Stevens-Johnson syndrome and toxic epidermal necrolysis

Contents


Classification

Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are severe, life-threatening mucocutaneous drug reactions characterised by epidermal detachment.

They represent a spectrum of the same condition, classified by the percentage of total body surface area (TBSA) with epidermal detachment:

Condition Detached/blistered BSA
SJS <10% TBSA
SJS-TEN overlap 10–30% TBSA
TEN >30% TBSA

TEN is distinct from staphylococcal scalded skin syndrome (SSSS — caused by exfoliative toxin, not drug hypersensitivity, primarily affects children, Nikolsky sign in normal skin, much lower mortality) and from toxic shock syndrome.


Causative Agents

SJS/TEN is almost always drug-induced. The most common causative agents:

High-Risk Drugs

Drug class Examples
Aromatic anticonvulsants Carbamazepine, lamotrigine, phenytoin, phenobarbitone — most common cause overall
Sulfonamide antibiotics Co-trimoxazole, sulfasalazine
Allopurinol Particularly in patients with HLA-B*5801 (South-East Asian populations)
NSAIDs (oxicam class) Piroxicam, meloxicam
Nevirapine (antiretroviral) Common in HIV-positive patients on HAART
Vancomycin Less common

HLA Associations

  • HLA-B*5701: carbamazepine-induced SJS/TEN (predominantly in Han Chinese, Thai, and other South-East Asian populations)
  • HLA-B*5801: allopurinol-induced SJS/TEN — pre-treatment HLA screening in high-risk populations now recommended
  • These associations underlie pharmacogenomics-guided prescribing in international guidelines

Latency

Typically develops 4–28 days after starting the causative drug. Latency can be longer with allopurinol (up to 8 weeks).


Pathophysiology

SJS/TEN is a cytotoxic T-lymphocyte (CTL)-mediated hypersensitivity reaction:

  1. Drug processing: the causative drug or its metabolite binds to HLA molecules on antigen-presenting cells, forming a drug-HLA-T-cell receptor (TCR) complex
  2. CD8+ CTL activation: drug-specific CD8+ T cells are activated; infiltrate the skin
  3. Keratinocyte apoptosis: activated CTLs release perforin, granzyme B, and FasL (CD95L) → keratinocyte apoptosis and epidermal detachment
  4. Granulysin: cytotoxic protein released by CTLs; identified as the key effector molecule in TEN — serum granulysin levels correlate with disease severity

The result is widespread keratinocyte death → epidermal sheet loss → denuded dermis resembling a severe burn.


Clinical Features

Prodrome (1–3 days before skin eruption)

  • Fever, malaise, sore throat, conjunctivitis
  • Burning sensation of skin
  • May be mistaken for viral illness or drug rash

Mucosal Involvement (essential feature)

  • Always present — distinguishes SJS/TEN from other severe drug reactions
  • Oral mucosa: painful erosions, pseudomembranes, inability to eat/drink
  • Ocular: conjunctival injection, erosions, pseudomembrane formation → risk of symblepharon, corneal scarring, blindness
  • Genital: erosions, urethritis

Skin Involvement

  • Nikolsky sign positive: tangential pressure on skin causes epidermal shearing/blister formation (reflects epidermal-dermal detachment)
  • Erythematous/dusky macules → blistering → epidermal detachment in sheets
  • Denuded dermis resembles a partial-thickness burn
  • Distribution: often starts on trunk, spreads to extremities; mucosae always involved

Systemic Complications

  • Respiratory: erosions of bronchial mucosa → airway compromise, pneumonia; worst prognosis complication
  • Hepatitis, nephritis, GI tract involvement
  • Fluid and electrolyte losses (as per burns — large fluid losses from denuded skin)

SCORTEN Score

SCORTEN (SCORe of Toxic Epidermal Necrolysis) is validated for TEN prognosis:

Variable Score
Age >40 years 1
Malignancy 1
Tachycardia >120 bpm 1
Epidermal detachment >10% BSA on Day 1 1
BUN >10 mmol/L 1
Blood glucose >14 mmol/L 1
Bicarbonate <20 mmol/L 1
SCORTEN Predicted mortality
0–1 3.2%
2 12.1%
3 35.3%
4 58.3%
≥5 >90%

ICU Management

1. Stop the Causative Drug Immediately

Every day of continued exposure increases mortality. Identify and stop the suspected agent urgently — earlier withdrawal correlates with better outcomes.

2. Specialist Referral

Immediate transfer to a specialist centre (burns unit with dermatological expertise or vice versa): best outcomes achieved in centres managing >3 cases/year.

3. Wound Care — Burns Model

  • Denuded skin managed as for partial-thickness burns: non-adhesive dressings, sterile or paraffin-based coverings
  • Avoid silver sulfadiazine (may be causative or co-contributory); use non-adherent dressings (Mepitel, Urgotul)
  • Strict infection control; barrier nursing; temperature control (patients are poikilothermic from skin loss)
  • Do NOT debride — let epidermis shed naturally

4. Fluid and Nutritional Support

  • Fluid requirements as for burns: Parkland-type formula (less aggressive than burns — skin barrier partly intact in SJS)
  • Titrate to urine output 0.5–1 mL/kg/h; daily weights; close electrolyte monitoring
  • Early enteral nutrition (ideally within 24–48h) — high protein requirements for skin healing

5. Eye Care

  • Ophthalmology involvement urgently — ocular involvement can lead to permanent blindness
  • Topical lubricants, topical antibiotics, topical corticosteroids (conjunctival)
  • Lysis of synechiae (adhesions between conjunctiva and cornea) may be required
  • Daily ophthalmological review in active disease

6. Analgesia and Sedation

  • Wound care is extremely painful; procedural sedation/analgesia
  • Oral analgesia limited by mucositis — IV route required
  • Opioid infusion + ketamine for dressing changes

7. Infection Surveillance

  • Regular wound swabs and blood cultures
  • Prophylactic antibiotics are NOT recommended — select resistant organisms
  • Treat proven infection promptly; Staphylococcus aureus and Gram-negative bacteria predominant

8. Airway and Respiratory

  • Monitor for respiratory mucosal involvement — hoarse voice, stridor, worsening gas exchange
  • Early intubation if airway compromise suspected — delay is dangerous as oedema can progress
  • Avoid nasogastric/nasotracheal routes where possible (nasal mucosal involvement); oral or surgical airway

Specific Therapies

There is no high-quality RCT evidence for any specific immunomodulatory therapy in SJS/TEN. The following are used based on observational data and pathophysiological rationale:

Corticosteroids

  • Controversial: early observational data suggested harm (increased infection, delayed healing); subsequent studies mixed
  • Some evidence for benefit in SJS (less extensive disease); harm/neutral in TEN
  • Not standard of care; used in some European centres for SJS in the first few days

Intravenous Immunoglobulin (IVIG)

  • Rationale: contains anti-FasL antibodies → blocks Fas-mediated keratinocyte apoptosis
  • Initial enthusiasm from early case series; subsequent larger observational studies and a meta-analysis have not confirmed benefit
  • Not recommended as standard therapy; used in some centres in severe TEN

Ciclosporin

  • Inhibits T-cell activation (calcineurin inhibitor → blocks IL-2 production)
  • Several case series and a small RCT suggest possible mortality benefit in TEN — faster re-epithelialisation
  • Growing evidence base; increasingly used in specialist centres; dose 3–5 mg/kg/day

TNF-α Inhibitors (Etanercept, infliximab)

  • Case reports and series suggest benefit; etanercept studied in small RCTs
  • Not standard; considered in refractory or severe TEN at specialist centres

Viva Questions

1. How do you distinguish TEN from staphylococcal scalded skin syndrome (SSSS) and why does it matter?

Both present with epidermal detachment and a positive Nikolsky sign, but they are completely different diseases. TEN is a drug-induced cytotoxic T-cell reaction causing full-thickness epidermal necrosis — the split is at the dermoepidermal junction; skin histology shows full-thickness epidermal necrosis with dermal inflammatory infiltrate. SSSS is caused by Staphylococcal exfoliative toxin (ET-A and ET-B) — serine proteases that cleave desmoglein-1 in the superficial epidermis (stratum granulosum). The split is high within the epidermis — denuded areas are superficial only. SSSS primarily affects children and immunocompromised adults; mortality is much lower in children (<5%) and higher in adults (~50% — immune clearance of toxin is impaired). Key distinguishing features: TEN shows mucosal involvement (oral, ocular, genital — always present), and the biopsy shows full-thickness epidermal necrosis; SSSS has no mucosal involvement and the biopsy split is superficial. Treatment also differs: TEN — stop drug, supportive care; SSSS — anti-staphylococcal antibiotics (flucloxacillin). The distinction matters because treating SSSS as TEN (steroids) or missing TEN drug cause is dangerous.


2. What is the SCORTEN score and how do you use it in clinical practice?

SCORTEN is a validated prognostic scoring tool for TEN, incorporating seven clinical variables: age >40, presence of malignancy, tachycardia >120 bpm, initial BSA detachment >10%, elevated BUN, hyperglycaemia, and acidosis (bicarbonate <20 mmol/L). Each variable scores 1 point. SCORTEN 0–1 predicts ~3% mortality; SCORTEN ≥5 predicts >90% mortality. It is calculated on Day 1 and Day 3 — Day 3 is more predictive of outcome than Day 1 as it reflects the trajectory of disease. In clinical practice, SCORTEN guides: the urgency of transfer to specialist centre; prognosis discussions with patients and families (bearing in mind scores are population-level probabilities, not individual predictions); and ITU resource planning. A SCORTEN ≥3 generally warrants HDU/ICU-level care. Importantly, SCORTEN was validated for TEN and SJS-TEN overlap — its discriminatory ability in pure SJS is less established.


3. A patient on carbamazepine develops SJS with 8% BSA epidermal detachment, oral ulceration, and bilateral conjunctivitis. What are your immediate management priorities?

Immediate priorities: (1) Stop carbamazepine immediately — the single most important intervention; every additional day of drug exposure worsens mortality; consult neurology about seizure risk and alternative agents urgently; (2) Specialist referral — contact burns unit and dermatology immediately for co-management; best outcomes in specialist centres; (3) Ophthalmology — bilateral conjunctivitis in SJS is a medical emergency; immediate ophthalmological assessment; topical lubricants, antibiotics, and corticosteroid eye drops; risk of permanent corneal scarring; (4) Wound care — nurse on non-adherent dressings; Mepitel or Urgotul; do not debride; strict infection control; (5) IV access and fluids — assess hydration; oral fluid intake limited by mucositis; IV fluid replacement; (6) Analgesia — IV opioids if mucositis prevents oral route; (7) Nutrition — NG tube early (nasogastric route may be limited by oral/nasal mucosal involvement — assess feasibility); (8) Monitor for progression — measure BSA involvement daily; SCORTEN score; watch for respiratory mucosal involvement (stridor, worsening oxygenation) — have low threshold to intubate early; (9) No prophylactic antibiotics; swab wounds regularly; (10) Consider ciclosporin in discussion with dermatology.