Stevens-Johnson Syndrome
A severe, immune-mediated mucocutaneous reaction characterised by widespread epidermal necrosis and mucosal erosions. SJS involves <10% BSA detachment; overlap SJS-TEN 10–30%; TEN >30%. Mortality: SJS ~5%, TEN ~25–30%. Most commonly drug-induced (allopurinol, anticonvulsants, antibiotics). Requires immediate drug withdrawal, supportive care in a burns unit, and consideration of ciclosporin.
Key Facts
SJS <10% BSA detachment; SJS-TEN overlap 10–30%; TEN >30%: spectrum of the same disease Mortality: SJS ~5%, TEN ~25–30%; SCORTEN score predicts mortality Drug-induced (>80%): allopurinol, carbamazepine, lamotrigine, phenytoin, co-trimoxazole, nevirapine Mucosal involvement: oral, ocular, genital erosions — present in >90% of cases Nikolsky sign positive: lateral pressure on skin → epidermal detachment SCORTEN score: predicts mortality — age >40, malignancy, heart rate >120, BSA >10%, urea >10, glucose >14, bicarbonate <20 Immediate drug withdrawal: cornerstone of management — delay increases mortality Ciclosporin 3–5 mg/kg/day: emerging evidence as beneficial treatment (reduce disease progression)
Overview
Key Facts
SJS and TEN represent a spectrum of severe cutaneous adverse reactions (SCARs) characterised by epidermal necrosis and detachment. They are dermatological emergencies requiring immediate drug withdrawal and multidisciplinary care.
Epidemiology
- Incidence: SJS ~1–6 per million/year; TEN ~0.4–1.2 per million/year
- All ages; increased risk with HIV (100× higher incidence)
- Mortality: SJS ~5%, TEN ~25–30%
Aetiology
- Drugs (>80%): onset typically 1–3 weeks after drug initiation
- Allopurinol (most common single cause)
- Anticonvulsants: carbamazepine, lamotrigine, phenytoin, phenobarbital
- Antibiotics: sulfonamides (co-trimoxazole), penicillins, cephalosporins, quinolones
- NSAIDs: piroxicam (oxicam class)
- Antiretrovirals: nevirapine
- Infections: Mycoplasma pneumoniae (especially in children — often SJS without clear drug cause)
- HLA associations: HLA-B5801 (allopurinol), HLA-B1502 (carbamazepine — SE Asian populations)
Pathophysiology
- CD8+ cytotoxic T-cells and NK cells → massive keratinocyte apoptosis via Fas-FasL, granulysin, and perforin/granzyme pathways
- Granulysin: key cytotoxic molecule — serum levels correlate with severity
- Full-thickness epidermal necrosis → sheet-like detachment → exposed dermis (like burn)
- Mucosal epithelial necrosis → oral, ocular, genital, respiratory tract involvement
Clinical Presentation
Prodromal Phase (1–3 Days)
- Fever >39°C, malaise, myalgia
- Sore throat, painful swallowing, stinging eyes
- May mimic upper respiratory infection
Mucocutaneous Phase
- Skin: dusky red/purple macules with central necrosis (atypical targetoid lesions)
- Positive Nikolsky sign: lateral pressure → epidermal separation
- Confluent areas of epidermal detachment — skin sloughs off in sheets
- SJS: <10% BSA; SJS-TEN: 10–30%; TEN: >30%
- Oral mucosa (90%): painful erosions of lips, buccal mucosa, tongue → difficulty eating/drinking
- Ocular (80%): conjunctival erosions, photophobia, pseudomembrane formation
- Genital (50–70%): erosions of vulva, penis, urethra → urinary retention
- Respiratory tract: cough, dyspnoea (bronchial epithelial involvement)
Red Flags
- Rapidly increasing BSA of detachment → transfer to burns unit
- Respiratory compromise → ITU
- Ocular involvement → urgent ophthalmology (risk of symblepharon, blindness)
- Sepsis from exposed dermis — leading cause of death
Differential Diagnosis
| Diagnosis | Key Features | Investigation |
|---|---|---|
| Erythema multiforme | Typical target lesions, acral, HSV trigger, NO mucosal sloughing | Clinical, HSV PCR |
| Staphylococcal scalded skin syndrome (SSSS) | Neonates/infants, Nikolsky +ve, NO mucosal involvement | Biopsy (subcorneal split) |
| Autoimmune blistering disease (PV, BP) | Chronic, different distribution | DIF biopsy |
| AGEP | Sterile pustules, no mucosal involvement, quick resolution | Biopsy |
| DRESS | Facial oedema, eosinophilia, organ involvement, no epidermal detachment | Bloods, biopsy |
| Burns | History of thermal/chemical exposure | History |
Diagnosis / Investigation
Bedside
- Nikolsky sign: positive — epidermal detachment with lateral pressure
- BSA assessment: calculate percentage of detached/detachable epidermis
- SCORTEN score: calculate within 24 hours of admission — predicts mortality
Bloods
- FBC: lymphopenia, neutropenia (common)
- U&Es: urea >10 mmol/L (SCORTEN parameter); renal function for fluid management
- LFTs: transaminitis common
- Glucose: >14 mmol/L (SCORTEN parameter)
- Bicarbonate: <20 mmol/L (SCORTEN parameter)
- Blood cultures: if sepsis suspected (exposed dermis = high infection risk)
- Mycoplasma serology/PCR: especially in children without clear drug trigger
Biopsy
- Skin biopsy: full-thickness epidermal necrosis with minimal dermal inflammation — characteristic
- Frozen section: can differentiate TEN from SSSS rapidly (subcorneal split in SSSS vs full-thickness necrosis in TEN)
Special Tests
- Ophthalmology assessment: urgent — assess for conjunctival erosion, pseudomembranes
- Chest X-ray: if respiratory symptoms
- Drug causality assessment: ALDEN algorithm — identifies causative drug by timing, known risk, rechallenge data
Management
Immediate
- STOP the causative drug immediately — single most important intervention; delay increases mortality by ~3–5% per day
- Transfer to burns unit/ITU: for TEN (>10% BSA); SJS may be managed on dermatology ward
- Calculate SCORTEN within 24 hours
Supportive Care (Burns Unit Principles)
- Fluid resuscitation: less than typical burns (~70% of Parkland formula); replace losses through exposed dermis
- Wound care: non-adherent dressings; minimise handling; leave detached epidermis in situ as biological dressing if possible
- Temperature regulation: heated room (30–32°C)
- Nutrition: high-calorie, high-protein; NG feeding if oral intake compromised
- Infection prevention: strict asepsis; avoid prophylactic antibiotics (only if signs of sepsis)
- Pain management: opioid analgesia usually required
- DVT prophylaxis: LMWH
Eye Care
- Ophthalmology review: daily in acute phase
- Preservative-free lubricating drops: hourly
- Topical corticosteroid eye drops (dexamethasone 0.1%)
- Break symblepharon: daily lysis of adhesions with glass rod
- Amniotic membrane transplant: for severe ocular involvement
Pharmacological
- Ciclosporin 3–5 mg/kg/day: increasing evidence of benefit — may halt disease progression; used in many UK centres
- Systemic corticosteroids: controversial — may be used early in disease; avoid in established TEN
- IVIg: evidence equivocal; used in some centres at 2 g/kg over 3 days
- Etanercept 50 mg SC single dose: some trial evidence (Paradisi trial)
Long-Term
- Document drug allergy comprehensively — patient, notes, GP, allergy card
- Avoid structurally related drugs: cross-reactivity (e.g. aromatic anticonvulsants)
- Ophthalmology follow-up: chronic dry eye, symblepharon, corneal scarring
- Dermatology follow-up: scarring, pigmentary changes
- Psychological support: PTSD, depression common after TEN
Referral Criteria
- Burns unit/specialist dermatology: all SJS-TEN
- Ophthalmology: all cases (even without initial eye symptoms)
- ICU: respiratory involvement, haemodynamic instability
- Psychology: long-term follow-up for psychological sequelae
Prognosis
- SJS: mortality ~5%
- SJS-TEN overlap: mortality ~10–15%
- TEN: mortality ~25–30%
- SCORTEN predicts mortality accurately (see table)
- Sepsis: leading cause of death
- Long-term complications: chronic dry eye/blindness (~20%), urogenital strictures, skin scarring, psychological trauma
- Mycoplasma-induced SJS (children): generally better prognosis
Other Relevant Information
SCORTEN Prognostic Score
| Parameter | Score |
|---|---|
| Age >40 years | +1 |
| Malignancy | +1 |
| Heart rate >120/min | +1 |
| BSA detachment >10% (day 1) | +1 |
| Serum urea >10 mmol/L | +1 |
| Serum glucose >14 mmol/L | +1 |
| Serum bicarbonate <20 mmol/L | +1 |
| Score | Predicted Mortality |
|---|---|
| 0–1 | 3% |
| 2 | 12% |
| 3 | 35% |
| 4 | 58% |
| ≥5 | >90% |
SJS vs TEN vs EM
| Feature | EM | SJS | TEN |
|---|---|---|---|
| Target lesions | Typical targets, acral | Atypical targets, trunk | Diffuse detachment |
| BSA detachment | 0% | <10% | >30% |
| Mucosal involvement | Mild or absent | Severe (≥2 sites) | Severe |
| Main trigger | HSV, Mycoplasma | Drugs | Drugs |
| Nikolsky | Negative | Positive | Positive |
| Mortality | <1% | ~5% | ~25–30% |