Atopic dermatitis

Meet the patient

A 6-year-old boy is brought in by his exhausted mother. For two years he has scratched his antecubital and popliteal fossae until they weep, wakes three times a night, and now refuses short-sleeved shirts at school. His cheeks were a crusted mess as a baby; he wheezes in spring and sneezes around the cat.[1]

Two exam questions are now live and you must answer both: why does his skin behave this way? (barrier and immunity) and what will actually give him his nights back? (the right rung of the treatment ladder). Everything below exists to answer those two questions at consultant depth.[1][2]

The three-word handle — red, dry, itchy

Atopic dermatitis is a chronic, relapsing, inflammatory skin disease defined by pruritus, an eczematous morphology, and a typical age-dependent distribution. It is the cutaneous first stop of the atopic march, travelling in company with asthma, allergic rhinitis and food allergy.[1]

The morphology shifts with time, and each phase asks for a different potency of cream — learn the four faces:[1]

The mantra: red, dry, itchy. Red is the inflammation, dry is the broken barrier, itchy is the IL-31 driving the scratch that keeps both going. Hold those three words and the disease, the scoring, and the drugs all fall into place.[1]

How common — and where it walks next (the atopic march)

AD is the commonest inflammatory skin disease of childhood, and it rarely travels alone. It affects up to 15 to 20 percent of children and 2 to 10 percent of adults worldwide, with rising prevalence in industrialised settings; about 60 percent begin in the first year of life, and a stubborn minority persists or recurs into adulthood.[1]

The atopic march is the sequence every candidate must recite: AD in infancy, then food allergy, then asthma, then allergic rhinitis — the same barrier leak and Th2 bias expressing themselves up the respiratory tree. The skin is the gateway, which is why controlling childhood eczema is not cosmetic.[1]

The risk-factor cluster runs on autopilot — group it once and it stays:[1]

• Genetic: family history of atopy; FLG loss-of-function mutations are the strongest known genetic risk factor.[6]
• Environmental: urbanisation, hard water, low humidity, aeroallergens, tobacco smoke, occupational irritants.
• Barrier defects: reduced ceramides, raised pH, tight-junction loss — more transepidermal water loss, more allergen entry.[7]
• Microbiome: Staphylococcus aureus colonises up to 90 percent of flares and tracks with severity.[2]

Why the barrier fails — filaggrin and the Th2 fire

AD is one disease with two engines: a leaky barrier and a type 2 inflammatory fire — and each fuels the other. Barrier failure lets allergen and bug in; the immune response then damages the barrier further. Break either loop and the disease improves.[1]

Filaggrin is the word that scores. It aggregates keratin filaments and is proteolysed into the natural moisturising factors that hold water in the stratum corneum. FLG loss-of-function mutations — or Th2-driven downregulation of filaggrin and the tight junctions — raise transepidermal water loss and open the door to allergen and microbe.[6][7]

Once the barrier leaks, keratinocytes sound the alarm, releasing TSLP, IL-25 and IL-33, which drive Th2 differentiation. The cytokines that follow each do a distinct job — name them by function, not order:[1]

• IL-4 and IL-13 — the central drivers: barrier dysfunction, IgE class switching, eosinophil recruitment, fibrosis.
• IL-31 — the itch cytokine, firing on peripheral nerves to drive the scratch you cannot stop.
• IL-5 and IL-9 — eosinophil and mast-cell survival.
• IL-22 and IL-17 — epidermal hyperplasia in chronic, lichenified lesions.[1]

Etymology for viva gold: filaggrin — "filament-aggregating protein". The name tells you its job; lose it and the keratin scaffolding collapses, the skin dries, and the march begins.[6]

The itch–scratch cycle — the engine of chronicity

Itch is not a symptom of AD; it is the mechanism that maintains it. Scratching damages the already-fragile barrier, drives S. aureus colonisation, releases more TSLP and IL-31, and lichenifies the skin — which itches more. Break the cycle and chronic eczema improves; feed it and no cream will keep up.[1]

This is why sedating antihistamines earn their place at night (for sleep, not for the itch per se), why emollients are foundation not afterthought, and why the itch-dominant phenotype is the one that most rewards an IL-31 blocker.[2]

Meet the patient at each age — distribution by age

Distribution changes with age, and the change is the single most examined fact about the clinical pattern. One line carries the marks: extensors and face in the infant, flexures in the child, hands and head-and-neck in the adult.[1]

The classic trap: a baby with crusted, weeping cheeks is not "just baby acne" or impetigo — the atopic distribution plus the family history makes AD the leading diagnosis until seborrhoeic dermatitis and impetigo are excluded.[1]

The secondary changes tell you how long it has gone on: excoriations, lichen simplex chronicus, prurigo nodules, and post-inflammatory hypo- or hyperpigmentation. The pigment change can distress patients more than the itch — name it and reassure.[1]

How severe? The scoring alphabet

You cannot escalate to a biologic without a number, and the number is EASI. Severity scoring separates "a bit dry" from "biologic-eligible", and every modern AD trial is built on the same handful of instruments.[1]

[1] [2]

Know the calculation, not just the name. Examiners ask what goes into each score, and the components are where the marks live:[1]

The thresholds that unlock a biologic

Four numbers gate the door to systemic therapy, and a fellowship candidate names them as a set. The AAD 2024 systemic guideline and the UK NICE TA814 framework converge on broadly comparable entry thresholds: EASI at least 16, IGA at least 3, DLQI at least 11 (or POEM at least 16), and BSA over 10%, after failure of optimised topical therapy for 4 to 8 weeks.[2][5]

The pivotal trials enrolled on these same numbers — dupilumab, upadacitinib and abrocitinib on EASI at least 16, IGA at least 3 and at least 10% BSA, with DLQI/POEM corroborating the patient burden; tralokinumab ECZTRA 1 and 2 used IGA at least 3 and EASI at least 16 with BSA at least 10%.[2]

Treat to a target, not to a prescription. Aim for EASI-75 or IGA 0/1 by week 16; if it is not met, escalate or switch rather than drift. A patient stuck at EASI-50 at week 16 is not "nearly there" — they need the next rung.[5]

The foundation — emollients, triggers, and the bath you did not think of

Every AD patient, however severe, starts and stays on emollients. They are foundation, not adjunct — applied liberally and frequently to all skin including unaffected areas, used as a soap substitute, and continued when the skin looks clear. For an adult, think 250 to 500 g per week as the working volume.[1]

The non-drug bundle that actually changes outcomes:[3]

• Emollients — liberally and frequently, all skin, including as soap substitute and bath additive.
• Trigger avoidance — fragrances, detergents, wool, overheating, aeroallergens, and the stressors that light the flare.
• Infection control — dilute bleach baths (0.005%) twice weekly cut S. aureus burden in recurrent superinfection; treat secondary infection promptly.
• Psychosocial support — sleep, anxiety, depression, school and work impact are part of the disease, not extras.[3]

The topical ladder — match potency to site

Topical anti-inflammatory therapy is the cornerstone, and the recurring trainee error is using the wrong potency on the wrong skin. The rule is simple and load-bearing: mild for face and flexures, moderate for trunk and limbs, potent only for thick plaques on palms and soles — and never the face with a potent steroid.[3]

[3]

The potency ladder, named as a memory device:[1]

• Face, eyelids, genitals, skin folds: mild (hydrocortisone 1%) — atrophy is irreversible.
• Trunk and limbs: moderate (e.g. betamethasone valerate 0.025%, clobetasone butyrate 0.05%) for moderate disease.
• Thick, lichenified plaques on palms, soles or limbs: potent to very potent (clobetasol propionate 0.05%, mometasone furoate 0.1%) for short courses only.
• Quantify with the fingertip unit — one fingertip covers two adult palms — so you prescribe a quantity, not a tube.[1]

Everyone forgets: topical calcineurin inhibitors do not cause skin atrophy, which is exactly why they own the face, eyelids and flexures for maintenance. The transient stinging on application is the reason patients stop them in week one — warn, and they stay on.[3]

Phototherapy — the middle rung

Narrowband UVB is the workhorse phototherapy for moderate-to-severe or widespread AD that has not answered to topicals. Typical regimens run two to three times weekly for 8 to 12 weeks with dose titration by skin type or minimal erythema dose. PUVA is second-line for refractory disease but carries greater long-term carcinogenic risk.[1]

Phototherapy bridges beautifully to slower systemic agents and combines with topical therapy — but it is a finite course, not a forever plan, because of cumulative UV dose.[1]

Systemic therapy — when topicals and light are not enough

Systemic therapy is for moderate-to-severe disease refractory to optimised topical therapy and phototherapy, or when quality of life has collapsed. The choice is no longer "steroids or nothing" — it is a deliberate ladder from conventional immunosuppressants to targeted biologics and JAK inhibitors.[1]

The AAD 2024 systemic guideline gives conditional recommendations for phototherapy, azathioprine, ciclosporin, methotrexate and mycophenolate, and recommends against systemic corticosteroids because of rebound flares and adverse effects. The 2025 focused update adds strong recommendations for dupilumab, tralokinumab, abrocitinib, baricitinib and upadacitinib, and for lebrikizumab, nemolizumab (with topical therapy), tapinarof and roflumilast creams.[5]

The whole spine, as one ladder:[1]

[1]

The conventional systemics are steroid-sparing workhorses, chosen by speed and safety profile. Ciclosporin is fastest but nephrotoxic, so cap continuous use at one year; methotrexate, azathioprine and mycophenolate are slower but durable, each with its own monitoring — azathioprine dosed by TPMT, mycophenolate watched for cytopenias, methotrexate with folic acid rescue.[1]

Systemic corticosteroids have no home in chronic AD. The AAD recommends against them: the rebound flare after a short oral course is often worse than the disease it was meant to calm. Reserve a brief course only for an exceptional acute flare with a clear exit plan.[1]

[1]

The three targeted classes, faced off — pick by mechanism, speed, and the eye:[2]

[1]

The quick-reference table for the viva — drug, target, and the trial that named it:[5]

[2] [5]

Before the first dose, screen. Hepatitis B surface antigen and core antibody, hepatitis C antibody, HIV, and TB (IGRA or Mantoux) per local guideline; bring inactivated vaccines up to date, because live vaccines are contraindicated once immunosuppression begins; and draw baseline FBC, renal and liver function, creatinine, LFT, lipids for JAK inhibitors, and a pregnancy test.[1]

The drug doses — named agents at named doses

The protocols below reflect standard induction and maintenance dosing used in pivotal trials and current AAD 2024-2025 guidelines. Local formularies and product labels may differ — individualise, and prescribe from local product information.[1]

[1] [1] [1] [1] [1] [1] [1] [1] [1] [1] [1] [1] [1]

The trials that changed practice

Four landmark programmes built the modern systemic ladder. Each moved a class from "interesting" to "guideline-strong", and a fellowship candidate names the trial with the drug.[1]

[8] [9] [10] [11]

The differential — discriminators, not lists

AD mimics hide in four patterns, and the discriminator is what earns the mark. Run the pattern, then the one-line tell.[1]

[1]

The one-line discriminator: well-demarcated with silvery scale on extensors is psoriasis; poorly demarcated and itchy in flexures is AD. Everything else refines around those two poles.[1]

Investigations — clinical first, then exclude

The diagnosis is clinical — investigations exist to exclude mimics, grade severity, and clear the patient for systemic therapy, never to confirm AD. Reach for them with a question already in mind.[1]

• Skin swab or culture if secondary bacterial (S. aureus) or viral (HSV) infection is suspected.
• KOH preparation to exclude tinea corporis in annular lesions.
• Patch testing if contact dermatitis is plausible — especially adults with hand or refractory disease.
• Total and specific IgE (RAST) may support atopic status but are not diagnostic.
• Biopsy if the diagnosis is uncertain, features are atypical, or treatment is failing.
• Baseline bloods before systemic immunosuppression — FBC, renal and liver function, creatinine, LFT, hepatitis B and C, HIV, TB screening (IGRA or Mantoux), pregnancy test where relevant.[1]

Histology is rarely needed, but know it: spongiosis (intercellular oedema) with epidermal thickening in chronic lesions, parakeratosis, and a superficial perivascular lymphocytic infiltrate with eosinophils. Unlike psoriasis, there are no regular acanthosis, no Munro microabscesses, no neutrophilic collections.[1]

Comorbidities — the atopic body and mind

AD is a systemic disease, and the comorbidities are part of the assessment, not a footnote. The AAD 2022 comorbidity guideline catalogues associations across allergic, atopic, immune-mediated, mental-health and bone-health domains, plus skin infections.[4]

Screen actively for asthma, allergic rhinitis and food allergy (the atopic march), and — just as importantly — for anxiety, depression and the sleep deprivation that chronic itch inflicts. A patient whose eczema is "controlled" but who is still not sleeping is not controlled.[4]

Special sites and populations

Site and life-stage change the first choice and the monitoring — state them explicitly.[3]

• Children: emollients and mild-to-moderate topical corticosteroids first-line; topical calcineurin inhibitors for facial and flexural maintenance; avoid potent steroids on face and folds; phototherapy and systemic therapy only under specialist care.
• Pregnancy and breastfeeding: emollients and mild topical corticosteroids are generally safe; avoid most systemic immunosuppressants and biologics unless specialist-directed; ciclosporin may be used in selected cases under specialist care.
• Hand and foot eczema: potent topical steroids under occlusion for thick plaques; patch test if an occupational or contact aetiology is suspected.
• Eyelid eczema: low-potency steroids or topical calcineurin inhibitors; avoid chronic potent steroid use because of glaucoma and cataract risk.[3]

Eczema herpeticum — the emergency

This is the complication that must never be missed, and there is nothing funny about it. A child or adult with AD who develops widespread, monomorphic, punched-out erosions with haemorrhagic crusting, fever, malaise, or a history of cold-sore contact has eczema herpeticum — a dermatology emergency caused by herpes simplex virus disseminating through broken barrier skin.[1]

Treat promptly with systemic aciclovir, and consider admission, especially in the young, the extensive, or the systemically unwell. Do not wait for the swab to return in a deteriorating patient — the clinical picture is enough to start.[1]

Secondary bacterial infection presents with weeping, crusting, pustules and rapid flare — S. aureus is the usual culprit; treat with antiseptics or antibiotics as appropriate. Molluscum contagiosum runs more extensive and persistent courses in AD skin.[1]

Generalised erythroderma — the whole skin inflamed and scaling — is an admission: the patient loses fluid, electrolytes and temperature control through the broken barrier, and needs inpatient stabilisation.[1]

[1]

How AD patients come to harm — the preventable list

• Eczema herpeticum missed as "a flare" — the delayed-antiviral harm.[1]
• A potent steroid used on the face or eyelids long-term, ending in skin atrophy, striae, glaucoma or cataract.[1]
• Systemic corticosteroids prescribed for chronic AD, ending in a rebound flare worse than the original.[1]
• A biologic or JAK inhibitor started without TB, hepatitis and vaccination screening.[1]
• Ciclosporin continued past a year without creatinine monitoring, ending in nephrotoxicity.[1]
• Azathioprine dosed without a TPMT result, ending in profound myelosuppression.[1]
• A JAK inhibitor given in pregnancy, or to a patient needing live vaccines.[1]
• An itchy, sleepless child labelled "just eczema" while depression and school refusal take hold.[4]

Prognosis and follow-up

AD is chronic and relapsing; the goal is control, not cure. Modern targeted therapies achieve high rates of EASI-75 or IGA 0/1, but relapse is the rule after stopping — so maintenance planning is part of the prescription, not an afterthought.[1]

Follow-up intervals track the therapy and severity:[1]

• Topical therapy: review at 4 to 8 weeks.
• Phototherapy: review every 2 to 4 weeks during induction.
• Systemic or biologic therapy: review at 8 to 16 weeks initially, then every 3 to 6 months, with laboratory monitoring as appropriate.[1]

[1] [2]

Guidelines and regional deltas

• AAD 2023 topical guideline — emollients, topical corticosteroids, calcineurin inhibitors, crisaborole and ruxolitinib.[3]
• AAD 2024 systemic and phototherapy guideline — strong recommendations for dupilumab, tralokinumab, abrocitinib, baricitinib and upadacitinib; conditional for phototherapy, ciclosporin, methotrexate, azathioprine and mycophenolate; recommends against systemic corticosteroids.[1]
• AAD 2025 focused update — adds strong recommendations for tapinarof cream, roflumilast cream, lebrikizumab and nemolizumab with topical therapy.[5]
• NICE CG57 — UK guidance for atopic eczema in children under 12 years: emollients, stepped topical corticosteroids, calcineurin inhibitors as second-line.
• EuroGuiDerm and IADVL — regional variations in phototherapy access, biologic eligibility and monitoring.[1]

The mantra, and the mnemonic

[1]

The mantra: emollients always, steroid the flare by site, calm the Th2 fire when the scores say so — and never miss eczema herpeticum.[1][2]

[1] [1]

Ward-round test — three stems, thirty seconds each

Stem 1 — the itchy child whose cream is on the wrong skin (answer)

A mother has been applying clobetasol propionate 0.05% to her 4-year-old's cheeks and eyelids for three months for "stubborn eczema". The cheeks are now thin and telangiectatic. What went wrong, and what do you do? Model: A potency–site mismatch. Clobetasol is a Class I ultra-potent steroid; the face and eyelids take only mild potency (hydrocortisone 1%) or — better still for maintenance — a topical calcineurin inhibitor such as tacrolimus 0.1% ointment, which is non-atrophogenic and owns the face and flexures. Stop the clobetasol immediately, switch to a calcineurin inhibitor plus emollients, screen for glaucoma and cataract from the periocular exposure, and educate on the potency ladder: mild for face and flexures, moderate for trunk and limbs, potent only for thick plaques on palms and soles.[1][3]

Stem 2 — punched-out erosions and a fever (answer)

A 7-year-old with known AD presents with a rapid flare, widespread monomorphic punched-out erosions with haemorrhagic crusting, fever and malaise. His sibling has a cold sore. What is this, and what is the first action? Model: Eczema herpeticum — herpes simplex virus disseminating through broken barrier skin, a dermatology emergency. Do not wait for the swab in a deteriorating child. Start systemic aciclovir immediately, consider admission (young, extensive, systemically unwell), and screen for and treat any secondary S. aureus infection. This is the complication that must never be missed — the monomorphic punched-out morphology against fever and cold-sore contact is the tell.[1]

Stem 3 — is this patient biologic-eligible? (answer)

A 28-year-old has AD despite optimised topical therapy and narrowband UVB for the last 8 weeks. Her EASI is 17, IGA 3, DLQI 13, BSA 14%. What is the threshold question, and what is first-line? Model: She meets systemic-therapy thresholds — EASI at least 16, IGA at least 3, DLQI at least 11, BSA over 10% — after documented failure of optimised topicals. Before the first dose, screen TB (IGRA), hepatitis B and C, HIV, and vaccinations, because live vaccines are contraindicated on biologics. Dupilumab (600 mg loading then 300 mg every other week) is first-line — the broadest type-2 blockade with the longest safety record; an anti-IL-13 (tralokinumab or lebrikizumab) is the choice if conjunctivitis is a concern. Treat to EASI-75 or IGA 0/1 by week 16; if not met, switch or escalate.[2][5][8]

References

1. [1]Langan SM, Irvine AD, Weidinger S. Atopic dermatitis Lancet, 2020.PMID 32738956
2. [2]Guttman-Yassky E, Renert-Yuval Y, Brunner PM Atopic dermatitis Lancet, 2025.PMID 39955121
3. [3]Sidbury R, Davis DMR, Cohen DE, et al. Guidelines of care for the management of atopic dermatitis in adults with topical therapies J Am Acad Dermatol, 2023.PMID 36641009
4. [4]Davis DMR, Drucker AM, Alikhan A, et al. American Academy of Dermatology Guidelines: Awareness of comorbidities associated with atopic dermatitis in adults J Am Acad Dermatol, 2022.PMID 35085682
5. [5]Davis DMR, Frazer-Green L, Alikhan A, et al. Focused update: Guidelines of care for the management of atopic dermatitis in adults J Am Acad Dermatol, 2025.PMID 40531067
6. [6]Drislane C, Irvine AD. The role of filaggrin in atopic dermatitis and allergic disease Ann Allergy Asthma Immunol, 2020.PMID 31622670
7. [7]Margolis DJ Atopic dermatitis: filaggrin and skin barrier dysfunction Br J Dermatol, 2022.PMID 35128630
8. [8]Simpson EL, Bieber T, Guttman-Yassky E, et al. Two Phase 3 Trials of Dupilumab versus Placebo in Atopic Dermatitis N Engl J Med, 2016.PMID 27690741
9. [9]Guttman-Yassky E, Teixeira HD, Simpson EL, et al. Once-daily upadacitinib versus placebo in adolescents and adults with moderate-to-severe atopic dermatitis (Measure Up 1 and Measure Up 2): results from two replicate double-blind, randomised controlled phase 3 trials Lancet, 2021.PMID 34023008
10. [10]Wollenberg A, Blauvelt A, Guttman-Yassky E, et al. Tralokinumab for moderate-to-severe atopic dermatitis: results from two 52-week, randomized, double-blind, multicentre, placebo-controlled phase III trials (ECZTRA 1 and ECZTRA 2) Br J Dermatol, 2021.PMID 33000465
11. [11]Silverberg JI, Guttman-Yassky E, Thaçi D, et al. Two Phase 3 Trials of Lebrikizumab for Moderate-to-Severe Atopic Dermatitis N Engl J Med, 2023.PMID 36920778