal levels of SHS exposure reported by the parent(s). This indicates that nicotine from SHS is being absorbed through the skin and that this absorption is directly correlated to the skin barrier dysfunction. Recent evidence supports that nicotine can be absorbed directly from the air at rates comparable to or higher than those observed via inhalation(28, 29) Given our findings, it is likely that children with AD absorb nicotine and allergens in the air and on surfaces through the skin at higher rates, along with increased levels of many of the other toxic substances found in cigarette smoke and other exposures.
The MPAACH cohort has many strengths including its size (thus far we have recruited 557 participants), rigorous phenotyping of AD, comprehensive biospecimen collection, extensive biomarker determinations, and its longitudinal design. MPAACH encompasses a wide degree of variation in AD severity, evident by the wide range in SCORAD scores from 0.6 to 85.5 and is a diverse population with 61.5% African-Americans. The first 250 participants enrolled were primarily between the ages of 2–3 years to evaluate the development of asthma at age 6–7 within 5 years. However, we are now enrolling children as young as 2–6 months to evaluate the timing of barrier dysfunction, sensitization and colonization and how these factors contribute to the severity of AD and progression to allergic comorbidities.
In summary, MPAACH is the first early life prospective cohort of children with AD in the US. The results herein collectively suggest that events that occur in the non-lesional skin in AD contribute to the co-sensitization and the development of skin inflammation and AD lesions (Fig 5). These findings are of high importance because very few children follow the traditional atopic march, emphasizing the need to better define AD disease phenotypes and endotypes that allow prediction of the natural history of AD.
Supplementary Material
What is already known on this topic?
Non-lesional skin in atopic dermatitis (AD) is abnormal, but the pathobiology of lesional and non-lesional skin and the definition of endotypes are poorly understood.
What does this article add to our knowledge?
This study reveals that events in the non-lesional, not the lesional skin, promote the subsequent development of AD severity and co-sensitization, which is a key risk factor for allergic co-morbidities.
How does this study impact current management guidelines?
Our study suggests that management of pediatric AD should include treatment of both lesional and non-lesional skin, because subclinical inflammation in normal-appearing areas may predispose to allergic co-morbidities and more severe disease.
Acknowledgements
We thank the participating families and staff on the MPAACH study.
Funding Source: This work was supported by NIH grant U19AI070235 (GKH, JBM, LJM and ABH).
Abbreviations
- AD
atopic dermatitis
- FA
Food Allergy
- MPAACH
Mechanisms of Progression of Atopic Dermatitis to Asthma in Children
- TEWL
transepidermal water loss
- SNP
single nucleotide polymorphism
- FLG
filaggrin
- CEQ
Children’s Eczema Questionnaire
- LAR
legally authorized representative
- CCHMC
Cincinnati Children’s Hospital Medical Center
- EMR
electronic medical record
- NVQ
New Visit Questionnaire
- SHS
secondhand smoke
- DDFC
Demographic Data and Family Contact Form
- IDQOL
Infant Dermatitis Quality of Life Questionnaire
- ACT
Asthma Control Test
- PROMIS
Patient-Reported Outcomes Measurement Information System
- SCORAD
Scoring for Atopic Dermatitis
- CyTOF
Cytometry by Time of Flight
- SPT
skin prick test
- ERMI
environmental relative moldiness index
- OFC
oral food challenge
- ATS
American Thoracic Society
- TRAP
traffic related air pollution
- ECAT
elemental carbon attributable to traffic
- OR
odds ratio
- 95%CI
95% confidence interval
Footnotes
The authors have no conflicts of interest.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Davidson WF, Leung DYM, Beck LA, Berin CM, Boguniewicz M, Busse WW, et al. Report from the National Institute of Allergy and Infectious Diseases workshop on “Atopic dermatitis and the atopic march: Mechanisms and interventions”. J Allergy Clin Immunol. 2019;143(3):894–913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Amat F, Saint-Pierre P, Bourrat E, Nemni A, Couderc R, Boutmy-Deslandes E, et al. Early-Onset Atopic Dermatitis in Children: Which Are the Phenotypes at Risk of Asthma? Results from the ORCA Cohort. PLoS One. 2015;10(6):e0131369. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Pinart M, Benet M, Annesi-Maesano I, von Berg A, Berdel D, Carlsen KC, et al. Comorbidity of eczema, rhinitis, and asthma in IgE-sensitised and non-IgE-sensitised children in MeDALL: a population-based cohort study. Lancet Respir Med. 2014;2(2):131–40. [DOI] [PubMed] [Google Scholar]
- 4.Voegeli R, Rawlings AV, Breternitz M, Doppler S, Schreier T, Fluhr JW. Increased stratum corneum serine protease activity in acute eczematous atopic skin. Br J Dermatol. 2009;161(1):70–7. [DOI] [PubMed] [Google Scholar]
- 5.Kezic S, O’Regan GM, Yau N, Sandilands A, Chen H, Campbell LE, et al. Levels of filaggrin degradation products are influenced by both filaggrin genotype and atopic dermatitis severity. Allergy. 2011;66(7):934–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Clausen ML, Slotved HC, Krogfelt KA, Agner T. Tape Stripping Technique for Stratum Corneum Protein Analysis. Sci Rep. 2016;6:19918. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Koppes SA, Brans R, Ljubojevic Hadzavdic S, Frings-Dresen MH, Rustemeyer T, Kezic S. Stratum Corneum Tape Stripping: Monitoring of Inflammatory Mediators in Atopic Dermatitis Patients Using Topical Therapy. Int Arch Allergy Immunol. 2016;170(3):187–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Winget JM, Finlay D, Mills KJ, Huggins T, Bascom C, Isfort RJ, et al. Quantitative Proteomic Analysis of Stratum Corneum Dysfunction in Adult Chronic Atopic Dermatitis. J Invest Dermatol. 2016;136(8):1732–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Berdyshev E, Goleva E, Bronova I, Dyjack N, Rios C, Jung J, et al. Lipid abnormalities in atopic skin are driven by type 2 cytokines. JCI Insight. 2018;3(4). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Clausen ML, Slotved HC, Krogfelt KA, Agner T. Measurements of AMPs in stratum corneum of atopic dermatitis and healthy skin-tape stripping technique. Sci Rep. 2018;8(1):1666. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Tham EH, Dyjack N, Kim BE, Rios C, Seibold MA, Leung DYM, et al. Expression and function of the ectopic olfactory receptor OR10G7 in patients with atopic dermatitis. J Allergy Clin Immunol. 2019;143(5):1838–48 e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Leitch CS, Natafji E, Yu C, Abdul-Ghaffar S, Madarasingha N, Venables ZC, et al. Filaggrin-null mutations are associated with increased maturation markers on Langerhans cells. J Allergy Clin Immunol. 2016;138(2):482–90 e7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Morita E, Takahashi H, Niihara H, Dekio I, Sumikawa Y, Murakami Y, et al. Stratum corneum TARC level is a new indicator of lesional skin inflammation in atopic dermatitis. Allergy. 2010;65(9):1166–72. [DOI] [PubMed] [Google Scholar]
- 14.Reisdorph N, Armstrong M, Powell R, Quinn K, Legg K, Leung D, et al. Quantitation of peptides from non-invasive skin tapings using isotope dilution and tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci. 2018;1084:132–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.McAleer MA, Jakasa I, Raj N, O’Donnell CPF, Lane ME, Rawlings AV, et al. Early-life regional and temporal variation in filaggrin-derived natural moisturizing factor, filaggrin-processing enzyme activity, corneocyte phenotypes and plasmin activity: implications for atopic dermatitis. Br J Dermatol. 2018;179(2):431–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Guttman-Yassky E, Diaz A, Pavel AB, Fernandes M, Lefferdink R, Erickson T, et al. Use of Tape Strips to Detect Immune and Barrier Abnormalities in the Skin of Children With Early-Onset Atopic Dermatitis. JAMA Dermatol. 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hulshof L, Hack DP, Hasnoe QCJ, Dontje B, Jakasa I, Riethmuller C, et al. A minimally invasive tool to study immune response and skin barrier in children with atopic dermatitis. Br J Dermatol. 2019;180(3):621–30. [DOI] [PubMed] [Google Scholar]
- 18.Leung DYM, Calatroni A, Zaramela LS, LeBeau PK, Dyjack N, Brar K, et al. The nonlesional skin surface distinguishes atopic dermatitis with food allergy as a unique endotype. Sci Transl Med. 2019;11(480). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Hanifin JM, Rajka G. Diagnostic features of atopic dermatitis. Acta Derm Venereol. 1980;92 (suppl):44–7. [Google Scholar]
- 20.von Kobyletzki LB, Berner A, Carlstedt F, Hasselgren M, Bornehag CG, Svensson A. Validation of a parental questionnaire to identify atopic dermatitis in a population-based sample of children up to 2 years of age. Dermatology. 2013;226(3):222–6. [DOI] [PubMed] [Google Scholar]
- 21.Bieber T Atopic dermatitis. N Engl J Med. 2008;358(14):1483–94. [DOI] [PubMed] [Google Scholar]
- 22.Alduraywish SA, Standl M, Lodge CJ, Abramson MJ, Allen KJ, Erbas B, et al. Is there a march from early food sensitization to later childhood allergic airway disease? Results from two prospective birth cohort studies. Pediatr Allergy Immunol. 2017;28(1):30–7. [DOI] [PubMed] [Google Scholar]
- 23.Johansson E, Biagini Myers JM, Martin LJ, He H, Ryan P, LeMasters GK, et al. Identification of two early life eczema and non-eczema phenotypes with high risk for asthma development. Clin Exp Allergy. 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Nissen SP, Kjaer HF, Host A, Nielsen J, Halken S. The natural course of sensitization and allergic diseases from childhood to adulthood. Pediatr Allergy Immunol. 2013;24(6):549–55. [DOI] [PubMed] [Google Scholar]
- 25.LeMasters GK, Wilson K, Levin L, Biagini J, Ryan P, Lockey JE, et al. High prevalence of aeroallergen sensitization among infants of atopic parents. J Pediatr. 2006;149(4):505–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Flohr C, Perkin M, Logan K, Marrs T, Radulovic S, Campbell LE, et al. Atopic dermatitis and disease severity are the main risk factors for food sensitization in exclusively breastfed infants. The Journal of investigative dermatology. 2014;134(2):345–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Chan SM, Turcanu V, Stephens AC, Fox AT, Grieve AP, Lack G. Cutaneous lymphocyte antigen and alpha4beta7 T-lymphocyte responses are associated with peanut allergy and tolerance in children. Allergy. 2012;67(3):336–42. [DOI] [PubMed] [Google Scholar]
- 28.Beko G, Morrison G, Weschler CJ, Koch HM, Palmke C, Salthammer T, et al. Dermal uptake of nicotine from air and clothing: Experimental verification. Indoor Air. 2018;28(2):247–57. [DOI] [PubMed] [Google Scholar]
- 29.Beko G, Morrison G, Weschler CJ, Koch HM, Palmke C, Salthammer T, et al. Measurements of dermal uptake of nicotine directly from air and clothing. Indoor Air. 2017;27(2):427–33. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.