Stevens-Johnson syndrome and toxic epidermal necrolysis
Adverse drug reactions account for most cases of Stevens-Johnson syndrome, a severe skin condition with a high mortality rate.
Source: Dr P. Marazzi / Science Photo Library
In this article you will learn:
- The drugs associated with an increased risk of developing Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN)
- The signs and symptoms of SJS and TEN
- How patients with SJS and TEN are managed in hospital
- How to assess the prognosis of patients with SJS and TEN
Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are rare, life threatening conditions usually associated with medicines use. They are severe forms of exfoliative dermatitis, and characterised by extensive epidermal erythema and blistering, which leads to necrosis and detachment of the epidermis. They can also cause mucocutaneous lesions.
There is no formal classification of SJS and TEN and there is a lack of consensus as to whether they are separate conditions or varying degrees of the same disorders differing only in severity. The more popular belief that they are similar conditions has not been universally accepted due to subtle histological differences.
The exact incidence of SJS and TEN is unknown. SJS is estimated to affect 1–6 patients per million population per year, and it is up to three times more common than TEN,. Incidence is higher in patient groups taking medicines that can trigger the conditions, such as antiepileptics and antiretrovirals, with an estimated incidence of 1–2 cases of SJS per 1,000 population each year in people treated for HIV,.
The precise cause of SJS and TEN is unknown, although it is believed the conditions are caused by an immune reaction. Pharmacogenomic studies indicate that ethnicity and certain human leukocyte antigen (HLA) types may predispose patients to SJS and TEN,. The HLA-B*1502 allele has a strong association with carbamazepine-induced SJS and TEN in the Han Chinese population, and the US Food and Drug Administration recommends screening all people of Asian and South Asian ethnicity for the allele before starting treatment with carbamazepine. A strong association has also been identified between the HLA-A*3101 allele and carbamazepine-induced SJS and TEN in patients of every ethnicity. Studies have also shown patients with the HLA-B*5801 allele are at increased risk of allopurinol-induced SJS/TEN. This allele is equally distributed among all ethnic groups, which makes genetic screening more difficult,.
Patients with systemic lupus erythematosus, cancer or collagen vascular disease, and those who have recently received a bone marrow transplant or radiotherapy, are at increased risk of developing SJS or TEN.
Signs and symptoms
SJS is less severe than TEN and the conditions are distinguishable predominantly by the total body surface area affected by skin detachment. SJS involves less than 10% of a patient’s total body surface area, whereas TEN involves more than 30%. The conditions overlap in patients for whom 10–30% of body surface area is affected,.
Initially, patients develop prodromal effects such as fever, malaise, cough, photophobia and keratoconjunctivitis; this usually occurs one to three weeks after starting the causative medicine in cases caused by an adverse reaction. Lesions typically appear one to three days later.
Skin lesions first appear as macules, often as targetoid lesions which look like a bullseye, and usually start on the trunk, face and neck before spreading to other areas. Diffuse erythema may be the first skin abnormality to occur in some cases.
Following the appearance of macules, vesicles and bullae form. The skin begins to slough within days and in some cases in hours. Patients may present with Nikolsky’s sign, in which the epidermal layer easily sloughs off when pressure is applied to the affected area ,.
The widespread blistering of the skin results from keratinocyte apoptosis, which is a series of biochemical reactions that lead to cell changes and ultimately cell death.
Mucus membranes, including the buccal, ocular and genital mucosae, are affected in more than 90% of patients with SJS or TEN. Usually more than one site is affected. The mucosae of the upper and lower respiratory tract may be affected, potentially requiring the need for mechanical ventilation.
The eyes are affected in up to 80% of cases. The most common symptom is severe conjunctivitis that sometimes leads to corneal ulceration, anterior uveitis and synechiae. Eye changes often resolve completely, but can sometimes result in the most debilitating complications, including blindness.
The acute phase of SJS and TEN typically lasts 8–12 days. Wound healing and regrowth of the epidermis begins after several days, and often takes two to four weeks; the duration of a patient’s recovery varies considerably, but they can usually be discharged from hospital shortly after re-epithelialisation begins.
Why does cell death occur?
Although medicines are the cause in the majority of SJS and TEN cases, the link between specific medicines and epidermal necrosis has not been determined. There are several proposed theories supporting immunopathologic pathways causing keratinocyte apoptosis in SJS and TEN.
Cytotoxic T lymphocytes, especially CD8 + T lymphocytes, and cytokines have been demonstrated to play an important part in this process ,  ,  . Recent evidence also suggests that granulysin, a cytolytic protein released from cytotoxic T lymphocytes and natural killer cells, might have a key role in keratinocyte death and the pathogenesis of SJS and TEN. Elevated concentrations of granulysin in blister fluid and cells from SJS/TEN patients have been associated with increased disease severity ,  .
Medicines are the leading cause of SJS and TEN, and are attributed in over 75% of cases in both adults and children. It is essential that a thorough drug history is taken for patients with SJS or TEN to identify the possible cause. Stopping a causative medicine early can improve the patient’s prognosis.
Specific attention must be made to the time interval between starting a medicine and the onset of symptoms. This is crucial in not only identifying the cause but also to prevent ‘innocent’ medicines the patient may have been taking from being stopped unnecessarily. The risk of developing SJS and TEN is highest when the causative medicine has been recently initiated, and is generally limited to the first eight weeks of treatment; medicines are unlikely to cause SJS or TEN once used for more than eight weeks.
In the 1990s, the first case control study for SJS and TEN was published. This looked at patients in selected hospitals in Europe and assessed the relative risk of SJS and TEN corresponding with commonly used medicines (see ‘Medicines implicated in causing SJS and TEN’). In 2007, a second large multinational case-control was published, which looked at newly marketed medicines following a number of case reports. This study identified newer drugs associated with SJS and TEN, including nevirapine, lamotrigine, sertraline, pantoprazole, and tramadol.
The second most common cause of SJS and TEN is infection, particularly in children. This includes upper respiratory infections, otitis media, pharyngitis, and Epstein-Barr virus, Mycoplasma pneumoniae and cytomegalovirus infections. The routine use of medicines such as antibiotics, antipyretics and analgesics to manage infections can make it difficult to identify if cases were caused by the infection or medicines taken.
Rare causes of SJS and TEN include vaccinations, radiotherapy, bone marrow transplantation, systemic diseases, contrast media, external chemical exposure and herbal medicines.
Medicines implicated in causing SJS and TEN
- Phenobarbital, phenytoin
- Piroxicam, tenoxicam
- Amoxicillin, ampicillin
- Azithromycin, clarithromycin, erythromycin
- Ciprofloxacin, levofloxacin, pefloxacin
- Non-steroidal anti-inflammatory drugs (NSAIDs)
- Pipemidic acid
- Valproic acid
There are no definitive treatments for SJS and TEN beyond stopping any causative medicine and managing complications. Treatment at a burns or intensive care unit has been linked with better patient outcomes and reduced mortality, and most patients will be transferred to specialised wards,.
Patients with SJS or TEN receive similar care to those with major burns. This includes wound care, fluid and electrolyte management, ocular care, nutritional support, temperature management, pain control and treatment of infections. A multidisciplinary approach is essential.
Dressings and topical antibacterial agents should be used to cover cleansed wounds and prevent infection. There is no evidence for the use of one type of dressing over another; currently the use of non-adherent nanocrystalline silver dressings is favoured in practice because these can be changed less frequently. Once the skin has regenerated, emollients are applied to keep the skin supple.
SJS and TEN can affect the eyes and lead to the development of synechiae, in which the iris adheres to either the cornea or lens, and blindness. Patients should therefore be referred to an ophthalmologist immediately for continuous assessment. Ocular care includes the use of topical antibiotics, topical corticosteroids, ciclosporin eye drops and daily lubrication with preservative-free drops or ointments. In some cases, an amniotic membrane graft may be considered, which involves taking the membrane from the innermost layer of a placenta and applying it to the surface of the eye. This acts as a biological contact lens to suppress inflammation and scarring ,.
Patients with SJS or TEN are at high risk of developing infections that can lead to sepsis. Despite these risks, administration of prophylactic antibiotics is not recommended as use has not been associated with improved survival and may increase the risk of resistance. Antibiotic therapy should therefore be reserved for suspected or proven infections.
Although there are no established treatments for SJS or TEN, several immunosuppressive or immunomodulating therapies have been used in clinical practice, including systemic corticosteroids, intravenous immunoglobulins (IVIG), cyclophosphamide, ciclosporin, thalidomide, plasmapheresis, granulocyte colony stimulating factor (GCSF) and anti-tumour necrosis factor (TNF) monoclonal antibodies. However, none has been adequately studied in randomised trials except thalidomide, which was associated with an increase in mortality.
Although TEN is listed in the Department of Health guidelines for use of IVIG, there is limited evidence showing benefit and some for potential harm. The use of IVIG for TEN remains controversial and treatment is at the discretion of the patient’s consultant.
SJS and TEN have a high overall mortality, ranging from 10% for SJS to more than 40% for TEN. The risk of mortality for these patients can be predicted within the first five days of hospitalisation using the SCORTEN scale, which is based on seven measured clinical and laboratory variables (see ‘SCORTEN scale’).
The most common causes of death in SJS and TEN are sepsis, acute respiratory distress syndrome and multiple organ failure. For patients who survive, long-term skin and ocular complications are common.
There is a high risk of recurrence in cases caused by an adverse reaction if the causative medicine is restarted. A recent cohort study found 7.2% of patients with SJS or TEN who survived their first episode were readmitted with a recurrence following discharge.
Following recovery, patients must be warned to avoid the trigger and of the potential risks of cross-sensitivity to other medicines. Because these patients may be genetically susceptible to drug-induced skin reactions, clinicians must take particular care when prescribing to this population.
|Heart rate (beats/min)||>120|
|Serum BUN (mg/dL)||>27|
|Detached or compromised body surface||>10%|
|Serum bicarbonate (mEq/L)||<20|
|Serum glucose (mg/dL)||>250|
|No of risk factors||Mortality rate|
|5 or more >90%||>90%|
John Bentley MPharm DipClinPharm is a surgical pharmacist and David Sie BPharm MSc is a specialist burns pharmacist, both at Chelsea and Westminster Hospital NHS Foundation Trust.
Citation: The Pharmaceutical Journal DOI: 10.1211/PJ.2014.20066728
Learning module unavailable
Learning module closed: 8 October 2015
Required score to pass: 70%
Sign in to take this article's CPD module.
CPD modules and learning certificates are only available to RPS members and subscribers.Sign in here
Recommended from Pharmaceutical Press