Adverse Drug Reaction
Retrospective analysis of Stevens-Johnson syndrome and toxic epidermal necrolysis over a period of 10 years
Abarna Devi Sanmarkan, Tukaram Sori, Devinder Mohan Thappa, TJ Jaisankar
Department of Dermatology and STD, Jawaharlal Institute of Postgraduate Medical Education and Research (JIPMER), Pondicherry - 605 006, India
Date of Web Publication: 10-Mar-2011
Devinder Mohan Thappa
Professor and Head, Department of Dermatology and STD, JIPMER, Pondicherry India
Background: Stevens-Johnson syndrome (SJS), and toxic epidermal necrolysis (TEN), are the acute emergencies in dermatology practice. Prompt diagnosis and management may reduce the morbidity and mortality in SJS/TEN patients. Early identification of the offending drug is necessary for early withdrawal and to prevent the recurrences of such a devastating illness. Aims : To study the demography, offending agents, clinical and laboratory features, treatment, complications, morbidity and mortality of SJS/TEN in our hospital. Materials and Methods: In this retrospective study, we reviewed the medical records of SJS, TEN, SJS/TEN overlap of inpatients over a period of 10 years Results: Maximum number of SJS/TEN cases were in the age group of 11-30 years. Males predominated in the SJS group with a ratio of 1.63:1, whereas females predominated the TEN group with a ratio of 1:2.57. Nonsteroidal anti-inflammatory drugs (NSAIDs) were the commonest group of drugs among the SJS group in 5/21 patients (23.8%). Antimicrobials were the commonest group of drugs causing TEN in 11/25 patients (44%). Mucosal lesions preceded the onset of skin lesions in nearly 50%. Our study had one patient each of SJS/TEN due to amlodipine and Phyllanthus amarus, an Indian herb. The most common morbidity noted in our study was due to ocular sequelae and sepsis leading to acute renal failure respectively. Kaposi's varicelliform eruption was found in three of our patients. Conclusion: Antimicrobials and NSAIDS are the common offending agents of SJS/TEN in our study.
How to cite this article:
Sanmarkan AD, Sori T, Thappa DM, Jaisankar TJ. Retrospective analysis of Stevens-Johnson syndrome and toxic epidermal necrolysis over a period of 10 years. Indian J Dermatol 2011;56:25-9
How to cite this URL:
Sanmarkan AD, Sori T, Thappa DM, Jaisankar TJ. Retrospective analysis of Stevens-Johnson syndrome and toxic epidermal necrolysis over a period of 10 years. Indian J Dermatol [serial online] 2011 [cited 2011 Apr 16];56:25-9. Available from: http://www.e-ijd.org/text.asp?2011/56/1/25/77546
Stevens-Johnson syndrome (SJS), SJS/TEN overlap, and toxic epidermal necrolysis (TEN), are variants of acute, rapidly progressive mucocutaneous reactions which differ only in their body surface area (BSA) involvement.  Drugs, infections, vaccines, radiological contrasts, vaginal suppositories,  acrylonitrates, graft versus host reaction,  and lupus erythematosus  in predisposed individuals result in immunologically mediated keratinocyte apoptosis, and hence, extensive necrosis and detachment of epidermis with significant morbidity and mortality. Several studies and case reports on SJS/TEN have been done in India as well as abroad since the description of SJS case by Stevens and Johnson in 1922  and TEN case by Lyell's in 1956.  A hospital-based retrospective study was done to study the demography, offending agents, clinical and laboratory features, treatment, complications, morbidity and mortality in SJS/TEN in our hospital.
Materials and Methods
Medical records of all inpatients, admitted with a diagnosis of SJS, TEN, and SJS/TEN overlap over a period of 10 years from June 98 to July 08, in the Department of Dermatology, at our institute were reviewed. Of the total 56 medical records, 46 with complete data were selected. Bastugi's criteria  formed the basis for classifying these severe mucocutaneous reactions into SJS, TEN and SJS/TEN overlap. Parameters like age, sex, co-morbid conditions, etiology, clinical features, past history of drug reactions, period of hospital stay, investigations, treatment modalities, course and outcome of these 46 patients were recorded and analyzed.
Age and gender
The number of patients with SJS and TEN was 21 each and with SJS/TEN overlap was four. Four patients in overlap group were included in TEN group. Maximum number of SJS cases was in the age group of 11-20 years and the maximum number of TEN cases was in the 20-30 years age group. Males predominated the SJS group with a ratio of 1.63:1, whereas females predominated in the TEN group with a ratio of 1:2.57. The youngest patient was a 3-year-old child and the oldest was a 78-year-old male.
Drug as an etiology was established in almost all the cases except in one case of SJS which was due to herpes labialis. The implicated drug was identified in only 36 cases. The major group of drugs causing SJS/TEN in our study was antimicrobials 15/46 (32.6%), followed by nonsteroidal anti-inflammatory drugs (NSAIDs) 12/46 (26.08%) and antiepileptic drugs 8/46 (17.3%). Besides this, there were two cases of TEN due to paracetamol and one case each due to nevirapine, isoniazid, fansidar (pyrimethamine and sulfadoxine), amlodipine, siddha medication, and Phyllanthus amarus (kizhanelli), a common medicinal plant used for hepatitis B in south India.
Considering the mucocutaneous reactions individually, NSAIDs were the commonest group of drugs among the SJS group in 5/21 patients (23.8%). Antimicrobials like fluroquinolones, penicillin group of drugs were the commonest group causing TEN in 11/25 patients (44%). At least more than eight patients (17.3%) developed SJS/TEN following consumption of over-the-counter medications for analgesia. Anacin, analgen (contains acetylsalicylic acid, paracetamol, codeine phosphate, and caffeine) and Vicks action 500 (contains paracetamol, phenylpropanolamine, and caffeine) were the common over-the-counter medications. The drugs which were implicated in causing SJS/TEN in our patients are shown in [Table 1] .
Table 1: Agents involved in SJS/TEN
Past history of drug reaction was present in seven cases of TEN. Nature of past and present reaction in patients with previous history of drug eruption is shown in [Table 2] .
Table 2: Nature of past and present reaction in patients with previous history of drug eruption
The co-morbid conditions for which our patients were taking these offending drugs were seizures (8), diabetes mellitus (5), hypertension (3), HIV (2), pulmonary tuberculosis (2), enteric fever (2), one case each of syphilis, non Hodgkin's lymphoma, hypothyroidism, astrocytoma, hemiparesis, carcinoma cervix, carcinoma breast, lichen planus and hepatitis B.
The mean duration between the drug intake and the onset of reaction was 6 days. The reaction time was less than a week in most of our patients [29/46 (63%)].
Fever, sore throat, myalgia and burning sensation were the major prodromal symptoms experienced by our patients. Anaphylaxis was observed in one of the SJS patient. SJS patients had either blanchable or non blanchable purpuric macules, typical and atypical targets lesions, vesicles, bullae and erosions over erythematous and urticated plaques, involving less than 10% of BSA. TEN patients presented with sudden onset of large sheets of epidermal necrosis along with severe constitutional features. Nikolsky's sign and skin tenderness were present in all TEN patients. Four TEN patients had crusted erosions over the scalp.
Mucosal lesions preceded the onset of skin lesions in nearly 50% of our patients. Also, 87% of SJS and 88% of TEN had oral mucosal involvement. All three (oral, conjunctiva, genital) mucosa were involved in 20/46 (43.47%) patients. Oral mucosal involvement was the most common followed by the conjunctiva. Other mucosal involvement noted was that of pharyngeal and anal. Abnormal laboratory parameters noted in our cases are given in [Table 3] .
Table 3: Laboratory parameters
Most of our patients had neutrophilic leukocytosis and raised erythrocyte sedimentation rate (ESR). Eosinophilia was seen in only three patients. Three TEN patients had thrombocytopenia and leukopenia. Neutropenia was noticed in two patients. Elevated liver enzymes were noticed in 18 of our patients.
Systemic complications observed in our study were acute renal failure (4), septicemia (4, all died, none on dexamethasone pulse therapy), metabolic encephalopathy (1), lower respiratory infection (2), congestive cardiac failure (2), pulmonary edema (2), hyperkalemia, intracranial bleed, aspirational pneumonia, psychosis, lung abscess, wound infection commonly due to Staphylococcus aureus (MRSA), followed by Pseudomonas, Enterococcus, Klebsiella. One case of herpes labialis and three cases of Kaposi's varicelliform eruption developed during the regression of the reaction. Post inflammatory pigmentation was noted in almost all cases. Other mucocutaneous complications that we noticed were oral candidiasis, gastritis in the form of hematemesis, phimosis parotitis, tonsillitis, and epiglottitis. The number of patients with ocular morbidity in our study was 23/46 (50%) comprising of symblepharon (23), corneal scarring, bacterial, viral conjunctivitis, blepharitis, and corneal xerosis.
Except for seven patients, all other patients received definitive therapy in the form of dexamethasone (17), prednisolone (11), methyl prednisolone pulse (2), dexamethasone pulse (2) and cyclosporine (7). Pulse therapy was given along with intravenous broad spectrum antibiotics. Overall, healing was noticed at 2-15 days after the onset of treatment. For dexamethasone pulse therapy, the onset of healing was on the third day of pulse, and for methylprednisolone pulse therapy, it was on the second day of pulse. The period of hospitalization ranged from 5 to 30 days, and it was lowest for methylprednisolone pulse therapy. Apart from hemoptysis due to gastritis in one patient, we did not record any other complications including sepsis and septicemia, in patients on dexamethasone pulse therapy.
The SCORTEN score for 21 TEN cases ranged from 0 to 5. Five of the SJS/TEN cases died giving a mortality rate of 10.86% (5/46). The cause for death in all our patients was sepsis leading to acute renal failure [Table 4] .
Table 4: The profile of patients who died of complications
This study was based on inpatient records; we did not record long-term sequelae in such patients.
Acute disseminated epidermal necrosis (ADEN 1, 2, 3) is the recently proposed terminology by Ruiz-Maldonado for SJS, overlap, and TEN, respectively. SJS/TEN are more common among individuals with lowered ability to detoxify reactive metabolites or alteration in detoxifying enzymes due to genetic basis (HLA-B 12, HLA-A29, HLA-DR7),  or functional basis (enzyme dysfunction due to AIDS or other disease).  Drug and/or their reactive metabolites behave as haptens and render the keratinocyte antigenic by binding to them. This elicits immune reactions resulting in extensive keratinocyte apoptosis which is mediated by cytotoxic T lymphocytes induced caspase cascade through perforin-granzyme route, FAS (CD95)-FAS ligand pathway, or nitric oxide synthase route in susceptible individuals.