Atopic dermatitis in children. Skin barrier dysfunction and therapeutic strategies. A structured review applying systematic-review methods

 

 

Nº de DOI: 10.34896/RSI.2026.79.83.002

 

 

AUTHORS

1. Adriana Yisely Ortiz Cabrera. General Practitioner. Affiliated with Centro Hospital Nuestro Señor de la Divina Misericordia. Graduate of Universidad Cooperativa de Colombia. Based in Puerres, Colombia. https://orcid.org/0009-0004-1013-1925
2. Ariana Marisol Andrade Sarauz. General Practitioner. Affiliated with Hospital Pediátrico Baca Ortiz. Graduate of Pontificia Universidad Católica del Ecuador. Based in Quito, Ecuador. https://orcid.org/0009-0009-4439-9686
3. María José Orbea Tovar. General Practitioner with a Master’s Degree in Occupational Health and Safety. Affiliated with Hospital General Latacunga. Graduate of Universidad Regional Autónoma de los Andes. Based in Latacunga, Ecuador. https://orcid.org/0009-0006-7260-6468
4. Daniela del Carmen Bermúdez Padilla. General Practitioner. Affiliated with Clínicas Privadas del Ecuador. Graduate of Universidad de las Américas. Based in Quito, Ecuador. https://orcid.org/0009-0000-6994-186X
5. Valeria Estefania Veloz Jara. General Practitioner. Affiliated with Consultorio Médico de Medicina General Integral Dr. Luis Molina. Graduate of Universidad Nacional de Chimborazo. Based in Latacunga, Ecuador. https://orcid.org/0009-0000-7878-5870

ABSTRACT

Atopic dermatitis (AD) is the most common chronic inflammatory skin disease of childhood, and impaired epidermal barrier function is regarded as one of its central mechanisms.

Objective: To synthesise current evidence on the causes and consequences of skin barrier dysfunction in children with AD and on the therapeutic strategies that target the barrier, the inflammation it permits, or both.

Methods: A criteria-based search and screening process was carried out in June 2026. Only peer-reviewed journal articles with a verifiable digital object identifier (DOI) were eligible, and 23 were retained. Findings were synthesised narratively because study designs and outcomes were too heterogeneous for pooling.

Results: Loss-of-function variants in the filaggrin gene are the strongest known genetic risk factor for AD; in one Irish case-control study, homozygous carriers had an odds ratio of 151 for moderate-to-severe childhood eczema. Barrier failure is amplified by reduced ceramides, increased protease activity, a type 2 and type 22 dominated immune response, and shifts in the skin microbiome, with Staphylococcus aureus expanding during flares. Trials of daily emollients from birth gave conflicting results: two small trials suggested benefit, whereas two large trials (n = 1,394 and n = 2,397) found no preventive effect. Emollients, topical corticosteroids and calcineurin inhibitors remain the basis of treatment, supplemented by the topical phosphodiesterase-4 inhibitor crisaborole and the topical Janus kinase inhibitor ruxolitinib. For severe disease, dupilumab has shown efficacy and an acceptable short-term safety profile in randomised trials in children from 6 months of age.

Conclusions: Barrier repair is a necessary foundation of paediatric AD care, but current evidence supports it as treatment rather than as a proven strategy for primary prevention. Long-term safety data, head-to-head comparisons and barrier-specific biomarkers are the main evidence gaps.

KEY WORDS

Atopic dermatitis, eczema, child, skin barrier, filaggrin, emollients, topical therapy, dupilumab, skin microbiome, atopic march.

RESUMEN

La dermatitis atópica (DA) es la enfermedad inflamatoria cutánea crónica más frecuente en la infancia, y el deterioro de la función de barrera epidérmica se considera uno de sus mecanismos centrales.

Objetivo: Sintetizar la evidencia actual sobre las causas y consecuencias de la disfunción de la barrera cutánea en niños con DA, así como sobre las estrategias terapéuticas dirigidas a la barrera, a la inflamación que esta propicia o a ambos aspectos.

Métodos: En junio de 2026 se llevó a cabo un proceso de búsqueda y selección basado en criterios preestablecidos. Solo se incluyeron artículos de revistas revisadas por pares con un identificador de objeto digital (DOI) verificable; finalmente, se seleccionaron 23 artículos. Los hallazgos se sintetizaron de forma narrativa, dado que los diseños de los estudios y las variables de resultado presentaban una heterogeneidad excesiva para realizar un metaanálisis.

Resultados: Las variantes de pérdida de función en el gen de la filagrina constituyen el factor de riesgo genético conocido más importante para la DA; en un estudio de casos y controles realizado en Irlanda, los portadores homocigotos presentaron una *odds ratio* de 151 para el eccema infantil de moderado a grave. El fallo de la barrera se ve agravado por la reducción de ceramidas, el aumento de la actividad de las proteasas, una respuesta inmunitaria dominada por los perfiles tipo 2 y tipo 22, y alteraciones en el microbioma cutáneo, con una proliferación de *Staphylococcus aureus* durante los brotes. Los ensayos sobre el uso diario de emolientes desde el nacimiento arrojaron resultados contradictorios: dos estudios pequeños sugirieron un beneficio, mientras que dos estudios de gran tamaño (n = 1.394 y n = 2.397) no hallaron ningún efecto preventivo. Los emolientes, los corticosteroides tópicos y los inhibidores de la calcineurina siguen constituyendo la base del tratamiento, complementados por la crisaborola (inhibidor tópico de la fosfodiesterasa 4) y el ruxolitinib (inhibidor tópico de la Janus quinasa). En casos de enfermedad grave, el dupilumab ha demostrado eficacia y un perfil de seguridad aceptable a corto plazo en ensayos aleatorizados realizados en niños a partir de los 6 meses de edad.

Conclusiones: La reparación de la barrera es un pilar fundamental en el manejo de la DA pediátrica; sin embargo, la evidencia actual respalda su uso como tratamiento y no como una estrategia demostrada de prevención primaria. La falta de datos sobre seguridad a largo plazo, de comparaciones directas entre tratamientos y de biomarcadores específicos de la barrera constituye las principales lagunas en la evidencia.

PALABRAS CLAVE

Dermatitis atópica, eccema, niño, barrera cutánea, filagrina, emolientes, terapia tópica, dupilumab, microbioma cutáneo, marcha atópica.

INTRODUCTION

La dermatitis atópica (DA) es la enfermedad inflamatoria crónica de la piel más frecuente en la infancia, y la alteración de la función de barrera epidérmica se considera uno de sus mecanismos centrales.

OBJECTIVE

The objective of this review was therefore threefold. The first aim was to summarise the evidence on how skin barrier dysfunction arises in children with AD, including genetic, lipid, immune and microbial contributions. The second aim was to appraise the evidence for barrier-directed strategies, namely emollient-based prevention and treatment. The third aim was to review the efficacy and safety evidence for the topical and systemic anti-inflammatory therapies that are used, or are emerging, in paediatric practice. The review is intended to help clinicians and researchers see where the evidence is solid, where it is conflicting, and where it is still missing, and to translate these findings into practical messages for the care of children and their families.

METHODOLOGY

Design and reporting:

This is a structured review that applies systematic-review methods: an explicit question, pre-specified eligibility criteria, a documented search strategy, a screening process, structured data extraction and a transparent narrative synthesis. Its reporting follows the spirit of the PRISMA 2020 statement. It must be stated plainly, however, that the review was not prospectively registered, that no direct Boolean searches of bibliographic databases such as MEDLINE or Embase were exported, and that a formal PRISMA flow diagram with database-level record counts therefore cannot be provided. The limitations that follow from these choices are discussed below, and readers should regard the conclusions as those of a rigorous structured review rather than of a formally registered systematic review.

Review question:

The question was framed with a population-intervention-outcome structure. The population was children and adolescents (0 to 17 years) with atopic dermatitis, together with studies of infants at risk of developing it and mechanistic studies that explain barrier dysfunction in this population. The interventions and exposures of interest were genetic and structural barrier defects, immune and microbial changes linked to the barrier, barrier-directed interventions such as emollients and moisturisers, and topical or systemic anti-inflammatory therapies. The outcomes were disease incidence, severity (for example, Investigator’s Global Assessment and Eczema Area and Severity Index scores), symptoms, quality of life and safety.

Search strategy:

Searches were run in June 2026 through a general web search engine that indexes publisher pages, PubMed records and institutional repositories. Search statements combined controlled and free-text terms in English, including “atopic dermatitis”, “eczema”, “children”, “skin barrier”, “filaggrin”, “emollient”, “moisturizer”, “prevention”, “crisaborole”, “ruxolitinib”, “tacrolimus”, “dupilumab”, “microbiome” and “guidelines”. Separate searches were made for each topic (genetics, immunology, microbiome, prevention, topical therapy, systemic therapy, guidelines), and the reference lists of key articles were used to identify further records. Approximately 35 candidate records were examined in full or in abstract.

Eligibility criteria:

Records were eligible if they met all of the following criteria. They had to be full articles published in a peer-reviewed journal; they had to have a digital object identifier (DOI) that could be cross-checked against at least one publisher, repository or indexing record; and they had to be written in English. Eligible designs were randomised controlled trials and their pre-specified or post hoc analyses, genetic association and cohort studies, mechanistic or transcriptomic studies, cross-sectional epidemiological surveys, clinical practice guidelines, and narrative or systematic reviews. The content had to address children or adolescents with AD, infants at risk of AD, or biological mechanisms of barrier dysfunction directly relevant to them. Studies in adults were accepted only as supporting evidence for mechanism or for drugs studied later in children.

Records were excluded if they were trial protocols, trial registry entries, conference abstracts, preprints, news items, commentaries without data, or web summaries. Publication in a journal alone was not sufficient: records without a verifiable DOI were excluded. One otherwise relevant cohort study on neonatal transepidermal water loss was set aside because an indexing record carried a retraction label that could not be resolved; the conclusions below do not rely on it. A conference abstract describing a Japanese moisturiser trial was replaced by the full journal article.

Screening and data extraction:

Titles and abstracts were screened by a single reviewer against the criteria above, and full texts or detailed abstracts were examined for the records that passed. No independent second screening was performed. For each retained article, the following information was extracted: first author, year, journal, study design, population and sample size, intervention or exposure, comparator, main outcomes and principal safety findings. Bibliographic details and DOIs were verified for every reference against at least one publisher, institutional repository or indexing record.

Appraisal and synthesis:

Formal risk-of-bias tools were not applied. Instead, each study was appraised narratively according to design features that affect internal validity and applicability: randomisation and allocation concealment, masking (including whether outcome assessors were masked when participants could not be), adherence to the intervention, loss to follow-up, trial duration, the age of participants, and funding or conflict-of-interest disclosures. The data were heterogeneous in design, comparators, outcomes and follow-up, so no meta-analysis was attempted. Findings are presented narratively under thematic headings, and effect sizes are quoted as reported by the original authors. Where a finding came from a post hoc or secondary analysis, this is stated, because such analyses are more exposed to selection and multiplicity problems than primary pre-specified analyses.

RESULTS

Characteristics of the included evidence:

Twenty-three journal articles were included. Six were reviews or overviews of the pathogenesis of AD and of filaggrin biology. Two were clinical practice guidelines, one European and one American. One was a large cross-sectional epidemiological survey. Four were genetic, transcriptomic or microbiome studies. Four were randomised trials of barrier-directed prevention in infants. Six reported randomised trials, or analyses of randomised trials, of topical or systemic treatment. The articles appeared in general medical journals (The Lancet, New England Journal of Medicine), specialist dermatology journals, and allergy and immunology journals, and they were published between 2007 and 2024. Table 1, at the end of this section, summarises the key randomised trials in children and infants.

Genetic and structural basis of barrier dysfunction:

The stratum corneum and filaggrin:

The stratum corneum, the outermost layer of the epidermis, is the principal physical barrier of the skin. It is often described with a “brick and mortar” analogy in which flattened corneocytes are embedded in lamellar lipids composed mainly of ceramides, cholesterol and free fatty acids. Filaggrin is a histidine-rich protein derived from its precursor profilaggrin, and it contributes to the aggregation of keratin filaments, to the formation of the cornified envelope, and, through its breakdown products, to skin hydration, acidity and antimicrobial defence⁷. The barrier it helps to build has to do several things at once: it limits water loss from the body, keeps out allergens, irritants and microbes, and maintains a mildly acidic surface that favours the normal skin flora and restrains proteases^1,2,7.

Filaggrin loss-of-function variants:

In a full sequencing study of the very large and repetitive FLG gene, investigators described 15 variants, seven of them prevalent, all of which were nonsense or frameshift mutations that abolished filaggrin production in representative cases⁸. In an Irish case-control study, the five most common European mutations were strongly associated with moderate-to-severe childhood eczema, with a heterozygote odds ratio of 7.44 (95% confidence interval 4.9 to 11.3) and a homozygote odds ratio of 151 (95% confidence interval 20 to 1,136) ⁸. The authors also identified three additional rare null mutations in the case series, which suggests that the genetic architecture of filaggrin-related AD includes both prevalent and rare risk alleles⁸. Reviews place the cumulative carrier frequency of FLG null alleles at around 10% in Europeans and emphasise that carrying a mutation does not make AD inevitable, since many carriers never develop eczema and many children with AD carry no mutation^1,5,6. Filaggrin deficiency should therefore be viewed as a strong but neither necessary nor sufficient cause.

From skin to systemic sensitization:

The significance of filaggrin loss extends beyond the skin. A mouse model, the “flaky tail” strain carrying a one-base-pair deletion (5303delA) analogous to common human FLG mutations, showed enhanced percutaneous allergen priming and elevated IgE sensitisation, providing experimental support for the hypothesis that antigen transfer through a defective barrier is a key step in the initiation of allergic disease⁹. In humans, loss-of-function FLG variants were a significant risk factor for IgE-mediated peanut allergy in a European study, with an odds ratio of 5.3 (95% confidence interval 2.8 to 10.2) in food-challenge-positive patients, replicated in a Canadian cohort with an odds ratio of 1.9 (95% confidence interval 1.4 to 2.6); the association remained significant after controlling for coexistent AD¹⁰. These findings underpin the concept of the atopic march, the progression from early AD to food allergy, allergic rhinitis and asthma, and the idea that skin barrier failure may initiate systemic sensitisation^6,7.

Lipids, proteases and pH:

Filaggrin deficiency is only one component of barrier failure. A recent review of the filaggrin literature notes that the most investigated causes of epidermal barrier impairment in AD, regarded as the starting point of the atopic march, are a lack of filaggrin accompanied by a decrease in ceramides and a marked activation of epidermal proteases⁷. Reduced filaggrin function also raises skin surface pH, which enhances protease activity and weakens antimicrobial defence, and it increases the activity of thymic stromal lymphopoietin (TSLP), an epithelial cytokine that acts through dendritic cells to promote Th2 adaptive and innate immune responses⁷. In this model, a structural defect and an immune response form a self-reinforcing loop: a leaky barrier lets in antigens, antigens trigger type 2 inflammation, and type 2 cytokines further suppress the expression of barrier proteins^2,4,5,7.

Immune dysregulation and the skin microbiome:

Type 2 and type 22 inflammation:

Gene-expression profiling of paired samples from non-lesional skin, acute lesions and chronic lesions in ten patients showed that the onset of acute lesions was associated with a striking increase in the terminal differentiation proteins S100A7, S100A8 and S100A9, together with significant increases in Th2 and Th22 cytokines and smaller increases in IL-17¹¹. Th1-associated genes were induced to a lesser degree in acute disease, and some were significantly upregulated in chronic disease; major Th22 and Th2 cytokines were further intensified between acute and chronic lesions¹¹. The authors proposed a model of progressive activation of the Th2 and Th22 axes from acute to chronic disease, which expanded the prevailing biphasic model and had important therapeutic implications¹¹. The Th2 cytokines IL-4 and IL-13 are the pathways targeted by dupilumab, discussed below.

Microbial dysbiosis:

Staphylococcal colonisation has long been associated with AD. A longitudinal 16S ribosomal RNA sequencing study followed 12 children with moderate-to-severe AD, aged 2 to 15 years, at baseline, during flares and after treatment of flares, and compared them with 11 healthy controls¹². Skin bacterial diversity depended on recent treatment, with even intermittent treatment linked to greater diversity than no recent treatment, suggesting that treatment diversifies skin bacteria before clinical improvement. The proportion of Staphylococcus aureus was greater during flares than at baseline or after treatment and correlated with disease severity, and S. epidermidis also increased during flares¹². These data link barrier and immune dysfunction to a dynamic microbial community and suggest that the microbiome is both a consequence of, and a contributor to, flares, although the sample was small and causality cannot be established from observational sequencing alone.

Barrier-directed prevention in infants:

If barrier failure precedes eczema, protecting the barrier from birth might prevent it. Two small trials suggested that this could be true. In a randomised trial of 124 neonates at high risk of AD in the United States and the United Kingdom, whole-body daily emollient from birth reduced the cumulative incidence of AD at 6 months (about 22% in the emollient group against 43% in controls), and the authors judged the approach feasible, safe and effective¹³. In a Japanese randomised trial of 118 high-risk neonates, an emulsion-type moisturiser applied daily during the first 32 weeks of life was associated with about 32% fewer cases of AD or eczema than petroleum jelly applied as needed (p = 0.012); 47 infants developed AD or eczema, 19 in the intervention group and 28 in the control group¹⁴.

Two much larger trials, published together, did not confirm these results. In the Barrier Enhancement for Eczema Prevention (BEEP) trial in England, 1,394 newborns with a family history of atopic disease were randomised to daily emollient (Diprobase cream or DoubleBase gel) for the first year plus standard skin-care advice, or to standard advice alone; 693 infants were assigned to the emollient group and 701 to the control group¹⁵. Adherence in the emollient group was 88% at 3 months, 82% at 6 months and 74% at 12 months. At 2 years of age, eczema was present in 139 of 598 infants (23%) with outcome data in the emollient group, and the investigators found no evidence that daily emollient during the first year prevents eczema, together with some evidence of an increased risk of skin infections¹⁵. In PreventADALL, a factorial cluster-randomised trial in Norway and Sweden, 2,397 infants from the general population were allocated to no intervention, skin emollients (bath additives and facial cream), early complementary feeding of peanut, cow’s milk, wheat and egg, or both interventions¹⁶. Neither skin emollients nor early feeding reduced AD by 12 months; the risk difference was 3.1% (95% confidence interval −0.3 to 6.5) for the skin intervention and 1.0% (95% confidence interval −2.1 to 4.1) for the food intervention, in both cases in favour of the control group¹⁶.

Topical therapy:

Foundations: emollients, corticosteroids and calcineurin inhibitors:

European consensus guidelines identify basic therapy as the treatment of disturbed barrier function with hydrating and lubricating topical products, combined with avoidance of specific and non-specific provocation factors³. Topical anti-inflammatory treatment based on glucocorticosteroids and calcineurin inhibitors is used for flare management and for proactive therapy to maintain long-term control, and the guidelines add that topical phosphodiesterase inhibitors may be an alternative where available [3]. The guidelines also stress education of patients and families as a component of care. None of the included trials altered this hierarchy.

Crisaborole:

Crisaborole 2% ointment is a nonsteroidal topical phosphodiesterase-4 inhibitor. Two identically designed phase 3 trials enrolled patients aged 2 years or older with mild-to-moderate AD and randomised them 2:1 to crisaborole or vehicle twice daily for 28 days¹⁷. The primary endpoint was an Investigator’s Static Global Assessment of clear or almost clear skin with at least a two-grade improvement at day 29. Crisaborole-treated patients reached treatment success and improvement in itch earlier than vehicle-treated patients (both p ≤ 0.001), and the authors reported a favourable safety profile¹⁷. The trials were short and vehicle-controlled, so they do not show how crisaborole compares with topical corticosteroids or calcineurin inhibitors, nor do they provide long-term data.

Ruxolitinib cream:

Ruxolitinib is a selective inhibitor of Janus kinases 1 and 2 formulated as a cream. In two identical phase 3 studies (TRuE-AD1 and TRuE-AD2), 631 and 618 patients aged 12 years or older with mild-to-moderate AD affecting 3% to 20% of body surface area were randomised 2:2:1 to 0.75% ruxolitinib cream, 1.5% ruxolitinib cream or vehicle twice daily for 8 weeks [18]. At week 8, Investigator’s Global Assessment treatment success was reached by 50.0% and 39.0% with 0.75% cream and 53.8% and 51.3% with 1.5% cream, against 15.1% and 7.6% with vehicle (p < 0.0001), with prompt antipruritic effects and good tolerability¹⁸. The authors noted that longer-term safety data were not yet available at the time of that analysis. Because the age floor was 12 years, these data inform adolescents but not younger children.

Systemic therapy:

Dupilumab in adults as the reference point:

Dupilumab is a monoclonal antibody against the alpha subunit of the interleukin-4 receptor, thereby blocking signalling of both IL-4 and IL-13. Two phase 3 trials in adults with moderate-to-severe AD established its efficacy against placebo, and the paediatric programme extended this evidence downward in age¹⁹.

Children aged 6 to 11 years

In a 16-week, double-blind phase 3 trial in 367 children aged 6 to 11 years with severe AD inadequately controlled by topical therapies, patients were randomised 1:1:1 to dupilumab 300 mg every 4 weeks, a weight-based regimen every 2 weeks (100 mg below 30 kg, 200 mg at 30 kg or above), or placebo, each with concomitant medium-potency topical corticosteroids²⁰. Both dupilumab regimens produced clinically meaningful and statistically significant improvements in signs, symptoms and quality of life against placebo in all prespecified endpoints. The response was weight-dependent: the optimal regimens for efficacy and safety were 300 mg every 4 weeks in children below 30 kg and 200 mg every 2 weeks in children at or above 30 kg. Conjunctivitis and injection-site reactions were more common with dupilumab, and the limitations were the short 16-week period and the restriction to severe disease²⁰.

Children aged 6 months to under 6 years:

A phase 3 trial randomised 162 children aged 6 months to under 6 years with moderate-to-severe AD, inadequately controlled by topical corticosteroids, to dupilumab (200 mg for body weight of 5 kg to under 15 kg; 300 mg for 15 kg to under 30 kg) every 4 weeks or placebo, both with low-potency topical corticosteroid (hydrocortisone acetate 1% cream), for 16 weeks²¹. Eighty-three children received dupilumab and 79 received placebo. At week 16, clear or almost clear skin (Investigator’s Global Assessment 0 to 1) was achieved by 28% with dupilumab against 4% with placebo (difference 24%, 95% confidence interval 13 to 34), and a 75% improvement in the Eczema Area and Severity Index (EASI-75) by 53% against 11% (difference 42%, 95% confidence interval 29 to 55; both p < 0.0001). Adverse events occurred in 64% of the dupilumab group and 74% of the placebo group, conjunctivitis occurred in 5% of dupilumab-treated children and none of the placebo group, and no dupilumab-related adverse events were serious or led to discontinuation²¹.

Sustained response:

A post hoc analysis combined data from two paediatric phase 3 trials and a one-year interim analysis of an open-label extension, covering 471 patients (304 children aged 6 to under 12 years and 167 adolescents aged 12 to under 18 years)²². Patients on dupilumab, with or without topical corticosteroids, had significantly lower SCORAD, objective SCORAD and individual SCORAD components than placebo from week 3 to week 16. The improvements were sustained or continued to improve over one year of open-label treatment: SCORAD-50 was reached by 91.3% to 91.8% of patients by week 52, and more than 86% had mild or absent pruritus and sleep loss at that time²².

Other systemic options:

The most recent American Academy of Dermatology guideline on phototherapy and systemic therapies reports strong recommendations for dupilumab, tralokinumab, abrocitinib, baricitinib and upadacitinib, conditional recommendations for phototherapy, azathioprine, cyclosporine, methotrexate and mycophenolate, and a recommendation against systemic corticosteroids²³. This guideline applies to adults; it illustrates the range of agents available, but it cannot be applied directly to young children.

DISCUSSION

This review brings together evidence that places skin barrier dysfunction at the centre of paediatric AD and shows that the clinical response to the barrier hypothesis has been uneven. On the mechanistic side, the evidence is strong and coherent. Genetic data identify filaggrin deficiency as a powerful risk factor, experimental data show that a defective barrier permits sensitisation, and transcriptomic and microbiome studies describe a feedback loop in which type 2 and type 22 inflammation, microbial change and barrier failure sustain one another^7,8,9,11,12. On the clinical side, the evidence supports barrier-directed care as the foundation of treatment, but it does not support emollients as a proven strategy for the primary prevention of AD in infants. The therapeutic landscape, meanwhile, has broadened: nonsteroidal topical agents are now available alongside corticosteroids and calcineurin inhibitors, and a targeted biologic has shown efficacy in randomised trials down to the age of six months^17,18,20,21.

The most instructive finding for practice is the divergence between the early prevention trials and the two large trials that followed them. The pilot trials were small, were conducted in high-risk infants, and showed relative reductions of roughly one third to one half^13,14. BEEP, with more than ten times the number of participants of either pilot, found no benefit and a possible increase in skin infections, and PreventADALL, in an unselected population, found a numerically higher rate of AD in the skin-intervention arm. Several explanations deserve consideration^15,16.

First, small trials are prone to overestimating effects, particularly when the outcome is a clinical diagnosis and masking of participants is impossible. In an emollient trial, parents know which arm they are in, and observer bias is limited only to the extent that assessors are masked. Second, the products differed: the trials used different emollients and application schedules, and a bath additive plus facial cream is not equivalent to whole-body daily application of a cream or gel^13,14,15,16. Third, adherence decreases over time. In BEEP, adherence in the emollient arm was 88% at 3 months but 74% at 12 months, and a prevention strategy that depends on daily application for a year may fail simply because it is difficult to deliver¹⁵. Fourth, the mechanism itself may be more complex than the simple barrier hypothesis implies. Barrier dysfunction is partly a consequence of early, subclinical inflammation as well as a cause of it, so moisturising a defective barrier from birth may not be enough to prevent the immune events that lead to AD^7,11. Fifth, the populations differed in baseline risk and in genotype, and none of the large trials could show whether infants with FLG null mutations, who have the clearest structural defect, respond differently from others⁸.

None of this undermines the use of emollients as treatment. In BEEP, the investigators stressed that their data relate only to prevention and do not challenge the practice of using emollients as first-line treatment for established eczema, which remains the recommendation in guidelines^3,15. The practical message for families is twofold: they should not be told that routine emollient use in a healthy infant will prevent eczema, and they should not be discouraged from moisturising a child who already has dry or eczematous skin. The signal of more skin infections in BEEP also argues for caution against presenting emollients as risk-free in all circumstances¹⁵. Other formulations, particularly those designed to replace lipids deficient in the atopic stratum corneum, remain under investigation, and the question of whether any barrier-directed intervention can alter the course of AD is still open.

The two nonsteroidal topical agents reviewed here have important strengths. Crisaborole was effective against vehicle in patients from 2 years of age, and ruxolitinib cream achieved Investigator’s Global Assessment success in about half of treated adolescents and adults, with a rapid effect on itch^17,18. Both offer an option for body sites or situations in which long-term corticosteroid or calcineurin inhibitor use is a concern, and their vehicles are themselves emollient. The evidence has limits that matter in clinical decisions. Both programmes were vehicle-controlled and short (28 days and 8 weeks), and neither compared the new agent with a topical corticosteroid or a calcineurin inhibitor^17,18. Consequently the review cannot rank these agents against established first-line therapy. The ruxolitinib trials enrolled patients aged 12 years or older, so evidence for younger children must come from other studies not included here¹⁸. Cost and access are also real barriers: a more expensive agent may not be justified where a low-potency corticosteroid is effective, and in many health systems access to newer topical drugs is restricted. Guidelines continue to describe the corticosteroid and calcineurin inhibitor approach, including proactive maintenance therapy, as the backbone of anti-inflammatory treatment³.

The dupilumab trials represent the most robust paediatric evidence for severe disease, given their randomised, double-blind, placebo-controlled design and consistent results across two age bands^20,21. They also highlight points that clinicians should weigh. First, dosing is weight-based, and the weight-dependent response in children aged 6 to 11 years shows why adult dosing cannot be extrapolated²⁰. Second, conjunctivitis was more frequent with dupilumab in both trials, with 5% affected in the youngest children, and ocular symptoms should be anticipated and monitored^20,21. Third, all the randomised periods lasted only 16 weeks and used concomitant topical corticosteroids, so the trials show the benefit of dupilumab as an add-on to topical care rather than as monotherapy^20,21. Fourth, the open-label data up to one year are encouraging, with SCORAD-50 reached by more than 90% of continuing patients, but open-label extensions are vulnerable to selection of patients who tolerate and respond to treatment, and they lack a control group²².

The wider range of systemic agents available to adults, including other biologics and oral Janus kinase inhibitors, is described in a recent guideline, but its recommendations apply to adults and the evidence in young children is far thinner²³. The decision to move a child from optimised topical therapy to a systemic drug should be taken by specialists, based on disease severity, effect on quality of life, comorbidity, safety monitoring and local access. Regulatory age thresholds differ between countries and change over time, and clinicians should consult current local labelling. The adult trials of dupilumab, which showed benefit in two large phase 3 studies, provide a safety and efficacy reference but cannot replace paediatric data¹⁹.

Taken together, the findings suggest a layered logic of treatment. The barrier is the base: emollients and avoidance of irritants address the structural defect and are justified by the mechanistic evidence^3,7. Topical anti-inflammatory therapy interrupts the loop between a leaky barrier and type 2 inflammation^3,11. Systemic targeted therapy blocks the dominant cytokine pathways when the loop cannot be interrupted topically^20,21. The microbiome adds a further dimension. The expansion of S. aureus during flares and the diversification of skin bacteria with treatment indicate that effective anti-inflammatory treatment may also normalise the microbial community, although the available data are small, observational and do not show whether microbial changes are a cause or an effect¹². Better evidence may come from trials that measure barrier function and microbial composition alongside clinical scores.

The strengths of this review are its explicit question and eligibility criteria, its restriction to peer-reviewed journal articles whose DOIs were cross-checked, and its combination of mechanistic and clinical evidence in one paediatric-focused synthesis. Its limitations are substantial and should be considered when interpreting the conclusions. The review was not registered in advance. The search was run through a web search engine rather than through exported searches of bibliographic databases, so relevant studies may have been missed and a formal PRISMA flow diagram cannot be provided. Screening and data extraction were performed by a single reviewer without independent verification. Risk of bias was appraised narratively with no formal instrument, and no meta-analysis was done. The included trials were largely sponsored by manufacturers or large public funders and were short, which limits conclusions about long-term safety. Finally, because the review required a verifiable DOI and a journal publication, some relevant paediatric evidence, such as trials reported only in registries or abstracts, was necessarily excluded, and several important topics, including calcineurin inhibitor long-term safety, wet-wrap therapy, bleach baths and probiotics, were not covered by the included sources. A full systematic review following PRISMA 2020, with a registered protocol, searches of at least two bibliographic databases, duplicate screening and a formal risk-of-bias assessment, would be the natural next step.

For practice, the evidence supports the following points. Barrier-directed skin care should be offered to every child with AD, taught to parents and caregivers, and reinforced at every visit. Emollients should not be promoted as a means to prevent AD in healthy or high-risk infants on the basis of the current evidence. Topical anti-inflammatory therapy should be used proactively and without undue fear of undertreatment. Nonsteroidal topical agents provide further options but have not been compared with established treatment. Systemic therapy with dupilumab is effective for severe disease from the age of six months in the trials reviewed, with weight-based dosing and attention to conjunctivitis.

For research, the main priorities are as follows. Trials of barrier-directed prevention should test products designed around the specific lipid and pH abnormalities of atopic skin, should report adherence objectively, and should stratify by FLG genotype to find whether any subgroup benefits. Long-term safety registries and extension studies should follow children treated with topical Janus kinase inhibitors and with biologics into adulthood. Head-to-head trials should compare nonsteroidal topical agents with corticosteroids and calcineurin inhibitors. Validated barrier biomarkers, measured early in life, are needed to identify infants who might benefit from intervention. Finally, trials should determine whether early, effective control of inflammation, rather than barrier protection alone, alters the progression to food allergy, allergic rhinitis and asthma that is described as the atopic march^6,7.

CONCLUSIONS

Skin barrier dysfunction is a central feature of atopic dermatitis in children. It results from genetic defects, most prominently loss-of-function variants in the filaggrin gene, and from reduced ceramides, increased protease activity, a raised skin pH, type 2 and type 22 immune activation and microbial change, which together form a self-reinforcing cycle. Because of this, restoring and protecting the barrier is justified as the foundation of treatment, and current guidelines endorse it.

The evidence does not, however, support emollients as a means of primary prevention. Two small trials with encouraging results were not confirmed by two large trials, one of which reported more skin infections in the emollient group. Families should be told that emollients treat dry, eczematous skin but have not been shown to prevent the disease.

For anti-inflammatory treatment, topical corticosteroids and calcineurin inhibitors remain the established first line. Crisaborole and ruxolitinib cream are effective in short vehicle-controlled trials, and dupilumab has produced significant, sustained improvement in children with moderate-to-severe or severe disease from 6 months of age, with conjunctivitis as the most notable adverse effect. Questions of long-term safety, comparative effectiveness, optimal sequencing of therapies and the prevention of the atopic march remain unresolved. Because this review was a structured review with the methodological limitations stated above, its conclusions should be confirmed by a formally registered systematic review before being used to support guideline recommendations.

REFERENCES

1. Langan SM, Irvine AD, Weidinger S. Atopic dermatitis. Lancet. 2020;396(10247):345-60. doi: 10.1016/S0140-6736(20)31286-1.
2. Odhiambo JA, Williams HC, Clayton TO, Robertson CF, Asher MI; ISAAC Phase Three Study Group. Global variations in prevalence of eczema symptoms in children from ISAAC Phase Three. J Allergy Clin Immunol. 2009;124(6):1251-8.e23. doi: 10.1016/j.jaci.2009.10.009.
3. Wollenberg A, Barbarot S, Bieber T, Christen-Zaech S, Deleuran M, Fink-Wagner A, et al. Consensus-based European guidelines for treatment of atopic eczema (atopic dermatitis) in adults and children: part I. J Eur Acad Dermatol Venereol. 2018;32(5):657-82. doi: 10.1111/jdv.14891.
4. Weidinger S, Novak N. Atopic dermatitis. Lancet. 2016;387(10023):1109-22. doi: 10.1016/S0140-6736(15)00149-X.
5. Weidinger S, Beck LA, Bieber T, Kabashima K, Irvine AD. Atopic dermatitis. Nat Rev Dis Primers. 2018;4:1. doi: 10.1038/s41572-018-0001-z.
6. Irvine AD, McLean WHI, Leung DYM. Filaggrin mutations associated with skin and allergic diseases. N Engl J Med. 2011;365(14):1315-27. doi: 10.1056/NEJMra1011040.
7. Čepelak I, Dodig S, Pavić I. Filaggrin and atopic march. Biochem Med (Zagreb). 2019;29(2):020501. doi: 10.11613/BM.2019.020501.
8. Sandilands A, Terron-Kwiatkowski A, Hull PR, O’Regan GM, Clayton TH, Watson RM, et al. Comprehensive analysis of the gene encoding filaggrin uncovers prevalent and rare mutations in ichthyosis vulgaris and atopic eczema. Nat Genet. 2007;39(5):650-4. doi: 10.1038/ng2020.
9. Fallon PG, Sasaki T, Sandilands A, Campbell LE, Saunders SP, Mangan NE, et al. A homozygous frameshift mutation in the mouse Flg gene facilitates enhanced percutaneous allergen priming. Nat Genet. 2009;41(5):602-8. doi: 10.1038/ng.358.
10. Brown SJ, Asai Y, Cordell HJ, Campbell LE, Zhao Y, Liao H, et al. Loss-of-function variants in the filaggrin gene are a significant risk factor for peanut allergy. J Allergy Clin Immunol. 2011;127(3):661-7. doi: 10.1016/j.jaci.2011.01.031.
11. Gittler JK, Shemer A, Suárez-Fariñas M, Fuentes-Duculan J, Gulewicz KJ, Wang CQF, et al. Progressive activation of TH2/TH22 cytokines and selective epidermal proteins characterizes acute and chronic atopic dermatitis. J Allergy Clin Immunol. 2012;130(6):1344-54. doi: 10.1016/j.jaci.2012.07.012.
12. Kong HH, Oh J, Deming C, Conlan S, Grice EA, Beatson MA, et al. Temporal shifts in the skin microbiome associated with disease flares and treatment in children with atopic dermatitis. Genome Res. 2012;22(5):850-9. doi: 10.1101/gr.131029.111.
13. Simpson EL, Chalmers JR, Hanifin JM, Thomas KS, Cork MJ, McLean WHI, et al. Emollient enhancement of the skin barrier from birth offers effective atopic dermatitis prevention. J Allergy Clin Immunol. 2014;134(4):818-23. doi: 10.1016/j.jaci.2014.08.005.
14. Horimukai K, Morita K, Narita M, Kondo M, Kitazawa H, Nozaki M, et al. Application of moisturizer to neonates prevents development of atopic dermatitis. J Allergy Clin Immunol. 2014;134(4):824-30.e6. doi: 10.1016/j.jaci.2014.07.060.
15. Chalmers JR, Haines RH, Bradshaw LE, Montgomery AA, Thomas KS, Brown SJ, et al.; BEEP study team. Daily emollient during infancy for prevention of eczema: the BEEP randomised controlled trial. Lancet. 2020;395(10228):962-72. doi: 10.1016/S0140-6736(19)32984-8.
16. Skjerven HO, Rehbinder EM, Vettukattil R, LeBlanc M, Granum B, Haugen G, et al. Skin emollient and early complementary feeding to prevent infant atopic dermatitis (PreventADALL): a factorial, multicentre, cluster-randomised trial. Lancet. 2020;395(10228):951-61. doi: 10.1016/S0140-6736(19)32983-6.
17. Paller AS, Tom WL, Lebwohl MG, Blumenthal RL, Boguniewicz M, Call RS, et al. Efficacy and safety of crisaborole ointment, a novel, nonsteroidal phosphodiesterase 4 (PDE4) inhibitor for the topical treatment of atopic dermatitis (AD) in children and adults. J Am Acad Dermatol. 2016;75(3):494-503.e6. doi: 10.1016/j.jaad.2016.05.046.
18. Papp K, Szepietowski JC, Kircik L, Toth D, Eichenfield LF, Leung DYM, et al. Efficacy and safety of ruxolitinib cream for the treatment of atopic dermatitis: results from 2 phase 3, randomized, double-blind studies. J Am Acad Dermatol. 2021;85(4):863-72. doi: 10.1016/j.jaad.2021.04.085.
19. Simpson EL, Bieber T, Guttman-Yassky E, Beck LA, Blauvelt A, Cork MJ, et al. Two phase 3 trials of dupilumab versus placebo in atopic dermatitis. N Engl J Med. 2016;375(24):2335-48. doi: 10.1056/NEJMoa1610020.
20. Paller AS, Siegfried EC, Thaçi D, Wollenberg A, Cork MJ, Arkwright PD, et al. Efficacy and safety of dupilumab with concomitant topical corticosteroids in children 6 to 11 years old with severe atopic dermatitis: a randomized, double-blinded, placebo-controlled phase 3 trial. J Am Acad Dermatol. 2020;83(5):1282-93. doi: 10.1016/j.jaad.2020.06.054.
21. Paller AS, Simpson EL, Siegfried EC, Cork MJ, Wollenberg A, Arkwright PD, et al. Dupilumab in children aged 6 months to younger than 6 years with uncontrolled atopic dermatitis: a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet. 2022;400(10356):908-19. doi: 10.1016/S0140-6736(22)01539-2.
22. Wollenberg A, Marcoux D, Silverberg JI, Aoki V, Baselga E, Zhang H, et al. Dupilumab provides rapid and sustained improvement in SCORing Atopic Dermatitis outcomes in paediatric patients with atopic dermatitis. Acta Derm Venereol. 2022;102:adv00726. doi: 10.2340/actadv.v102.854.
23. Davis DMR, Drucker AM, Alikhan A, Bercovitch L, Cohen DE, Darr JM, et al. Executive summary: guidelines of care for the management of atopic dermatitis in adults with phototherapy and systemic therapies. J Am Acad Dermatol. 2024;90(2):342-5. doi: 10.1016/j.jaad.2023.08.103.

APPENDICE

TABLE 1. Key randomised trials in children or infants included in this review

Trial (reference)	Population	Intervention vs comparator	Main finding
Simpson 2014 [13]	124 high-risk neonates, USA and UK	Daily whole-body emollient vs no emollient	Lower incidence of AD at 6 months (about 22% vs 43%); judged feasible and safe
Horimukai 2014 [14]	118 high-risk neonates, Japan	Daily emulsion-type moisturiser for 32 weeks vs petroleum jelly as needed	About 32% fewer cases of AD or eczema by week 32 (p = 0.012)
BEEP, Chalmers 2020 [15]	1,394 high-risk newborns, England	Daily emollient for the first year plus standard care vs standard care	No reduction in eczema at 2 years; some evidence of more skin infections
PreventADALL, Skjerven 2020 [16]	2,397 unselected newborns, Norway and Sweden	Bath additives and facial cream, early complementary feeding, both, or neither	No reduction in AD at 12 months with either intervention
Paller 2016 [17]	Patients aged 2 years and over with mild-to-moderate AD (two phase 3 trials)	Crisaborole 2% ointment twice daily vs vehicle for 28 days	Earlier treatment success and itch relief than vehicle
Papp 2021 [18]	Patients aged 12 years and over, TRuE-AD1 (n = 631) and TRuE-AD2 (n = 618)	Ruxolitinib cream 0.75% or 1.5% vs vehicle for 8 weeks	IGA success 39.0% to 53.8% vs 7.6% to 15.1% with vehicle
Paller 2020 [20]	367 children aged 6 to 11 years with severe AD	Dupilumab plus medium-potency corticosteroid vs placebo plus corticosteroid, 16 weeks	Significant improvement in signs, symptoms and quality of life; more conjunctivitis
Paller 2022 [21]	162 children aged 6 months to under 6 years	Dupilumab plus low-potency corticosteroid vs placebo plus corticosteroid, 16 weeks	IGA 0 to 1 in 28% vs 4%; EASI-75 in 53% vs 11%

Abbreviations: AD, atopic dermatitis; EASI-75, at least 75% improvement in the Eczema Area and Severity Index; IGA, Investigator’s Global Assessment.

Source: Prepared by the authors.