Since the 1950s, it has been customary to consider erythema multiforme major (with multipolar mucosal erosions), Stevens-Johnson syndrome (SJS) and ectodermosis erosive pluriorificial to be one and the same entity, although the two latter disorders had been clearly distinguished from erythema multiforme by the authors who originally described them.
Lyell described the clinical picture of toxic epidermal necrolysis (TEN) in 1956, also clearly differentiating it from erythema multiforme. It was subsequently noticed that SJS and TEN could be triggered by the same cause and that an initial clinical picture of SJS could develop into TEN.
In 1993, a return to the original clinical classification system was suggested with, on the one hand, erythema multiforme minor and majus with its distinctive target-like lesions and acral distribution and, on the other, SJS and TEN which are drug-induced and present as erythematous macules which never have the target-like appearance described above and converge to form more or less extensive confluent plaques with a predominantly "central" distribution on the chest and at the tops of the limbs . In this classification system, SJS and TEN are seen as variants of the same extremely serious condition, Epidermal Necrolysis, which is mainly drug-induced and whose only distinctive feature is the extent of the skin detachment it induces (< 10 percent of the total body surface for SJS, more than 30 percent for TEN and from 10 to 30 percent for overlapping forms).
These are exceptional accidents, with 1.5-2 cases reported per 1 million inhabitants per annum. The risk is exacerbated by inflammatory rheumatoid disease, particularly systemic lupus erythematosus, cancer, radiotherapy and, above all, HIV infection.
The onset of the disease is characterised by non specific symptoms (fever, sore throat, burning eyes and a red rash), followed one to three days later by the distinctive mucocutaneous lesions. The rash starts on the face and upper torso and spreads very rapidly. It consists of purple, nearly purpuric, macules rapidly becoming confluent on the torso.
Limited in SJS, this confluence is extensive in TEN. With SJS, the epidermal separation occurs in patches, whereas with toxic epidermal necrolysis large sheets of epidermis can be peeled off or removed with the slightest pressure on the necrotic areas, exposing the bright red dermis beneath The lesions reach their peak in 3 to 4 days, exceptionally more rapidly or slowly. The body surface area affected is an important prognostic factor and must be assessed using a burns victims table or the "palm and fingers = 1 percent" rule.
More than 90 percent of patients have mucosal lesions. These are painful diffuse erosions generally affecting several areas of the body. Particular attention must be paid to the eyes owing to the risk of sequelae.
Cutaneous biopsies show lesions combining massive epidermal cell apoptosis  and a slight mononuclear infiltrate in the superficial dermis and epidermis.
There are no specific signs of the condition on laboratory tests. All patients develop anaemia in a few days, as a result of blood loss from the skin and transient erythroblastopenia. Moderate neutropenia is common. Nearly all patients will present with lymphocytopenia. TEN is unlike other drug reactions in that patients rarely present with eosinophilia or hyperleukocytosis.
Several abnormal laboratory test result from “acute skin failure”, including hypoalbuminemia, hypophosphoremia, hyperglycemia, glycosuria, altered renal function and water-electrolyte balance and elevated enzymes (ASAT, ALAT, amylase, CPK, LDH, etc.).
The complications of epidermal necrolysis are extremely serious. The main complications are a direct result of epidermal destruction and are proportionate to the surface area affected. These are loss of water and electrolytes (2 to 3 l/day in an adult with 50 to 100 percent detachment of the skin surface area), loss of heat and risk of systemic infection as a result of the loss of skin’s barrier function. The "stress" hormones increase catabolism and cause insulin resistance, often accompanied by hyperglycaemia and glycosuria.
Other complications are a result of the damage to other epitheliums caused by the same destruction processes affecting the epidermis. These are not necessarily related to the extent of the skin lesions and may include rarely gastrointestinal damage with abdominal pain and bloody diarrhoea and less rarely, bronchopulmonary complications causing diffuse, severe bronchopulmonary disease with hypoxemia, rapidly progressing towards respiratory distress. In such patients, bronchoscopy shows detachment of the tracheal and bronchial epithelium. The risk of bronchial involvement is not confined to patients with toxic epidermal necrolysis; it has also been reported in cases of Stevens-Johnson syndrome in which the cutaneous damage is limited.
The EN-related death rate is 22-25 percent, from 5-10 percent with the limited forms up to 40 percent when 30 percent or more of the epidermis is lost. Death most often results from pulmonary damage and/or infection . The number of patients with a serious pre-existing condition dying some weeks later must also not be overlooked.
The epidermis starts to regenerate after a few days. Progress is rapid and healing is virtually complete within two to three weeks. The mucosal erosions last longer. Almost all patients experience late complications. Pigmentation disorders are common and more than 50 percent of surviving patients will present with ocular damage, a combination of varying degrees of dry eye syndrome with kerato-conjunctivitis, irritation caused by dystrophic eyelashes and conjunctival – possibly even corneal - malpighian metaplasia. All these symptoms may progress over several months and severely impair the sight. Other sequelae are also common and have an impact on quality of life. These include abnormal nail regrowth, alopecia, salivary anomalies and odontopathies.
According to the currently accepted definition, epidermal necrolysis is essentially drug-induced . Sixty to 70 percent of cases can clearly be imputed to a specific drug and 10 to 20 percent arise in patients taking a variety of medicines without it being possible to identify the "culpable" medicinal product. In the remaining 20 or so percent of cases, an obvious drug cause cannot be identified; few of these have been convincingly attributed to infection (Mycoplasma pneumoniae and hepatitis B infection, in particular), and the rest are believed to be "idiopathic".
The time to onset is remarkably standard, from 1 to 4 weeks after first intake of the medicinal product, the median time being 12 to 14 days. The reaction may start several days after the medicinal product has been withdrawn if the elimination half-life is long. When a patient has been taking a medicinal product for more than 2 months, the risk is negligible.
The high risk products are allopurinol, antiepileptic medicines (phenobarbital, lamotrigine, carbamazepine, phenytoine), antibacterial sulphonamides (sulfamethoxazole-trimethoprim and sulfadiazine) certain non-steroidal anti-inflammatory drugs (oxicam derivatives), and the antiretroviral drug nevirapine. A lower risk has been demonstrated for several antibiotics (aminopenicillins, cephalosporins, quinolones and cyclins) .
It is difficult to establish a causal link on an individual basis because, for the moment, no in vitro or in vivo sensitive and specific tests have been established.
The keratinocytes are destroyed through apoptosis, which in itself is probably induced by cytotoxic lymphocytes invading the epidermis. The contrast between the extent of the necrosis and the low leukocyte count in situ indicates that cytokin amplification phenomena are probably involved. TNF-α and fas-Ligand were initially suspected but are now not considered to play a major role, unlike granulysin .
Contrary to the long accepted hypothesis of an immune response directed against reactive metabolites, it is now currently accepted that the cytotoxic lymphocytes present at the site of the lesions and react against the parent form of the medicinal product.
7.1 - MECHANISMS
This is a serious "drug allergy". It is an extremely rare and unforeseeable kind of allergy. It causes more or less extensive destruction of the uppermost layer of the skin, like a second degree burn. The lining of the mouth, eyes and genitals are also affected. Once the rash starts, it spreads for about 5 days. Its severity of the reaction depends heavily on how much of the skin and mucosa is affected. The skin starts to regenerate rapidly, on average 10 to 15 days after the rash is first observed. The mucous membrane recovers more slowly and the eye lesions may persist for several months. Once it has healed, the skin almost always recovers a normal texture although its colour may have changed permanently.
It is essential to identify the drug that may have caused the reaction as if it – or another similar drug – is taken again, the same or an even more severe reaction may occur. In exceptional cases, particularly in children, the disease can be triggered by a bronchopulmonary infection (Mycoplasma pneumoniae). In some cases, a cause is never found. To the best of our knowledge, there is no risk of recurrence in such cases.
7.2 - OBSERVATION AND MONITORING
Patients must be hospitalised for between 10 and 30 days, depending on the severity of the condition. Serious cases will have to be transferred to a specialist care facility to ensure that patients receive continuous care and the intravenous treatments required over the two weeks during which the skin is no longer able to fulfil its barrier function.
The rash may continue to spread for a number of days. At the present time, we do not have any treatments proven as capable to halt the spread of the rash but it is possible that the medical team will offer the patient to participate in a trial of a new medicine.
Once cured, the patient will require several months’ medical monitoring and will need care for any residual damage caused by the rash. This particularly concerns the eyes. Patients are often left with lighter and darker patches on their skin. These colour changes fade over time but use of high factor sun protection is generally required.
7.3 - GENETIC RISKS
This disease is neither contagious nor transmittable. Although exceptional, another family member may be allergic to the same drug, as it only seems “permitted” in certain genes. As a precautionary measure, the drug suspected to be the cause of the reaction should not be given to other family members unless they have already taken it without any ill effects.
7.4 - IMPACT ON SOCIAL LIFE AND WORK
Patients will generally have to stay off work for a few weeks after leaving the hospital if their eyes have not been affected by the disease. If there is severe eye damage (less than one in ten cases), it may be impossible for patients to return to their previous jobs.
In France, there is a statutory compensation scheme for no-fault medical accidents (generally the case for these diseases), but its application in practical terms appears difficult, although this is not the case in other countries (Sweden, Japan, etc.).
Patients must always carry with them a card listing the medicines they cannot take. Other medicines may be used without any particular risks.
8.1 - DURING THE ACUTE PHASE
The extent of the cutaneous lesions, expressed as a percentage of body surface area, must be assessed on a daily basis. Areas of erythema must be distinguished from areas in which the skin is detached or "detachable" (positive Nikolsky’s sign). The necrolytic surface (detached + detachable), determines the amount of fluids to be given to the patient, is an indicator of disease progression and has prognostic significance. This evaluation is difficult and the surface affected is usually vastly over-estimated. It should be measured segment by segment using the tables developed for burns victims. It is useful to bear in mind that the surface area of the palm and fingers represents 1 percent of the total body surface. Generally speaking, small, even very disseminated blisters cover much less than 10 percent (or ten hands).
Clinical monitoring must include respiratory function (respiratory rate, pulse and x-rays), state of consciousness and haemodynamic status (heart rate, blood pressure and hourly diuresis volume).
Daily laboratory tests must include measurement of blood and urine electrolytes, blood and urine glucose, phosphoremia, ASAT, ALAT, amylasemia, CPK and CBC.
8.2 - EVALUATION OF PROGNOSIS
The two most important factors to be taken in to account when assessing the prognosis are age and the extent of cutaneous detachment, which is expressed as a percentage of total body surface area.
A scoring system, SCORTEN, has been validated by several teams . It consists of seven items, each of which, if present, counts for one point.
– age > 40 years;
– recent cancer;
– surface area affected on entry > 10 percent;
– heart rate > 120/min;
– urea > 10 mmol/l;
– serum bicarbonates < 20 mmol/l;
– glycaemia > 14 mmol/l.
Patients with a score of not more than 2 have a 90 percent chance of surviving; this decreases to less than 50 percent if the score is not less than 4.
8.3 - MONITORING OF COMPLICATIONS
The cutaneous long-term complications of the disease are sweating abnormalities and pigment disorders which can last for several years. Although rare, some scars, particularly those on pressure points, may become hypertrophic.
The most serious lasting consequence is eye damage. About half of patients will present with varying degrees of ocular involvement . The effects range from dry eye syndrome, to abnormal eyelash regrowth, synechia, keratitis and corneal epithelium malpighian metaplasia, all of which may have severe functional repercussions. As the condition of the eyes may not stabilise for a number of years, patients must be seen by an ophthalmologist before their discharge and a few weeks later.
The severity of the reaction must be spotted as soon as possible (present median delay of 2 days) and the patient referred to an appropriate hospital department (present median delay of 4 days). Assistance with diagnosis and whether or not to transfer the patient can be obtained in France from a reference centre. (E-mail: email@example.com Internet site: http://dermatosebulleuse.fr/ URGENCES : +33 (1) 49 81 21 11, extension 36053)
The first step of treatment is to stop administering the drug suspected to have caused the reaction. The prognosis is more serious if a drug causing the reaction is continued after the distinctive signs of epidermal necrolysis.
Recommendations for treatment have been established by burns units in the USA (5) and in France, under the supervision of the National Health Authority (HAS) .
All agree on the following important point: NO "SPECIFIC" TREATMENT HAS PROVEN ITS WORTH AND TREATMENT IS ESSENTIALLY SYMPTOMATIC.
9.1 - SPECIFIC TREATMENTS
Untreated, the lesions progress for a mean of 5 to 6 days after the first symptoms are observed, i.e. 2 to 3 days after the patient is hospitalised. To be effective, a treatment intended to halt progression of the lesions must work almost immediately effect and be administered as soon as the patient is admitted to hospital.
In view of the suspected immunopathological pathogenesis of the reaction, several authors have treated the necrolysis with systemic corticosteroids, immunosuppressants (cyclophosphamide) and anticytokins (anti-TNF, immunoglobulins).
An oral dose of 60 to 80 mg/day of prednisone administered throughout the progressive phase, or intravenous bolus injections of methylprednisolone for 2 to 3 days continue to be recommended in some countries.
Thalidomide is considered to be a potent inhibitor of TNF synthesis and has been tested in a placebo-controlled, double-blind clinical trial. This is the only controlled clinical trial to be undertaken in toxic epidermal necrolysis and was interrupted owing to the high death rate which, once the blind was broken, proved to be significantly related to the administration of thalidomide . Thalidomide must therefore not be given to patients with epidermal necrolysis. Without contraindicating the use of other anti-TNF agents, this result means that extreme caution is called for.
The theoretical basis for this treatment was the inhibition of fas/fas-L mediated apoptosis in highly apoptosis-sensitive lymphocytes observed with IVIg in vitro. We now know that the keratinocyte apoptosis observed in EN is not fas-L dependent .
In the first open clinical trial in ten patients with EN receiving a 2 g/kg dose of IVIg, the course of the disease was unusually short and the outcomes positive. Considered to be a major treatment breakthrough, this result was subsequently "backed up" by several articles reporting on a number of compiled series in which the treatment outcomes were also positive. However, these compilations included so many duplicate cases that their credibility is doubtful .
Analysis of a large French-German (level of evidence lower than that of controlled trials but higher than in series) found that neither IVIg nor systemic corticosteroids were associated with a significant reduction in mortality, versus symptomatic treatment alone .
At the present time, the therapeutic arsenal for epidermal necrolysis is therefore limited to symptomatic treatments.
The increasingly striking analogies with transplant rejection phenomena nonetheless indicate that the studies of corticosteroid bolus injections and ciclosporin, whose preliminary results are rather encouraging, are worth pursuing.
9.2 - SYMPTOMATIC TREATMENT
Patients must be admitted to a specialist intensive care facility or burns unit for care as soon as possible. During their transfer, they must be handled with extreme care. In particular, adhesive electrocardiogram electrodes are to be avoided.
The principles of symptomatic treatment are the same as for extensive burns.
The environmental temperature must be between 28 and 30°C to reduce energy loss, shivering and stress. Use of special beds will improve the patient’s comfort and facilitate nursing.
Cutaneous erosions are extremely painful. Recourse to morphine is almost always necessary. Pain relief both reduces stress and makes nursing the patient easier.
A venous line is indispensible. If possible, it should be inserted at a distance from the damaged areas of skin. The quantity of infusion initially administered is calculated using the sum of the detached and "detachable" areas and is less than that delivered to a burns victim (about two thirds to three quarters of the amount). For example, a patient weighing 70 kg and with 50 percent detached/detachable epidermis requires about 5 litres a day, part of which is administered by enterally.
Electrolyte solutions and potentially plasma replacement solutions (liquid gelatine [Plasmion®], starch [Hyperhes® Voluven ®]) are to be preferred.
All patients will receive continuous enteral nutrition via a nasogastric tube. This provides patients with part of their daily fluid intake, limits lengthy use of an IV drip and reduces the risk of bacterial translocation from the gastrointestinal tract.
The damaged skin is rapidly colonised, initially by staphylococci and then by Gram negative bacteria from the gastrointestinal flora. If introduced too early, antibiotic treatment may encourage bacterial resistance. Infection entering the blood stream through the skin, gastrointestinal tract or an IV catheter is a major source of morbidity and mortality. It is difficult to diagnose and must be suggested as soon as any untoward symptoms are observed. There is no consensus on how to prevent and treat such infections. We recommend:
– isolation of the patient and aseptic nursing technique to prevent the risk of nosocomial contamination;
– twice daily bathing/spraying with an antiseptic solutions (chlorhexidine); as a precautionary measure, silver sulfadiazine is to be avoided if there is a possibility that a sulfamide antibiotic could be responsible for the reaction.
– broad spectrum antibiotics at the first sign of infection. The antibiotic used will be guided by the microorganisms present in the damaged skin (bacterial mapping tests done every two days) and an understanding of the bacteriological environment in the intensive care unit. As the distribution volume of antibiotics is often boosted by transepidermal water loss, patients must first be treated with the highest doses in the therapeutic range with the subsequent doses determined on the basis of the plasma concentrations assayed.
Hyperglycaemia is common and should only be tolerated if glycosuria is not also present. In such cases, continuous insulin is required, generally at high doses (insulin-resistance).
Hypophsophoremia may be severe and must be corrected as it increases insulin resistance and enhances the risk of respiratory failure as a result of muscle degeneration.
Patients presenting with the early symptoms of hypoxemia, which is often a sign of erosive lesions in the bronchi, must be transferred to an appropriate unit for mechanical ventilation.
Certain burns units continue to debride dead skin, remove exsudates on a daily basis and then cover the dermis with allografts or skin substitutes. We feel that this is an unnecessarily aggressive attitude (particularly in view of the repeated general anaesthetics involved). We recommend leaving the necrotic epidermis in place. Clinical experience has taught us that epidermal regeneration is much faster under the dead epidermis than on an exposed dermis. Exposed areas can be treated with vaseline coated dressings, antiseptic dressings or non-adhesive hydrocellular dressings.
Treatment of mucosal lesions
Mucosal lesions may develop into synechia. During the acute stage of the disease, the priority is to prevent synechia from forming, paying particular attention to the eyes and female genital mucosa. Regular examination of these tissues by an expert is essential.
At the very least, the eyes should be kept constantly hydrated (saline solution, methyl cellulose, etc) and any nascent synechia destroyed with a non-traumatic instrument. Corticosteroid eyedrops do not appear to prevent sequellae.
Genital care includes application of an ointment to the erosions and manual destruction of any nascent synechia. For diffuse vulvar or vaginal erosive lesions, we recommend the use of a pessary.
9.3 - TREATMENT OF COMPLICATIONS
The consequences of the disease are more common, longer-lasting and debilitating than originally thought. Over 80 percent of survivors continue to present complications a year after the acute episode. All patients must therefore be monitored to allow the prompt detection and management of sequellae.
Patients should be advised to apply appropriate sun protection measures to prevent exacerbation of the hyperpigmented areas of skin induced by the disease.
Although very unusual, hypertrophic scares may develop. These must be treated immediately with compression clothing whose efficacy has been demonstrated in a burns setting.
Artificial tears, destruction of eyelashes and vitamin A creams (although their efficacy has not been proven) may be required for the treatment of ocular complications. The development and use of scleral lenses seems to have been a major breakthrough .
9.4 - SUBSEQUENT USE OF MEDICINES
Further use of the suspect agent(s) as determined using a specific drug causality approach  must be contraindicated in writing. All the – very few - cases of recurrence reported in the literature were attributed to the same generic medicine or to chemically similar substances. Contraindicating all the "high risk" medicines is therefore not justified. The existence of a few familial cases and the discovery of a genetic predisposition  justify extending this contraindication to all blood relatives.
The address of victims associations must be given to the patient and his/her family. (In France: Amalyste http://amalyste.free.fr )
Reference centre for toxic and autoimmune acquired bullous dermatoses, Co-ordinators for the Ile de France: Professor Catherine Prost, Professor Pierre Wolkenstein
Hôpital Henri Mondor, Assistance Publique - Hôpitaux de Paris
Service de dermatologie
51 Avenue du Maréchal de Lattre de Tassigny
94010 Créteil Cedex
SCAR Study Group. Correlations between clinical patterns and causes of erythema multiforme majus, Stevens-Johnson syndrome, and toxic epidermal necrolysis: results of an international prospective study. Arch Dermatol, 2002, 138 : 1019-1024. Check on Pubmed
A clinical classification of cases of toxic epidermal necrolysis, Stevens-Johnson syndrome and erythema multiforme. Arch Dermatol, 1993, 129 : 92-96. Check on Pubmed
Granulysin is a key mediator for disseminated keratinocyte death in Stevens-Johnson syndrome and toxic epidermal necrolysis. Nat Med 2008, 14 : 1343-1350. Check on Pubmed
Bacteremia in Stevens-Johnson syndrome and toxic epidermal necrolysis: epidemiology, risk factors, and predictive value of skin cultures. Medicine (Baltimore), 2010, 89 : 28-36. Check on Pubmed
Toxic epidermal necrolysis clinical guidelines. J Burn Care Res, 2008, 29 : 706-712. Check on Pubmed
Toxic epidermal necrolysis and Stevens-Johnson syndrome : does early withdrawal of causative drugs decrease the risk of death ? Arch Dermatol, 2000, 136 : 323-327. Check on Pubmed
Performance of the SCORTEN during the first five days of hospitalization to predict the prognosis of epidermal necrolysis. J invest Dermatol, 2006, 126 : 272-276. Check on Pubmed
Risk factors for the development of ocular complications of Stevens-Johnson syndrome and toxic epidermal necrolysis. Arch Dermatol, 2009, 145 : 157-162. Check on Pubmed
A European study of HLA-B in Stevens-Johnson syndrome and toxic epidermal necrolysis related to five high-risk drugs. Pharmacogenet Genomics, 2008, 18 : 99-107. Check on Pubmed
Stevens-Johnson syndrome and toxic epidermal necrolysis: assessment of medication risks with emphasis on recently marketed drugs. The EuroSCAR-study. J Invest Dermatol, 2008, 128 : 25-44. Check on Pubmed
Br J Dermatol, 1996, 134 : 710-714. Check on Pubmed
ALDEN, an Algorithm for Assessment of Drug Causality in Stevens-Johnson Syndrome and Toxic Epidermal Necrolysis: Comparison With Case-Control Analysis. Clin Pharmacol Ther , 2010, 88 : 60-68. Check on Pubmed
Effects of treatments on the mortality of Stevens-Johnson syndrome and toxic epidermal necrolysis: A retrospective study on patients included in the prospective EuroSCAR Study. J Am Acad Dermatol, 2008, 58 : 33-40. Check on Pubmed
Vision-related function after scleral lens fitting in ocular complications of Stevens-Johnson syndrome and toxic epidermal necrolysis Am J Ophthalmol, 2009, 148 : 852-859. Check on Pubmed
Toxic epidermal necrolysis and intravenous immunoglobulin: a review. Semin Cutan Med Surg, 2006, 25 : 91-93. Check on Pubmed
Randomised comparison of thalidomide versus placebo in toxic epidermal necrolysis. Lancet, 1998, 352 : 1586-1589. Check on Pubmed
Therapeutics in Dermatology, Fondation René Touraine © 2001-2012