Stevens-Johnson Syndrome

Overview

Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are two ends of a single, life-threatening spectrum of mucocutaneous drug reactions characterised by widespread keratinocyte death and full-thickness epidermal detachment.

Aetiology

Symptoms begin 1-8 weeks after starting the culprit drug. Key culprits include lamotrigine, carbamazepine, phenytoin, allopurinol, sulfonamides, NSAIDs, and nevirapine. Non-drug triggers include Mycoplasma pneumoniae and HSV (especially in children).

💡

Screen for HLA-B*15:02 before starting carbamazepine in patients of Han Chinese or Thai descent - carriers have dramatically elevated SJS/TEN risk (NICE-recommended).

Classification

Classified by % total body surface area (BSA) showing epidermal detachment:

•SJS - <10% BSA; mortality ~5-10%

•SJS/TEN overlap - 10-30% BSA

•TEN - >30% BSA; mortality ~25-35%

Presentation

•Prodrome (days 1-3): fever, malaise, sore eyes, odynophagia, burning skin pain - before visible rash

•Cutaneous: dusky erythematous patches beginning centrally (face, trunk); flaccid blisters; epidermal sheets peel leaving raw, oozing dermis

•Nikolsky sign positive - lateral pressure on normal-appearing skin causes epidermal shearing

•Mucosal erosions - at least two sites (oral, ocular, genital, respiratory); haemorrhagic crusting of lips

•Ocular - conjunctival injection, photophobia; occurs in up to 80%

🎯

Mucosal involvement at ≥2 sites distinguishes SJS/TEN from erythema multiforme major (EM). EM has true three-zone target lesions, acral distribution, is infection-triggered, and carries far lower mortality. SJS/TEN has atypical targets, central distribution, and is drug-triggered.

Investigations

•Skin biopsy (perilesional) - confirms full-thickness epidermal necrosis and dermo-epidermal separation; distinguishes from other blistering disorders

•FBC, U&E, LFTs, CRP, glucose - baseline; lymphopenia and elevated CRP typical; hyponatraemia is a SCORTEN variable

•Blood cultures and wound swabs - identify secondary infection (leading cause of mortality)

•SCORTEN scoring - calculate day 1 and day 3 (day 3 is stronger predictor)

•Ophthalmic slit-lamp examination - essential even with mild eye symptoms; subconjunctival fibrosis begins early

Differential diagnosis

Management

•Immediate drug withdrawal - stop the culprit drug as soon as SJS/TEN is suspected

•Admission - all SJS patients admitted; TEN/overlap managed in burns unit or equivalent specialist facility

•Supportive care - IV fluid resuscitation (mirrors burns management), temperature regulation, nutritional support, non-adherent wound dressings

•Ophthalmology review - urgent, even with mild eye symptoms

•Analgesia - skin pain is severe

•Infection surveillance - blood cultures and wound swabs; targeted antibiotics only with microbiological evidence

⚠️

Do NOT use prophylactic antibiotics routinely - drives resistant organisms without improving outcomes. Do NOT routinely use systemic corticosteroids in established TEN - risk of immunosuppression and sepsis outweighs unproven benefit.

Complications

•Sepsis and bacteraemia - leading cause of death; S. aureus and Pseudomonas aeruginosa are key pathogens

•Acute respiratory failure - bronchial epithelial sloughing

•Hypovolaemia and electrolyte imbalance - mirrors major burn

•Ocular sequelae (most common long-term complication) - symblepharon, dry eye, corneal scarring, visual impairment in up to 35% of survivors

•GI/urogenital strictures - oesophageal strictures, vaginal adhesions, urethral strictures

Prognosis

•The causative drug must be permanently avoided and clearly documented in all medical records; the patient should carry a medication alert card

•Family members should be counselled - pharmacogenomic risk can be shared

•Structured follow-up required with ophthalmology, dermatology, and relevant specialties

SCORTEN

Each variable = 1 point. Calculate on day 1 and repeat day 3.

•Age >40 years

•Heart rate >120 bpm

•BSA detachment >10% on day 1

•Serum urea >10 mmol/L

•Serum glucose >14 mmol/L

•Serum bicarbonate <20 mmol/L

•Presence of malignancy

SCORTEN	Estimated mortality
0-1	~3%
2	~12%
3	~35%
4	~58%
≥5	>90%