DermNet NZ

Facts about the skin from DermNet New Zealand Trust. Topic index: A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Stevens Johnson Syndrome & Toxic Epidermal Necrolysis

Stevens Johnson Syndrome (SJS) and Toxic Epidermal Necrolysis (TEN) are now believed to be variants of the same condition, distinct from erythema multiforme. SJS/TEN is a very rare, acute, serious, and potentially fatal skin reaction in which there is sheet-like skin and mucosal loss. Using current definitions, it is nearly always caused by medications.

Who gets SJS/TEN?

Fortunately SJS/TEN is a very rare complication of medication use (estimated to be 1-2/million each year for SJS, and 0.4-1.2/million each year for TEN). But anyone on medication can develop SJS/TEN unpredictably. It can affect all age groups, both sexes and all races. It is more common in association with human immunodeficiency virus infection (HIV), which may reflect the increased use of medications by HIV patients.

More than 200 medications have been reported in association with SJS/TEN. It is more often seen with drugs with long half-lives compared to even a chemically similar related drug with a short half-life. A half-life of a medication is the time that half of the delivered dose remains circulating in the body. The medications are usually systemic (taken by mouth or injection) but TEN has been reported after topical use.

SJS/TEN has rarely been associated with vaccination and infections such as mycoplasma. Infections are generally associated with less severe disease than when drugs are the cause.

The most commonly implicated medications are antibacterial sulfonamides.

Drugs that most commonly cause SJS/TEN
  • Sulfonamides, e.g., cotrimoxizole; beta-lactams i.e., penicillins, cephalosporins
  • Imidazole antifungals
  • Nevirapine (non-nucleoside reverse-transcriptase inhibitor)
Nonsteroidal anti-inflammatory drugs (NSAID)
(oxicam type mainly)
  • Naproxen
  • Ibuprofen
  • Carbamazepine
  • Phenytoin
  • Phenobarbital
  • Valproic acid
  • Lamotrigine

The adverse drug reaction SJS/TEN usually develops within the first week of antibiotic therapy but up to 2 months after starting an anticonvulsant. For most drugs the onset is within a few days up to 1 month.

Clinical features of SJS/TEN

Before the rash appears, there is usually a prodromal illness of several days duration resembling an upper respiratory tract infection or 'flu-like illness. Symptoms may include:

There is then an abrupt onset of a tender/painful red skin rash starting on the trunk and extending rapidly over hours to days onto the face and limbs. The maximum extent is usually reached by 4 days. The skin lesions may be:

The blisters then merge to form sheets of skin detachment, exposing red, oozing dermis. The Nikolsky sign is positive in areas of skin redness. This means that blisters appear when the skin is rubbed gently.

Toxic epidermal necrolysis Toxic epidermal necrolysis Toxic epidermal necrolysis
Skin signs in SJS/TEN

Mucosal involvement is prominent and severe, although not forming actual blisters. At least 2 mucosal surfaces are affected including:

Toxic epidermal necrolysis Stevens Johnson Syndrome Stevens Johnson Syndrome
Mucosal involvement in SJS/TEN

The patient is very ill, extremely anxious and in considerable pain. In addition to skin/mucosal involvement, other organs may be affected including liver, kidneys, lungs, bone marrow and joints.

How is the diagnosis made?

The diagnosis is suspected clinically and classified based on the skin surface area detached at maximum extent.

Classification of TEN
  • Skin detachment <10% of body surface area (BSA)
  • Widespread erythematous or purpuric macules or flat atypical targets
Overlap SJS/TEN
  • Detachment between 10% and 30% of BSA
  • Widespread purpuric macules or flat atypical targets
TEN with spots
  • Detachment >30% of BSA
  • Widespread purpuric macules or flat atypical targets
TEN without spots
  • Detachment of >10% of BSA
  • Large epidermal sheets and no purpuric macules

As these areas of skin detachment are at maximum extent, the category cannot always be defined with certainty on initial presentation. The diagnosis may therefore change during the first few days in hospital.

If the test is available, elevated levels of serum granulysin taken in the first few days of a drug eruption may be predictive of SJS/TEN.

Skin biopsy is usually required to confirm the clinical diagnosis and to exclude Staphylococcal Scalded Skin Syndrome (SSSS) and other generalised rashes with blisters.

The histopathology shows keratinocyte necrosis (death of individual skin cells), full thickness epidermal/epithelial necrosis (death of an entire layer of skin), minimal inflammation (very mild lymphocytic infiltrate of the superficial dermis). The direct immunofluoresence test on the skin biopsy is negative, indicating the disease is not due to deposition of antibodies in the skin.

Blood tests do not help to make the diagnosis but are essential to make sure fluid and vital nutrients have been replaced, to identify complications and to assess prognostic factors (see below). Abnormalities may include:

Patch testing rarely identifies the culprit in SJS/TEN following recovery, and is not recommended.

What is SCORTEN?

SCORTEN is an illness severity score that has been developed to predict mortality in SJS and TEN cases. One point is scored for each of seven criteria present at the time of admission. The SCORTEN criteria are:

The risk of dying from SJS/TEN depends on the score.

SCORTEN predicted mortality rates
SCORTEN 0-1 >3.2%
SCORTEN 2 >12.1%
SCORTEN 3 >35.3%
SCORTEN 4 >58.3%
SCORTEN 5 or more >90%

What is the treatment for SJS/TEN?

Care of a patient with SJS/TEN requires:

Case reports and small patient series have reported benefit from active adjuvant treatments delivered during the first 24-48 hours of illness. Ciclosporin 3-5 mg/kg/day, cyclophosphamide, intravenous immunoglobulin (IVIG) 2-3 g/kg, and plasmapheresis have had variable responses. As SJS-TEN is fortunately a rare condition, controlled trials of therapies in large numbers of patients are difficult.

The role of systemic corticosteroids (cortisone) remains controversial. Studies giving high doses of corticosteroids for a short time at the start of the reaction have been shown them to be beneficial. However concerns have been raised that they may increase the risk of infection, impair wound healing and other complications. They are not effective later in the course of the illness.

The biologic agents infliximab and etanercept, which inhibit tumour necrosis factor alpha (TNFα), have in isolated case reports been reported to be useful. Thalidomide, trialled because of its anti-TNFα effect, increased mortality, and should not be used.

Some units recommend using heparin routinely to prevent thromboembolism (blood clots).

Specialised nursing care is essential as is frequent reassessment with appropriate adjustments in care.

Complications of SJS/TEN

This condition can be fatal due to complications in the acute phase. The mortality rate is up to 10% for SJS and at least 30% for TEN.

During the acute phase, potentially fatal complications include:

Longterm complications include:

It may take weeks to months for symptoms and signs to settle.

Dyspigmentation due to toxic epidermal necrolysis
Nail loss after toxic epidermal necrolysis
Nail shedding
Symblepharon after toxic epidermal necrolysis
Complications of SJS/TEN

Proposed mechanism

SJS/TEN is a rare and unpredictable reaction to medication. The mechanism has still not been understood and is complex.

  1. Drug specific CD8+ cytotoxic lymphocytes can be detected in the early blister fluid. They have some natural killer cell activity and can probably kill keratinocytes by direct contact.
  2. Cytokines implicated include perforin/granzyme, Fas-L and Tumour Necrosis Factor (TNF) alpha.

There are probably two major pathways involved:

  1. Fas-Fas ligand pathway of apoptosis has been considered a pivotal step in the pathogenesis of TEN. The Fas ligand (FasL), a form of tumour necrosis factor, is secreted by blood lymphocytes and can bind to the Fas ‘death’ receptor expressed by keratinocytes.
  2. Granule-mediated exocytosis via perforin and granzyme B resulting in cytotoxicity (cell death). Perforin and granzyme B can be detected in early blister fluid and it has been suggested that levels may be associated with disease severity.

Prevention of SJS-TEN

People who have survived SJS-TEN must avoid taking the causative drug or structurally related medicines. Cross -reactions can occur between:

In the future, we may be able to predict who is at risk of SJS-TEN using genetic screening, as specific at-risk HLA alleles have been recognised in certain populations.

Allopurinol should be prescribed for good indications (e.g. gout with hyperuricaemia) and commenced at low dose (100 mg/day), as SJS-TEN is more likely at doses >200 mg/day.

Related information


  1. Carr DR, Houshmand E, Heffernan MP. Approach to the acute, generalized, blistering patient. Semin. Cutan. Med. Surg. 2007; 26:139-146.
  2. Cotliar J. Approach to the patient with a suspected drug eruption. Semin. Cutan. Med. Surg. 2007; 26:147-154.
  3. Meth MJ, Sperber KE. Phenotypic diversity in delayed drug hypersensitivity: An immunologic explanation. The Mount Sinai Journal of Medicine 2006; 73: 769-776.
  4. Mockenhaupt M. Severe drug-induced skin reactions: clinical pattern, diagnostics and therapy. Journal der Deutschen Dermatologischen Gesellschaft 2009; 7: 142-162.
  5. Roujeau J-C. Clinical heterogeneity of drug hypersensitivity. Toxicology 2005; 209: 123–129.
  6. Wolf R, Orion E, Marcos B, Matz H. Life-threatening acute adverse cutaneous drug reactions. Clinics in Dermatology 2005; 23 (2); 171-181.
  7. Wolkenstein P, Chosidow O, Fléchet ML, Robbiola O, Paul M, Dumé L, Revuz J, Roujeau JC. Patch testing in severe cutaneous adverse drug reactions, including Stevens-Johnson syndrome and toxic epidermal necrolysis. Contact Dermatitis. 1996;35:234-6.
  8. Fujita Y, Yoshioka N, Abe R, et al. Rapid Immunochromatographic Test for Serum Granulysin Is Useful for the Prediction of Stevens-Johnson Syndrome and Toxic Epidermal Necrolysis. J Am Acad Dermatol. 2011;65:65-68
  9. Fernando SL. The management of toxic epidermal necrolysis. Aust J of Dermatol 2012; 53: 165-71

On DermNet NZ:

Other websites:

Medscape Reference:

Books about skin diseases:

See the DermNet NZ bookstore

Author: original SJS and TEN pages were written by Vanessa Ngan and Dr Amanda Oakley in 2003, and updated by Dr Delwyn Dyall-Smith in 2009.

DermNet NZ does not provide an online consultation service.
If you have any concerns with your skin or its treatment, see a dermatologist for advice.