8/6/2019 03mb3_4

1/4

VOL.11 NO.5 MAY 2006

Medical Bulletin

17

VOL.13 NO.10 OCTOBER 2008

Introduction

Stevens-Johnson Syndrome(SJS) and Toxic EpidermalNecrolysis (TEN) are variants of a spectrum ofconditions characterised by erythematous maculesevolving to epidermal detachment and mucous

membrane erosions. In SJS there is less than 10% bodysurface area involvement, in TEN more than 30% and10-30% overlap cases.

It is important to be able to recognise Stevens-Johnsonsyndrome (SJS) and toxic epidermal necrolysis (TEN)and manage them properly. The mortality rate of SJSand TEN is high: even in moderately severe cases itcould be up to 30%. For those who survive, there could be troublesome late complications. Moreover, since theu se o f d ru g i s t he m os t i mp or ta nt c au se , t heidentification and removal of the causative medicationis of paramount importance to halt the progression ofthe conditions and to prevent recurrence from

inadvertent re-challenge.

Recognising SJS and TEN

We could recognise SJS and TEN early if we are familiarwith their clinical features, especially those earlier ones.Many patients with SJS and TEN begin with theprodromal symptoms of fever, headache and myalgia.The SJS and TEN skin eruptions1 first appear aserythematous then dusky or purpuric macules. Thelesions are usually irregularly shaped, discrete in the beginning then coalesce with one another. Atypicaltarget lesions could be seen but they are not the three-

zone target lesions seen in erythema multiforme. Therash first appears on the face and upper part of thetrunk and proximal part of the extremities and spreadrapidly to the rest of the body. Lesions soon developedinto flaccid blisters. For those non-blistered rash,Nikolsky sign (separation of epidermis from dermiswith lateral pressure) can be demonstrated, which is animportant though not pathognomonic sign. Finally thenecrotic epidermis comes off leaving large areas of redexudative dermis exposed.

The mucous membrane is always involved in SJS andTEN, commonly precede the rash but sometimes after.Erythema is followed by painful erosions on the buccal,

ocular and genital mucosae, and usually more then onesite are involved. More than 80% of patients haveconjunctival involvement, sometimes leads to cornealulceration, anterior uveitis and synechiae. Ocularinvolvement in SJS and TEN could result in the mostdebilitating late complications.

SJS and TEN do not limit themselves to the skin.Pulmonary and digestive system involvements are notuncommon. A quarter of patients have shortness of breath, hypoxia and haemoptysis, and the degree ofpulmonary involvement is not necessary in proportionto the degree of skin involvement. Chest X-ray couldshow features of interstit ial involvement but

differentiation from infection is important, which could b e h el pe d w it h f ib re op ti c b ro n ch os co py .Gastrointestinal tract involvement will result indiarrhoea, malena and oesophageal necrosis. Renalinvolvement will result in proteinuria, haematuria andazotaemia.

The prodrome of fever, myalgia, headache; theappearance of dusky rash on the face and proximallimb; mucosal erosion, and the positive history of drugexposure should alert the physician to the possibility ofSJS and TEN

Managing Patients with SJS and TENSJS and TEN are life threatening conditions that needintensive care with experienced physicians and specialisnurses and multidisciplinary team work. Theframework of the management is depicted in Chart 1.

Diagnosis

All suspected cases of SJS and TEN should bec o nf i rm e d b y s ki n b i op s y f o r h i st o lo g ic a ndimmunofluorescence examinations. Early lesion showssuprabasal layer apoptotic keratinocytes. Later lesionshows full-thickness epidermal necrosis and separationof epidermis from dermis. A number of important

Diagnosis and Management of Stevens-JohnsonSyndrome and Toxic Epidermal Necrolysis

Dr. HHF Ho

Dr. HHF Ho

MBBS(HK), MRCP(UK), FHKCP, FHKAM(Med)

Cheung Sha Wan Dermatological Clinic, Social Hygiene Service

Diagnosis: biopsy

Removal of cause

Specific treatment

Complications management

Fluid, electrolyte,infection

Wound care,pain control

Chart 1 Management of SJS/TEN

8/6/2019 03mb3_4

2/4

Medical Bulletin

18

VOL.13 NO.10 OCTOBER 2008

conditions mimic SJS and TEN (Table 1) hence ahistological evidence is important. Since 90% SJS andTEN has mucous membrane involvement the absence ofsuch should prompt one to consider alternativediagnosis.

Erythema multiforme (EM) could easily be mixed withSJS since both present with rash and oral mucosal

erosion. The classification by Bastuji-Garin2 separatederythema multiforme from SJS/TEN although it is notuniversally agreed. EM is different from SJS and TEN inmany ways (Table 2). Infection is the major cause of EMand the commonest implicated infections are Herpessimplex and Mycoplasma pneumoniae although someother infectious agents have been reported, whereasdrug is considered as an uncommon cause. The typicaltarget lesions in EM have three concentric zones: centraldusky disk, middle pale ring, outermost erythematoush al o a nd t he y a re n ot f ou nd i n S JS a nd T EN .Characteristically all lesions of EM are papular and inacral distribution at least initially whereas in SJS andTEN rash start on face and proximal limbs. Although in

EM there could also be mucosal involvement they aremostly limited to oral mucosa.

Staphylococcal scalded skin syndrome (SSSS) presentsinitially as a macular exanthema which might quicklyevolve to blistering eruption with positive Nikolsky'ssign and mimic SJS and TEN. SSSS more commonlyoccurs in infants or adults with renal failure. A Tzancksmear will find acantholytic cells in SSSS but not TEN.Skin biopsy with frozen section examination will findintradermal cleavage with acantholysis in thesubgranular layer whereas in SJS and TEN full-thickness epidermal necrosis and dermal-epidermalseparation are found. The diagnosis of SSSS instead ofSJS/TEN will enable the early use of antibiotics againstStaphylococcus.

Erythema multiforme major

Staphylococcal scalded skin syndrome

Purpura fulminant

Disseminated intravascular coagulation with skin necrosis

Acute generalised exanthematous pustulosis

Generalised bullous fixed drug eruption

Chemical toxicity (methotrexate, colchicines etc)

Burns

Graft-versus-host disease

Pemphigus

Removal of Cause

In 70% of SJS and TEN drug cause could be identifiedand more than 100 agents have been reported3. Drughistory taken carefully and repeatedly, involving familymembers, enquiring family doctors, and taking over-the-counter non-prescription items into consideration are

necessary before the causative agent can be identified.Since many patients could be taking several agents at thesame time, the true causative agent could be hard toisolate. The temporal relationship between the intake ofthe agent and onset of condition is an important factor.SJS and TEN usually begins less than 8 weeks but morethan 4 days from the first intake of the agent. Look fordrugs that were added within this period. Only re-challenged drugs will elicit the condition in a few hours.Some medications have higher risk of causing SJS andTEN whereas in some other medications SJS and TENhas not been reported (Table 3). Infection is not acommon cause of SJS and TEN although there have beencase reports of Mycoplasma pneumonia. The identified

culprit should be removed immediately and labelled"allergic" so that it would not be re-challengedinadvertently. In case of complicated drug history and adefinite single causative agent could not be identified,only the necessary medications should be retained.

Specific Treatment

SJS and TEN are life threatening conditions. The successof treatment depends on early recognition of thecondition, prompt removal of the causative medicationsand intensive supportive care in a well-equippedhospital.4 Several agents with anti-inflammatory or

immunosuppressive properties have been tried to alterthe course of the disease but no single agent has theirefficacy clearly proven by clinical trials.

1. Intravenous Immunoglobulin (IVIG)Prepared from pooled plasma, IVIG contains immuneantibodies that interfere with the apoptotic pathwaymediated by the Fas ligand and receptor. Theoretically itis best to give IVIG early (within 24-72 hours from firstappearance of bullae)4,5 before Fas ligand and receptorbinding has occurred, although it may still be effective ifnew bullae are still appearing. Sucrose-depleted IVIG ispreferred since it has lower possibility of renal toxicity.Patient with IgA deficiency will develop anaphylaxis toIVIG. It is best to obtain a patient's IgA level beforeadministering but awaiting the report might delaytreatment. History of recurrent sinopulmonary infectionand gastrointestinal infection may help to identify thosewith IgA deficiency which is very rare.

Table 1 Mimickers of SJS and TEN

Table 2 Comparing EM and SJS/TEN

EM SJS/TENInfection: Herpes simplex,Mycoplasma pneumoniae

Papular erythematous lesions

Typical targets with three zones

Aral distribution initially

70-% mucosal involvement butlimited to oral mucosa usually

Fever and constitutional symptomsabsent

Most

8/6/2019 03mb3_4

3/4

VOL.11 NO.5 MAY 2006

Medical Bulletin

19

VOL.13 NO.10 OCTOBER 2008

Results of studies of IVIG on SJS and TEN has beenconflicting, and IVIG should not be considered as aroutine treatment. Some studies have suggested thehigher dose of 3g/kg total dose given over 3 days hasbetter effects over the lower dose of 2g/kg total dose.

2. Systemic CorticosteroidSome studies have advocated the use of systemiccorticosteroids in the early stage of SJS and TEN. Otherstudies failed to prove the effect of the agent and havedemonstrated an increase in the chance of sepsis andother complications. Balancing available evidencee sp ec ia ll y t he m or e r ec en t o ne s, 1 systemiccorticosteroids cannot be recommended in TEN. Its usei n S JS i s s ti ll con tr ov er si al b ut s ho ul d n ot b erecommended when extensive skin loss has alreadyoccurred.

3. Cyclosporin ASupported by favourable outcomes6 in several casereports and series, which used cyclosporin A at a dose

of 3-4mg/kg/day in short term, thus avoiding its sideeffects which commonly occur in long term use, thisagent seems promising but more comprehensive studiesare needed.

4. Other AgentsTheoretically removal of the offending medication, itsmetabolites or cytokines by plasmapheresis orhaemodialysis could help the improvement of SJS andTEN. However the lack of good clinical evidence andthe risk of sepsis associated with in-dwelling catheterdoes not support these as recommendable treatments.Thalidomide7based on its anti-TNF effect has been triedbut the study was prematurely terminated since excessmortality was reported.

Management of Fluid, Electrolyte,Respiration and Infection

In the absence of proven effective specific agent, thesuccess in treating SJS and TEN depends very much onsupportive care. Since SJS and TEN can deterioraterapidly, intensive care unit or burn centre care isrecommended. Fluid loss and electrolyte imbalanceshould be closely monitored and corrected. Peripheralline is more recommendable than central, which has a

higher chance of infection, but good peripheral venousaccess is difficult to find. All lines should be checked forsigns of infection daily and changed two times a weekwith tips of lines and catheters sent for culture.Nutrition support with nasogastric tube helps healing.Respiratory rate and oxymeter monitoring areimportant. Raised urea level, blood glucose above14mm/L and neutropenia are unfavourable prognosticfactors and should be monitored.

Sepsis is the main cause of death. Cultures should betaken frequently from the cutaneous erosions, mucosalerosions, blood and urine to obtain the microbiologyand their sensitivity profile. Signs of infection should bemonitored closely and systemic antibiotics should bepromptly administered when signs of infections (feveror falling body temperature, rigour, hypotension,decrease in urinary output, respiratory failure, poorglycaemic control and impaired consciousness, etc) are

detected. Prophylactic antibiotic is contraindicated sincethis will encourage the appearance of resistant strains.

Wound Care and Pain Control

Painstaking wound care is the backbone to managementof SJS and TEN.1,4 Good wound care reduces the chanceof infection and pain. There is no standard protocol onthe wound dressing. Various non-stick dressing has beenused but sulfa-containing material should be avoided toprevent systemic sensitisation and leucopenia. Use air-fluidised mattresses to prevent pressure sore. Theenvironment temperature is maintained at 28-30 degreeCelsius to prevent hypothermia. Debridement of necroticepidermis is not necessary. Adequate pain control manya time needs morphine group of analgesics. Respiratorydepression should be watched out if opiates are used.

Oral mucosal ulceration is very painful. Chlorhexidinerinses help in maintaining good hygiene and white-soft

paraffin on the lips relieves the pain. Complications ont he e ye s c ou ld r es ul t i n b li nd ne ss a nd a nophthalmologist's care is necessary. Artificial tears,antibiotics eye drops every two hourly and mechanicaldisruption of early synechiae is needed.

Complications

Sepsis is the most important cause of mortality. Extensiveerosions put patients at risk of infection by bacteria andfungi which will result in pulmonary complications andmulti-organ failure. If respiratory failure develops,ventilation support is needed.

Late ophthalmic complications are seen in up to 75% ofp at ie nt s, h en ce e ar ly t re at me nt i s n ee de d.Hyperpigmentation and hypopigmentation are commonand sometimes scars and nail dystrophy may result.Genital adhesions resulting in dyspareunia, pain and bleeding should be watched out. Gastrointestinal, bronchial, urethral and anal complications are lesscommon. Post-traumatic stress disorder is also possibleand some patients may need psychiatrist's care. Allpatients recovering from SJS and TEN should befollowed-up for development of these complicationswhich could be delayed but debilitating.

Prognosis

Depending on the severity, the clinical course of SJS andTEN may last up to a few weeks. It should be noted thatthe prognosis is not related to type or dose of thecausative medication. A SCORTEN prognostic scoringsystem8 has been developed to correlate mortality withselected parameters.

Pro gno stic factors Po ints SCORTEN Mo rtalityRate

Age > 40

Heart rate >120/minCancer or haematologic malignancy

>10% body surface area

Serum urea >10mm/L

Serum bicarbonate14mm/L

1

11

1

1

1

1

0-1

23

4

>5

3.2%

12.1%35.8%

58.3%

90%

8/6/2019 03mb3_4

4/4

Medical Bulletin

20

VOL.13 NO.10 OCTOBER 2008

Conclusion

Successful management of SJS and TEN requires theearly recognition of the conditions, diagnosis with biopsy, identification and removal of the causativedrugs and intensive multidisciplinary management in ahospital with experienced medical and nursing

personnel.

Chave TA, Mortimer NJ et al: Toxic epidermal necrolysis: currentevidence, practical management and future directions. Br JDermatol 2005;153:241-253Bastuji-Garin S et al: A clinical classification of cases of toxicepidermal necrolysis, Stevens-Johnson syndrome and erythemamultiforme. Arch Dermatol 1993;129:92-96

1.

2.

References

Stern RS, Chan HL: Usefulness of case report literature indetermining drugs responsible for toxic epidermal necrolysis. J AmAcad Dermatol 1989;21:317-322Fromowitz J, Ramos-Caro et al: Practical guidelines for themanagement of toxic epidermal necrolysis and Stevens-JohnsonSyndrome. Int J Dermatol 2007;46:1092-1094Prins C, Kerdel FA et al: Treatment of toxic epidermal necrolysiswith high-dose intravenous immunoglobulins: multicenter

retrospective analysis of 48 consecutive cases. Arch Dermatol 2003;139: 26-32Zaki I, Patel S et al: Toxic epidermal necrolysis associated withsevere hypocalcaemia, and treated with cyclosporine. Br J Dermatol1995; 133:337-8Wolkenstein P, Latarjet J, Roujeau J-C et al: Randomizedcomparison of thalidomide versus placebo in toxic epidermalnecrolysis. Lancet 1998;352:1586-9Bastuji-Garin S, Fouchard N et al: SCORTEN: a severity-of-illnessscore for toxic epidermal necrolysis. J Invest Dermatol 2000;115:149-153

3.

4.

5.

6.

7.

8.