<% vol = 15 number = 1 nextlink = 22 prevlink = 12 titolo = "TREATMENT OF TOXIC EPIDERMAL NECROLYSIS (TEN)" volromano = "XV" data_pubblicazione = "March 2002" header titolo %>

Brambilla G.1, Brucato F.1, Angrisano A.2, Palmieri G.1

1 Divisione di Medicina Interna II, Ospedale Niguarda Ca’ Granda, Milan, Italy
2 Divisione di Pediatria, Ospedale “Caduti Bollatesi”, Bollate, Milan

SUMMARY.Toxic epidermal necrolysis (ten) is a common and potentially dangerous mucocutaneous disease characterized by rash, bullae, and diffuse exfoliation of wide cutaneous surface areas. A description is given of the disease, also known as Lyell’s syndrome, and it is distinguished from similar conditions. Information is provided regarding the epidemiology, aetiology, and pathogenesis of TEN, as well as its clinical features. Diagnosis and therapy are also considered. Pharmacological therapy is required, particularly with cyclosporin.


Toxic epidermal necrolysis (TEN), also known as Lyell’s syndrome, is a widespread life-threatening mucocutaneous disease that is particularly observed secondary to drug-taking. It characterized by rash, bullae, and diffuse exfoliation of wide cutaneous surface areas, as in second-degree burns. Separation of the dermal-epidermal junction causes Nikolsky’s sign and gives the skin the typical “wet dressing” appearance.


In 1956 Lyell distinguished two entities in the description of toxic epidermal necrolysis: the Staphylococcal Scalded Skin Syndrome (SSSS) and what is now known as TEN.1

SSSS is a generalized exfoliative dermatitis due to local infections of Staphylococcus aureus capable of producing both in vivo and in vitro a 24 kD toxin, which, when injected into the peritoneum of newborn mice, rapidly causes a bullous and exfoliative dermatitis. This disease usually occurs within the first three months of life and is rare in young and adult people. Nowadays the disease develops in isolated cases or in small epidemics in neonatology departments, probably caused by contagion from an asymptomatic carrier of the micro-organism on the hospital staff.

TEN differs from SASS because it is drug-related (no infectious causes have been demonstrated) and because not only the skin but also the mucosae are commonly involved. Furthermore, while in SSSS we observe a wide formation of superficial intra-epidermal blisters followed by desquamation of large sheets, with a good outcome except in patients suffering from hypovolaemic shock (mortality 2-3%), TEN is characterized by necrosis of all skin layers with a serious deterioration of health conditions and high mortality (approximately 30%). Other problems of classification are due to the relationship between TEN and erythema multiforme (EM). EM is a cell-mediated hypersensitivity reaction to many different immunological insults, including infections (both viral and bacterial) and drugs. EM minor, the most frequent form, consists of a cutaneous eruption symmetrically affecting the extensorial surfaces of the limbs, hands, and feet; in EM major, the eruption also involves the mucosae. Some researchers consider the Stevens-Johnson syndrome (SJS) to be a synonym for EM major. This syndrome was initially described as a cutaneous eruption with erosion of the mucosae, particularly the ocular mucosa. The most severe form of SJS can develop into extensive epidermolysis, and this is the reason why SJS has sometimes been classified as the severest form of EM. Differences in the clinical presentation and aetiology suggest that SJS should be classified as a separate entity from EM. EM major and minor are in fact hypersensitivity pathologies that often occur in response to infectious agents, while SJS and TEN are idiosyncratic reactions caused by drug-taking. A current definition distinguishes between SJS and TEN on the basis of the extent of cutaneous involvement: less than 10-30% in SJS and more than 10-30% in TEN. Although this scheme is reproducible, the nosological controversies continue.1,2


TEN is a rare affection, without sex or race predilection; it is commoner in elderly people even though it may occur in any age group.3,4 The greater incidence in the adult population is justified by the greater consumption of drugs compared with children and adolescents. The incidence of TEN is about 0.4-1.2 cases per million. The disorder seems to be commoner among persons with HIV-infection, patients with LES, and bone marrow recipients.


Many factors are involved in the aetiology of TEN, such as adverse reactions to drugs, vaccinations, malignant disorders, and graft versus host disease (GVHD).5 There are also idiopathic cases, in which is not possible to identify any certain cause.

With the diffusion of HIV infection and the use of trimethoprim-sulphamethoxazole for the prevention or treatment of Pneumocystis carinii pneumonia, a considerable increase has been observed in the number of cases of TEN: this probably reflects both an increased sensitivity to drugs among patients who are already immunodeficient and an increasing tendency to prescribe high-risk medicine.4

The most important drugs involved in the pathogenesis of TEN are sulphonamides, aminopenicillin, quinolones, cephalosporin, carbamazepine, phenobarbital, phenytoyn, NSAIDs, and allopurinol. Surprisingly, the use of corticosteroids may also be associated with TEN.6


It has been recently shown that TEN produces necrosis in all the skin layers because of an important process of apoptosis of the keratinocytes; the mechanism is not fully understood, but some elements suggest that the destruction of cutaneous and mucosal tissue is an immune-mediated reaction, as a result of a drug-induced reaction. T-lymphocytes directed to a particular drug have in fact been isolated in the serum and cutaneous lesions of patients with skin involvement induced by drugs.

Another interesting aspect is that the onset of TEN symptoms usually occurs 7-21 days after the beginning of medical treatment, and the rare cases of relapse initiate within 48 h of taking the drug, thus suggesting the existence of an immunological memory.

Finally, immunohistological studies on cutaneous biopsies of patients affected by early TEN have shown a widespread dermal infiltrate in the dermal-epidermal junction, composed primarily by helper CD4 T-lymphocytes and CD8 T-cells, suggesting a cell-mediated reaction versus keratinocytes.

It has been shown that in the fluid contained in the bullae of patients with TEN, T-lymphocytes predominate in the initial phases, while subsequently the cells of the monocyte-macrophagic line prevail7 and probably contribute to the progression of the necrosis through TNF-· production.8

Comparative studies have in fact demonstrated higher levels of soluble TNF-· and its receptors in the fluid of bullae of patients affected by TEN than in the blisters of patients with second-degree burns.

The possible role of genetic factors in the pathogenesis of TEN remains to be clarified. For example, a reduced acetylating capacity (slow acetylator) can facilitate the onset of TEN after drug taking.9-12

Clinical features

The initial symptoms of TEN are fever and weakness, with the possible association of conjunctivitis, pharyngitis, and itching. These symptoms usually last 2-3 days, and may suggest infection of the upper airways.

Mucosal lesions usually precede involvement of the skin by 1-3 days. In order of frequency, these involve the oropharynx, conjunctivae, genitalia, and the perineal areas. Subsequently an exanthema appears, first located in the head, neck, and shoulders and quickly extending and converging, with flaccid bullae and widespread areas of skin detachment. This detachment is typically at the level of the dermal-epidermal junction, as in second-degree burns.13 Nikolsky’s sign, i.e. the peeling-off of the epidermis in large sheets caused by a sliding touch even in areas that are apparently unaffected, is intensively positive. The lesions generally progress in 3-4 days, but can extend quickly in a few hours or more slowly in a week; the scalp is generally not involved. Erosive lesions of the mucosae are present in 97% of cases, in the oral cavity (93%), the eyes (78%), and the genitalia (63%).14 The anal mucosa may also be affected. The lips are commonly crusted and covered by scabs; swallowing is painful. Alterations of the general state include fever, indisposition, and pains. Electrolytic alterations can begin quickly and progress within a short time. Pancytopenia caused by diserythropoiesis and haemolysis can occur;4 prognosis in these cases is extremely bad.

Although mucocutaneous involvement is predominant, TEN must be regarded as a multisystemic pathology. With regard to the respiratory apparatus, there may be erosions of the bronchial mucosa and respiratory insufficiency as a result of increased capillary permeability; pneumonia ab ingestis is also possible.5 Glomerulonephritis and tubular necrosis can occur, as also alteration of gastrointestinal mucosa permeability. There may be hepatic involvement, pancreatitis, alterations of glycaemia, and hypophosphoraemia, which can cause disturbances of consciousness. In many patients death occurs in the first or second week after onset, generally because of haemodynamic shock, intestinal haemorrhages, pulmonary oedema, renal failure, systemic infections, fluid loss, or diffuse intravascular coagulation.15,16 The mortality rate is about 30% (range, 20-70%).


Diagnosis is not difficult in advanced phases of the pathology. In the initial phases problems can arise in comparisons with SSSS, which normally affects newborn babies but may also occur in adults with immunodeficiency conditions or in the presence of other associated pathologies.

A biopsy test is fundamental for diagnosis and shows necrosis of basal layer cells without massive inflammatory infiltration of the dermis; in specimens with mature lesions it is possible to observe necrosis of keratinocytes involving all the epidermis with diffuse detachment of the epidermis at the level of the basal membrane.


The treatment of TEN in burn units has considerably improved patients’ prognosis and survival because these centres are competent in most aspects of the appropriate care and nursing, such as protection of the cutaneous and mucosal surfaces involved, monitoring of the electrolytic balance, fluid replacement, nutritional support, and the prevention and treatment of infection. With regard to this last aspect, antibiotic prophylaxis is not recommended as a routine measure because of the risk of cross reactivity with the drug involved and the possibility of infection with resistant bacteria.16,17

Rather than having recourse to specialist support treatment, we seek to evaluate possible medical treatment for this pathology, on the basis of a survey of the literature and considering the rationale for each drug used, remembering that is mandatory to suspend administration of any suspect drug.


The use of corticosteroids is a much-debated question. Some reports have described a dramatic improvement in patients with TEN treated with corticosteroids.1 However, TEN can occur in patients subjected to long-term corticosteroid therapy: a study in Germany showed that 5% of patients with TEN had been on steroid therapy for at least a week before the clinical onset.17

According to Kelemen et al.15 the mortality rate in TEN is not correlated to the extent of the surface affected by the necrotic process, to the reduction in the white blood cell count, or to corticosteroid administration for more than 48 h.

Halebian compared a first group of 15 patients affected by TEN or SJS and treated with high-dose corticosteroid therapy with a second group of patients not so treated after admission to a burn unit.16 The survival rate in the second group was almost double (66%) that in the group treated with corticosteroids (33%). The incidence of septic complications was not significantly different in the two groups, but increased mortality was observed in cases of sepsis.

This study, widely reported in the literature, has been subjected to various criticisms - the Authors considered subjects affected by both TEN and SJS, without specifying their exact distribution; the doses and duration of the corticosteroid treatment were not clearly specified; and, in the group described as untreated, 11 patients (73%) received corticosteroids for an average period of 3.5 days before hospitalization. Considering the low incidence rate and the extreme severity of TEN, a clinical randomized trial focused on the effectiveness of systemic steroid therapy is difficult to realize. It must however be borne in mind that Authors who support the use of corticosteroids recommend high doses of corticosteroids (desametasone 8-16 mg daily, quickly tapering in 7-10 days) within 24-48 h, in order to prevent evolution of the process.

Fine13 recommends the cautious administration of prednisone per os (1-2 mg/kg daily) or intravenously for 3-5 days, immediately upon onset of the pathology. This approach may be useful if there are no contraindications to steroid therapy, in the hope that early treatment (within the first 72 h) in a patient with a still limited involvement of the skin surface (less than 25%) may be in time to prevent widespread diffusion.

Conversely, many reports in the literature suggest that the administration of corticosteroids in TEN can be contraindicated (increased mortality owing to septic accidents) and the consideration of other therapeutic solutions would therefore seem to be justified.

Other possible medical therapies for the treatment of TEN that are reported in the literature include the use of plasmapheresis, intravenous immunoglobulins, and cyclosporin.


Plasmapheresis is used in antibodies-mediated pathologies such as the Goodpasture syndrome and myasthenia and in phenomena associated to immune complexes such as LES and cryoglobulinaemia.

In TEN there is no evidence of the formation of immune complexes or activation of the complement system: the rationale of therapy is to identify and remove the drug responsible and its metabolites.

A number of studies confirm the effectiveness of plasmapheresis in the Lyell syndrome. Chaiademenos18 used plasmapheresis to treat seven patients affected by TEN in 30-80% of the skin surface and involving mucous surfaces. In six patients the therapeutic scheme was used on alternate days, while the seventh patient was treated daily. Each patient had one to four sessions, with the replacement of 2500 ml of plasma with 4% albumin solution. None of the patients died and there were no side effects or complications in a follow-up lasting eight years. We recommend the alternate day scheme. Although plasmapheresis is an expensive practice and requires a venous access, we consider it to be a valid method in the therapy of TEN that makes it possible to obtain early relief from pain and rapid cessation of necrolysis. The mortality rate in patients treated with plasmapheresis (< 10%) is comparable with that of patients admitted and treated in burn centres. We must not however forget that the patient is exposed to the risks related to transfusion therapy.19

Intravenous immunoglobulin

Not many cases of the use of intravenous immunoglobulins (IVIG) have been reported in the literature. Generally, doses of 400 mg/kg daily are administered for a period of 3-5 days, obtaining rapid re-epithelialization of the cutaneous and mucous lesions, resolution of the systemic features, and a reduction in the period of hospitalization.

The use of IVIG is particularly indicated in patients with HIV, who have a higher risk of TEN than the normal population and frequently take trimethoprim-sulphamethoxazole as a prophylactic measure or for treatment of pneumonia caused by Pneumocystis carinii. In such patients the use of corticosteroids is avoided, in order not to induce further immune suppression.20

The mechanism of action of intravenous immunoglobulins is not fully known: they probably inhibit the activity of the cytokines by reducing the expression of adhesion molecules and/or increasing the activity of T-suppressor cells. According to a recent study, IVIG prevent the apoptosis of keratinocytes mediated by a particular cell receptor called FAS (CD95).21

The efficacy of IVIG can be explained by their powerful anti-inflammatory activity and by the protection they provide from possible infections by common pathogen agents (sepsis is the principal cause of death in patients with TEN).21,22


Cyclosporin was introduced into the treatment of TEN in 1986.

The effectiveness of this drug in a number of conditions is well known and accepted, e.g. in the prevention of transplant rejection, therapy of rheumatoid arthritis, primary biliary cirrhosis, ulcerative colitis, Crohn’s disease, GVHD, and psoriasis. Cyclosporin inhibits the principal cellular populations involved in the pathogenesis of TEN (activated T-lymphocytes, macrophages, keratinocytes), interferes with the metabolism of TNF-·, and possesses an anti-apoptosis property, and we know that apoptosis is the mechanism that causes the death of keratinocytes in TEN.23-25

Recently Arévalo et al.10 performed a comparative study between a group of 11 patients treated with cyclosporin and another group of six patients treated with cyclophosphamide (150 mg i.v. b.i.d.) and methylprednisolone (> 1 mg/kg daily). Cyclosporin was initially given at a dose of 3 mg/kg/die per os for two weeks. After complete re-epithelialization was obtained, the drug was reduced by 10 mg every 48 h. The incidence of failure of various organs (lungs, heart, kidneys, liver) and leucopoenia requiring the use of the granulocyte colony stimulating factor (GB < 1.000/mm3) was significantly less in the group treated with cyclosporin than in the group treated with cyclophosphamide and corticosteroids.

Even though wide cutaneous surfaces were affected in all the patients, none of the patients treated with cyclosporin died, while in the other group the mortality rate was 50%.

The effectiveness of cyclosporin could be due to its capacity to interrupt the disease’s evolution, allowing early re-epithelialization of the cutaneous surfaces affected.

In this study no side effects were observed and, in particular, no significant nephrotoxicity was found: cyclosporin’s toxicity probably manifests itself at higher doses (> 15 mg/kg/day) than those habitually used in cases of TEN. In the event of hepatic insufficiency or suspicious involvement of the P450-cytochrome, monitoring of the drug’s haematic levels is recommended.


In conclusion, it is mandatory:

As TEN is a life-threatening condition, a “wait and see” approach is not possible. A pharmacological therapy is therefore necessary, the choice of which - for ethical and practical reasons, in view of the pathology’s rarity - cannot be based on wide-scale clinical trials but only on restricted observations and isolated reports.

Having evaluated the drugs used in the treatment of TEN and described in the literature, we believe that cyclosporin presents the most valid rational basis, acting at all the levels of the pathogenetic mechanism (pro-inflammatory cytokines, apoptosis of keratinocytes, T- lymphocyte activation). It is relatively safe at the dosage employed (3-4 mg/kg daily) and inexpensive.

RESUME. La nécrolyse épidermique toxique (NET) est une maladie mucocutanée commune et potentiellement dangereuse caractérisée par une éruption cutanée, des bulles et l’exfoliation diffuse de larges zones cutanées. Les Auteurs décrivent la maladie, connue aussi comme le syndrome de Lyell, et la distinguent d’autres conditions semblables. Après avoir décrit l’épidémiologie, l’étiologie et la pathogenèse de TEN, comme aussi ses caractéristiques cliniques, ils considèrent le diagnostic et la thérapie. La thérapie pharmacologique est nécessaire, particulièrement avec l’emploi de la cyclosporine.


  1. Becker D.S.: Toxic epidermal necrolysis. Lancet, 351:1417-20, 1998.
  2. Rasmussen J.E. et al.: Erythema multiforme, Stevens-Johnson syndrome and toxic epidermal necrolysis. Dermatol. Nurs.,7: 37-43, 1995.
  3. Roujeau J.C. et al. : Medication use and the risk of Stevens-Johnson syndrome or toxic epidermal necrolysis. N. Eng. J. Med., 333: 1600-7, 1995.
  4. Revuz J.: Sindrome di Lyell (Necrolisi epidermica tossica). In: “Manuale di Dermatologia e Venereologia”, Saurat J.H., Grosshans E., Laugier P., Lachapelle J.M., Italian edition, Milan, 274-5, 1995.
  5. Le Cleach L. et al.: Blister fluid T-lymphocytes during toxic epidermal necrolysis are functional cytotoxic cells which express human natural killer (NK) inhibitory receptors. Clin. Exp. Immunol., 119: 225-30, 2000.
  6. Paquet P., Pierard G.E.: Soluble fractions of tumor necrosis factor-alpha, interleukin-6 and of their receptors in toxic epidermal necrolysis: A comparison with second degree burns. Int. J. Molec. Med., 1: 459-62, 1998.
  7. Dietrich A., Kawakubo Y., Rzany B. et al.: Low N-acetylating capacity in patients with Stevens-Johnson syndrome and toxic epidermal necrolysis. Exp. Dermatol., 4: 313-6, 1995.
  8. Wolkenstein P., Carrière V., Charue D. et al.: A slow acetylator genotype is a risk factor for sulphonamide-induced toxic epidermal necrolysis and Stevens-Johnson syndrome. Pharmacogenetics, 5: 255-8, 1995.
  9. Leyva L. et al.: Anticonvulsivant-induced toxic epidermal necrolysis: Monitoring the immunologic response. J. Allergy Clin. Immunol., 105: 157-65, 2000.
  10. Arévalo J. et al.: Treatment of toxic epidermal necrolysis with cyclosporin A. J. Trauma, 48: 473-8, 2000.
  11. Smoot E.C.: Treatment issues in the care of patients with toxic epidermal necrolysis. Burns, 25: 439-42, 1999.
  12. Besana C.: Reazioni da ipersensibilità a farmaci. In: “Manuale di Medicina Interna Sistemica”, Rugarli C., Masson, Milan, vol. II: 1204-11, 1993.
  13. Fine J.D.: Drug therapy: Management of acquired bullous skin diseases. N. Eng. J. Med., 333: 1475-84, 1995.
  14. Power W.J., Ghoraishi M., Merayo-Lloves J. et al.: Analysis of the acute ophthalmic manifestations of the erythema multiforme/ Stevens-Johnson syndrome/toxic epidermal necrolysis disease spectrum. Ophthalm., 102: 1669-76, 1995.
  15. Kelemen J.J. et al.: Burn center care for patients with toxic epidermal necrolysis. J. Am. Coll. Surg., 180: 273-8, 1995.
  16. Halebian P.H. et al.: Improved burn center survival of patients with toxic epidermal necrolysis managed without corticosteroids. Ann. Surg., 204: 503-12, 1986.
  17. Pasricha J.S. et al.: Toxic epidermal necrolysis. Int. J. Dermatol., 35: 523-7, 1996.
  18. Chaiademenos G.C. et al.: Plasmapheresis in toxic epidermal necrolysis. Int. J. Dermatol., 36: 218-21, 1997.
  19. Egan C.A. et al.: Plasmapheresis as an adjunct treatment in toxic epidermal necrolysis. J. Am. Acad. Dermatol., 40: 458-61, 1999.
  20. Sanwo M. et al.: Use of intravenous immunoglobulin in the treatment of severe cutaneous drug reactions in patients with AIDS. J. Allergy Clin. Immunol., 98: 1112-5, 1996.
  21. Viard I. et al.: Inhibition of toxic epidermal necrolysis by blockade of CD95 with human intravenous immunoglobulin. Science, 282: 490-3, 1998.
  22. Magina S. et al.: A case of toxic epidermal necrolysis treated with intravenous immunoglobulin. Br. J. Dermatol., 142: 191-2, 2000.
  23. Paquet P. et al.: Would cyclosporin A be beneficial to mitigate drug-induced toxic epidermal necrolysis? Dermatol., 198: 198-202, 1999.
  24. Criton S. et al.: Toxic epidermal necrolysis: A retrospective study. Int. J. Dermatol., 36: 923-5, 1997.
  25. Zaki I. et al.: Toxic epidermal necrolysis associated with severe hypocalcaemia, and treated with cyclosporin. Br. J. Dermatol., 133: 337-8, 1995.
  26. Dobrosavljevic D., Milinkovic M.V., Nikolic M.M.: Toxic epidermal necrolysis following morbilli-parotitis-rubella vaccination. J. Eur. Academy Derm. Venereol., 13: 59-61, 1999.
  27. Rzany B., Schmitt H., Schopf E.: Toxic epidermal necrolysis in patients receiving glucocorticosteroids. Acta Derm. Venereol., 71: 171-2, 1991.
<% riquadro "This paper was received on 22 September 2001.

Address correspondence to: Dr G. Brambilla, Divisione di Medicina Interna II, Ospedale Niguarda Ca’ Granda, Milan, Italy." %>

<% footer %>