Report
A clinicotherapeutic analysis of Stevens–Johnson syndrome and toxic epidermal necrolysis with an emphasis on the predictive value and accuracy of SCORe of Toxic Epidermal Necrolysis Shuchi Bansal, MD, Vijay K. Garg, MD, MNAMS, Kabir Sardana, MD, DNB, MNAMS, and Rashmi Sarkar, MD, DNB, MNAMS
Department of Dermatology, Maulana Azad Medical College and Lok Nayak Hospital, New Delhi, India Correspondence Kabir Sardana, MD, DNB, MNAMS 466, Sector 28 Noida Uttar Pradesh 201303 India E-mail: [email protected] Funding: None.
Abstract Background The SCORTEN scale (SCORe of Toxic Epidermal Necrolysis) is widely used to predict mortality in patients with Stevens–Johnson syndrome (SJS) or toxic epidermal necrolysis (TEN). Although it is largely a useful tool, the predictive ability of the scale is variable, and modifications to the existing scale have been suggested. In addition, there is controversy regarding the roles of active therapy and conservative management. Objectives This study was conducted to assess outcomes in SJS/TEN patients managed conservatively and to analyze the predictive performance of SCORTEN. Methods Sixty patients were studied prospectively from October 2008 to September 2011. The accuracy of SCORTEN in predicting mortality was analyzed on days 1, 3, and 5 of admission. All patients were managed conservatively. Results The discriminative power of SCORTEN was acceptable, being best on day 5. The
Conflicts of interest: None.
agreement between actual and predicted mortality was poor, as indicated by a low P-value of the Hosmer–Lemeshow statistic. Only three parameters (heart rate, blood urea, and serum bicarbonate) were found to be significant on multivariate analysis, and all of these represented components within the original SCORTEN scale. The mortality rate was 16.7%. All deaths occurred within the first two weeks (8–12 days). Conclusions Serial analysis using SCORTEN is better than analysis performed only on day 1. Although the scale is largely useful, its performance is influenced by the demographic profile of the study population; minor modifications based on the population studied may increase the predictive accuracy of the original SCORTEN. In addition, conservative management is a valid therapeutic option, is preferable to treatment with steroids and immunosuppressants, and is highly recommended.
Introduction
e18
Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are life-threatening drug-induced reactions characterized by the occurrence of distinctive skin lesions, severe mucosal involvement, and systemic symptoms; severity is categorized based on the percentage of epidermal detachment.1 Drugs have been implicated as the most important etiological factors.2–5 Mortality ranges from 5% in SJS to 10–15% in the SJS–TEN overlap and 30–40% in TEN.1–5 Substantial endeavor has been made to determine the predictors, both clinical and laboratory-based, of prognosis in patients with SJS and TEN.6 The SCORTEN scale (SCORe of Toxic Epidermal Necrolysis), which provides International Journal of Dermatology 2015, 54, e18–e26
a specific score indicating the severity of illness, was developed and validated by Bastuji-Garin et al.,7 in 2000 to determine disease severity and prognosis in patients with SJS or TEN. The scale is based on seven parameters, each of which is given a score of 1 and calculated on day 1 of admission. A score of 1–7 predicts a probability of mortality ranging from 3.9 to 98.5% as calculated using the equation to predict death proposed by the authors.7 Since its development, SCORTEN has been widely used to predict mortality in patients with SJS or TEN, as well as to help determine the efficacy of therapeutic interventions.8–10 Although the scale performs well, there are issues regarding its predictive power as well as its accuracy.11–13 In a study of 144 patients conducted by Guegan et al.,11 SCORTEN was found to perform best on day 3, rather ª 2013 The International Society of Dermatology
Bansal et al.
than on day 1 as originally proposed by the authors. Vaishampayan et al.14 and Sekula et al.13 have called for a reevaluation of the existing SCORTEN parameters and have also suggested modifications to the original scale. Spornraft-Ragaller et al.15 observed a considerable overestimation and concluded that SCORTEN did not perform well in severely affected patients. Imahara et al.16 found that the predictive performance of SCORTEN is influenced by the treatment protocol used, which has changed significantly since the introduction of the scale. Notwithstanding the pioneering work of its original authors7 and its validity,8,9 there is a need to fine-tune the use of the scale. As no studies have as yet attempted to confirm the accuracy of SCORTEN in a large cohort of Indian patients, we aimed to analyze the predictive performance of SCORTEN in SJS and TEN patients using analysis on days 1, 3, and 5 of admission and to analyze other prognostic factors that have been suggested by other authors but not included in SCORTEN. The clinical course and final outcomes in patients were assessed using a conservative line of management with the aim of comparing our therapeutic outcome with those reported by other studies to date. Materials and methods Study design and settings This prospective observational study was conducted in the Department of Dermatology, Venereology, and Leprology in a tertiary care hospital from October 2008 to September 2011. Patients Sixty-four consecutive patients with SJS or TEN were admitted to this department from October 2008 to September 2011. The diagnosis of disease was based upon the widely accepted classification system suggested by Bastuji-Garin et al.1 Because some of these patients had been referred from primary centers, only those patients who had not been treated with any modality other than conservative therapy were included in the study. Four patients did not fulfill this criterion and were thus excluded. Study protocol Patients were subjected to a detailed assessment of their clinical history and a thorough general physical as well as mucocutaneous examination. In addition to routine hematological and biochemical investigations, all the necessary investigations required for analysis of SCORTEN were carried out. A plain skiagram of the chest along with culture samples from skin, blood, and urine were obtained in all patients. Patients were managed using a conservative protocol that included the withdrawal of the suspected causative drug, fluid replacement, nutritional support, temperature regulation, and the provision of daily dressings and prophylactic antibiotics. ª 2013 The International Society of Dermatology
Therapeutic and SCORTEN analysis in SJS/TEN
Report
SCORTEN scores were calculated for all patients on days 1, 3, and 5 of admission. Patients were hospitalized until complete resolution of skin and mucosal lesions and were subsequently followed for a further three months after discharge.
Endpoint The primary endpoint for evaluation was mortality rate at 1 month from the onset of SJS or TEN. This was compared with SCORTEN-predicted mortality as calculated using the predictive death equation: PðdeathÞ ¼ e logit 1 þ e logit where logit = 4.448 + 1.237 (SCORTEN).
Statistical analysis The discriminative power of SCORTEN was assessed by the area under the receiver operator characteristic (ROC) curve. The area under the curve (AUC) ranges from 0 to 1; a value of 1 indicates perfect discrimination by the test, whereas a value of 30% Delay in hospitalization (days)
suggested
to
augment
Vaishampayan et al.14 Palmieri et al.17 Revuz et al.6 de Prost et al.18 Revuz et al.6 Revuz et al.6 Hague et al.12 Namdar et al.19 Sekula et al.13 Palmieri et al.17
International Journal of Dermatology 2015, 54, e18–e26
e19
e20
Report
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
three years. Twenty patients had been referred from various primary centers, and the remaining 40 had been admitted directly to our center. Final diagnoses were SJS in 30 patients, overlapping SJS and TEN in 10 patients, and TEN in 20 patients. The mean standard deviation age of the patients was 23.3 4.3 years (range: 7–54 years). A positive history of drug intake was present in 96.6% of patients. Anticonvulsants represented the most common causative drug (43.3% of patients), followed by analgesics (26.6% of patients). Most of the patients had deranged hematological and biochemical results. Bacterial skin colonization was present in 26 (43.3%) patients, nine of whom also had systemic infection. Four patients showed pulmonary infiltrates on chest x-ray. Coexistent morbidities were present in 11 (18.3%) patients at the time of disease onset in the form of tuberculosis, cardiovascular disease, leprosy, diabetes, and hepatitis. Malignancy workup was negative in all patients.
respectively, showed that although the SCORTEN values for deceased patients were, as expected, significantly higher than those for surviving patients at all three timepoints, the difference was most significant at day 5 (P < 0.001; Table 3). A comparison of observed and predicted mortalities for each score value showed poor agreement, as indicated by the significantly low P-values (30% also had a better correlation with mortality than did epidermal detachment of >10% (P = 0.070 and P = 0.299, respectively). Deranged electrolytes, skin colonization, and delayed hospitalization did not show any significant association with mortality. Multivariate analysis
All of the parameters found to be significant on univariate analysis (both the original SCORTEN parameters, as well as the additional parameters) were included in multiple regression analysis. For the SCORTEN parameters, only day 5 values were entered in the logistic regression because the discriminative power of SCORTEN was found to be best at that time-point. Respiratory involvement, although found to be significantly important on univariate analysis, could not be included in the logistic regression because none of the surviving subjects had respiratory involvement. After backward elimination, only three parameters remained significant on multivariate analysis: blood urea, serum bicarbonate, and heart rate (Table 5). Discussion Stevens–Johnson syndrome and TEN are life-threatening conditions that contribute to a large percentage of the ª 2013 The International Society of Dermatology
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
Report
Table 3 SCORTEN values and corresponding mortality (days 1, 3 and 5) Day 1
Day 3 Deaths, actual/ predicted, n
SCORTEN score
Patients, n
0 1 2 3 4 5 6 Total Mean SCORTEN (P-value) SMR (95% CI) AUC H–L test
0 – 26 1/1.04 20 2/2.40 8 6/2.60 3 0/1.86 3 1/2.60 0 – 60 10/10.50 1.78a/2.80b (0.002) 0.952 (0.362–1.54) 0.787 0.001
Patients, n
Day 5 Deaths, actual/ predicted, n
0 – 28 1/1.12 19 2/2.28 7 5/2.24 3 1/1.86 3 1/2.60 0 – 60 10/10.10 1.70a/2.90b (0.001) 0.990 (0.376–1.60) 0.811 0.013
Patients, n
Deaths, actual/ predicted, n
0 – 30 1/1.20 17 1/2.04 6 4/1.92 4 2/2.48 3 2/2.55 0 – 60 10/10.19 1.60a/3.30b ( 75%), it was best at day 5, which is contrary to the findings of the original study.7 A similar observation in a sample of Indian patients was made by Vaishampayan et al.,14 who found the performance of SCORTEN to be best on day 5, although the small sample size of their study precluded any statistical analysis. Guegan et al.11 found the accuracy of SCORTEN for predicting mortality to be best on day 3 of admission, whereas a study in Taiwanese patients, conducted by Ho et al.,22 found it to be best on day 1 of admission. We hypothesize that this variation is associated with the fact that the acute phase of SJS/TEN lasts 8–12 days, during which the disease progresses,25 and that the SCORTEN scale will perform best when the disease is at its peak. The mean delay in admission in our patients was 2.5 days, whereas mean delays of 5.2 and 6.75 days, respectively, were reported by Guegan et al.11 and Ho et al.22 Moreover, the majority (n = 40/60) of patients in the present study presented to us directly and were not referred from other centers. Hence, the disease reached its peak in our patients a few days further into hospital admission than in the two International Journal of Dermatology 2015, 54, e18–e26
e21
Report
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
90 SCORTEN-predicted mortality
80
Mortality on day 1 70 Mortality on day 3 Mortality (%)
e22
60 Mortality on day 5 50 40 30 20 10 0 1
2
3 SCORTEN score
4
5
Figure 2 Line diagram comparing mortality rates on days 1, 3 and 5 with SCORTEN-predicted mortality and showing that mortality is underestimated for scores of 2 and 3, and overestimated for scores of 4 and 5
Table 4 Univariate analysis of factors prognostic for mortality
Prognostic factor
Surviving patients(n = 50) (%)
Deceased patients(n = 10) (%)
P-value
Age >40 yearsa Presence of malignancya Heart rate >120 bpma Epidermal detachment >10%a Blood urea >28 mg/dla Serum glucose >252 mg/dla Serum bicarbonate 30%b Delay in hospitalization (days)b
15/50 0/50 12/50 23/50 13/50 4/50 13/50 6/50 20/50 4/50 5/50 0/50 8/50 14/50 Mean
2/10 (20) 0/10 (0) 7/10 (70) 7/10 (70) 8/10 (80) 1/10 (10) 8/10 (80) 5/10 (50) 6/10 (60) 5/10 (50) 1/10 (10) 4/10 (40) 2/10 (20) 6/10 (60) Mean 2.40
0.709 _ 0.008 0.299 0.002 1.0 0.002 0.013 0.305 0.004 1.0 120 bpm
44.133 25.445 6.823
0.000 0.001 0.050
3.787 3.237 1.920
a
After applying backward elimination.
International Journal of Dermatology 2015, 54, e18–e26
earlier studies,11,22 in which the disease was already at its peak at admission. Thus, the duration of disease from the time of onset, rather than the duration of hospital stay, is crucial in influencing the predictive performance of SCORTEN. Consequently, a serial analysis of SCORTEN would serve as a better indicator of mortality, especially in patients who present in the early stages of the reaction. Analyses of individual SCORTEN parameters showed that only three parameters (heart rate, blood urea, and serum bicarbonate) had a significant association with ª 2013 The International Society of Dermatology
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
Report
Table 6 Seminal studies on the efficacy of SCORTEN in patients with Stevens–Johnson syndrome (SJS) or toxic epidermal necrolysis (TEN) Results Authors (year)
Patients, n
SMR (95% CI)
AUC
H–L statistic
Conclusions
Trent et al.8 (2004)
24 TEN patients
0.912 (0.393–1.800)
–
–
Guegan et al.11 (2006)
144 Patients
–
>80%
>0.30 Highest on day 3
Vaishampayan et al.14 (2008)
10 SJS–TEN/TEN
Statistical analysis not done
Cartotto et al.9 (2008)
61 Patients 7% SJS, 39% overlap, 54% TEN
1.17 (0.695–1.853)
0.82
1.381
George et al.20 (2008)
145 Patients (including erythema multiforme)
–
0.762
–
Morales et al.21 (2010) Ho et al.22 (2010)
82 Patients
Correlation coefficient 0.97, P = 0.005
51 Patients 37 SJS, 8 overlap, 6 TEN
1.12 (0.48–2.2)
>80% on days 1–5
Highest on day 1 (0.88)
Sekula et al.13 (2011)
166 Patients 72 SJS, 59 overlap, 35 TEN
–
0.75
–
Hu et al.27 (2013)
101 TEN patients
1.625 (1.229–2.109)
–
–
Present study (2013)
60 Patients 30 SJS, 10 overlap, 20 TEN
Day 1: 0.952 (0.362–1.54) Day 3: 0.990 (0.376–1.60) Day 5: 0.990 (0.376–1.60)
Day 1: 0.787 Day 3: 0.811 Day 5: 0.861
Day 1: 0.001 Day 3: 0.013 Day 5: 0.095
No significant difference between actual and SCORTEN-predicted mortality SCORTEN performance during the first 5 days of hospitalization was excellent and best on day 3 Systemic diseases should be included in SCORTEN BSA >30% should be given greater weight SCORTEN is an accurate scoring system for estimation of mortality among SJS–TEN patients Discriminative power of SCORTEN was better than that of the APACHE II score (AUC = 0.737), but not as good as that of ICNARC (AUC = 0.795) SCORTEN value correlates well with mortality rate Discrimination and calibration of SCORTEN were good and best on day 1 SCORTEN remains the tool of choice, but the auxillary score might be an alternative in retrospective settings with missing laboratory data Mortality rate was directly proportional to the SCORTEN score Renal insufficiency and failure significantly affected outcome Discriminative power of SCORTEN is good and best on day 5, but the calibration between observed and predicted mortalities is poor Performance of SCORTEN is influenced by the demographic profile of the study population
AUC, area under the receiver operator characteristic curve; SMR, standardized mortality ratio; 95% CI, 95% confidence interval; H–L test, P-value of the Hosmer–Lemeshow goodness-of-fit test; BSA, body surface area; APACHE, Acute Physiology and Chronic Health Evaluation; ICNARC, Intensive Care National Audit and Research Center.
ª 2013 The International Society of Dermatology
International Journal of Dermatology 2015, 54, e18–e26
e23
e24
Report
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
mortality, whereas age, blood glucose, and epidermal detachment of >10% did not show significant associations (Table 4). Interestingly, we noticed that the mean age of patients in our study was 23.3 years, which is much lower than the mean age of 42.3 years reported in the original study on SCORTEN.7 This explains why age, which has been regarded as an important prognostic factor in most Western studies, was not found to have significant influence in the present study (Table 4). This tendency to affect younger patients has also been noted in previous Indian studies on SJS and TEN.26 Similarly, malignancy, which is an important parameter in Western populations, lacks significance in the current setting as a result of its low incidence in the study population. This also has important bearing on how ethnicity influences outcomes in TEN patients. Factors such as increasing age and malignancy, which are known to have negative effects on outcome, are less commonly seen in Asian populations, which may indicate a favorable outcome in comparison with those in Western populations. The inclusion of blood glucose as one of the parameters in SCORTEN was not found to be relevant, which echoes the findings of Hu et al.27 A re-evaluation of the existing SCORTEN parameters was suggested by Sekula et al.13 and Vaishampayan et al.14 in their respective studies. Analyses of other possible risk factors found that hypernatremia, delayed admission, and skin colonization did not show any impact on mortality. Hague et al.12 suggested respiratory involvement to be a poor prognostic factor that is not reflected in SCORTEN. In the present study, we were unable to assess this factor on multivariate analysis, but because there is a strong correlation between serum bicarbonate and respiratory involvement, it is not necessary to add respiratory involvement as a factor. The inclusion of coexistent morbidities has also been suggested, but this factor did not show a significant association when other prognostic factors were controlled for. Moreover, although it is difficult to define the actual cause of death in SJS/TEN, death within the first six weeks is considered to be related to the reaction itself.28 This claim is supported by events in the current study, in which all patient deaths occurred within the first two weeks. Thus the inclusion of coexistent morbidities in the calculation of SCORTEN scores is probably unnecessary. Of the 60 patients included in the present study, 10 succumbed to death, giving a mortality rate of 16.7%. Mortality rates of 10–66% have been reported in various previous studies,29–31 including a population-based registry in which mortality was as high as 48%.28 That we observed a low mortality rate of 16.7% in the present study while using a conservative line of treatment has important therapeutic implications for patients with SJS or TEN. International Journal of Dermatology 2015, 54, e18–e26
As well as conservative management strategies, various therapies such as corticosteroids, intravenous immunoglobulins, cyclosporine, thalidomide, and plasmapheresis have been tried in various studies, with conflicting results. However, most of these studies represented either small case series or retrospective non-randomized studies.10,32–39 A Cochrane review on interventions in TEN reported that, to date, only a single, double-blind, randomized, placebo-controlled study had used thalidomide in SJS and TEN.40 However, the study had to be stopped midway because of excessive mortality in the thalidomide group. Retrospective studies do not show any benefit of the use of steroids or other immunosuppressants in TEN, and higher rates of morbidity and mortality have been reported in patients treated with steroids.28,31,41 Many instances of SJS and TEN have involved patients who were using high-dose steroids for pre-existing diseases, which is eloquent proof that steroids do not alter the pathogenesis or course of TEN.28,41 Although SCORTEN predicted mortality reasonably well in the current study, only three parameters (heart rate, serum bicarbonate level, and blood urea) were found to be significant. However, we admit that the testing of 15 variables (some of which were evaluated at three timepoints) may have led to an unavoidable margin of error. In addition, we did not attempt to mirror the rigorous revalidation of SCORTEN in three populations performed previously by other authors.11 The key messages of the present study relate to the value of using serial analysis of SCORTEN values obtained on a daily or alternate-day basis, which is better than using analysis on day 1 of admission only. In addition, the performance of the scale is influenced by the demographic profile and characteristics of the study population, and minor modifications based on the population under study may increase the predictive accuracy of the original scale. Moreover, different standards of care should be factored into the use of SCORTEN. It is probably time to review SCORTEN. Just as change is the only constant in life, so it is in SCORTEN. The benefits and risks of steroid therapy are subject to international controversy, as are those of other immunomodulatory therapies. In view of the low mortality rate observed in the present study, the conservative line of treatment followed in our set-up appears to represent a valid therapeutic option that is preferable to the administration of high-dose steroids and immunosuppressants and is highly recommended. References 1 Bastuji-Garin S, Rzany B, Stern RS, et al. Clinical classification of cases of toxic epidermal necrolysis, ª 2013 The International Society of Dermatology
Bansal et al.
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Stevens–Johnson syndrome, and erythema multiforme. Arch Dermatol 1993; 129: 92–96. Roujeau JC, Guillaume JC, Fabre JP, et al. Toxic epidermal necrolysis (Lyell syndrome): incidence and drug etiology in France, 1981–1985. Arch Dermatol 1990; 126: 37–42. Schopf E, Stuhmer A, Rzany B, et al. Toxic epidermal necrolysis and Stevens–Johnson syndrome: an epidemiological study from West Germany. Arch Dermatol 1991; 127: 839–842. Roujeau JC, Kelly JP, Naldi L, et al. Medication use and the risk of Stevens–Johnson syndrome or toxic epidermal necrolysis. N Engl J Med 1995; 333: 1600–1608. Devi K, George S, Criton S, et al. Carbamazepine – the commonest cause of toxic epidermal necrolysis and Stevens–Johnson syndrome: a study of 7 years. Indian J Dermatol Venereol Leprol 2005; 71: 325–328. Revuz J, Penso D, Roujeau JC, et al. Toxic epidermal necrolysis: clinical findings and prognosis factors in 87 patients. Arch Dermatol 1987; 123: 1160–1165. Bastuji-Garin S, Fouchard N, Bertocchi M, et al. SCORTEN: a severity of illness score for toxic epidermal necrolysis. J Invest Dermatol 2000; 115: 149–153. Trent JT, Kirnser RS, Romanelli P, et al. Use of SCORTEN to accurately predict mortality in patients with toxic epidermal necrolysis in the United States. Arch Dermatol 2004; 140: 890–892. Cartotto R, Mayich M, Nickerson D, et al. SCORTEN accurately predicts mortality among toxic epidermal necrolysis patients treated in a burn center. J Burn Care Res 2008; 29: 141–146. Trent JT, Kirsner RS, Romanelli P, et al. Analysis of intravenous immunoglobulin for the treatment of toxic epidermal necrolysis using SCORTEN. Arch Dermatol 2003; 139: 39–43. Guegan S, Bastuji-Garin S, Poszepczynska-Guigne E, et al. 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. Hague JS, Goulding JM, Long TM, et al. Respiratory involvement in toxic epidermal necrolysis portends a poor prognosis that may not be reflected in SCORTEN. Br J Dermatol 2007; 157: 1294–1296. Sekula P, Liss Y, Davidovici B, et al. Evaluation of SCORTEN on a cohort of patients with Stevens– Johnson syndrome and toxic epidermal necrolysis included in the RegiSCAR study. J Burn Care Res 2011; 32: 237–245. Vaishampayan SS, Das AL, Verma R. SCORTEN: does it need modification? Indian J Dermatol Venereol Leprol 2008; 74: 35–37. Spornraft-Ragaller P, Theilen H, Gottschlich G, et al. Treatment of toxic epidermal necrolysis. Experience with nine patients with consideration of intravenous immunoglobulin. Hautarzt 2006; 57: 185–194. Imahara SD, Holmes JHIV, Heimbach DM, et al. SCORTEN overestimates mortality in the setting of a
ª 2013 The International Society of Dermatology
Therapeutic and SCORTEN analysis in SJS/TEN
17
18
19
20
21
22
23
24
25
26
27
28
29 30
Report
standardized treatment protocol. J Burn Care Res 2006; 27: 270–275. Palmieri TL, Greenhalgh DG, Saffle JR, et al. A multicenter review of toxic epidermal necrolysis in US burn centers at the end of twentieth century. J Burn Care Rehabil 2002; 23: 87–96. de Prost N, Ingen-Housz-Oro S, Duong T, et al. Bacteremia in Stevens–Johnson syndrome and toxic epidermal necrolysis: epidemiology, risk factors, and predictive value of skin cultures. Medicine (Baltimore) 2010; 89: 28–36. Namdar T, von Wild T, Siemers F, et al. Does hypernatremia impact mortality in toxic epidermal necrolysis? Ger Med Sci 2010; 8: 30. George SMC, Harrison DA, Welch CA, et al. Dermatological conditions in intensive care: a secondary analysis of the Intensive Care National Audit and Research Centre (ICNARC) case mix program database. Crit Care 2008; 12: S1. Morales ME, Purdue GF, Verity SM, et al. Ophthalmic manifestations of Stevens–Johnson syndrome and toxic epidermal necrolysis and relation to SCORTEN. Am J Ophthalmol 2010; 150: 505–510. Ho YL, Chang YT, Chu YT, et al. Performance of the SCORTEN in Taiwanese patients with Stevens–Johnson syndrome and toxic epidermal necrolysis. Dermatologica Sinica 2010; 28: 15–20. Kannenberg SM, Jordaan HF, Koegelenberg CF, et al. Toxic epidermal necrolysis and Stevens–Johnson syndrome in South Africa: a 3-year prospective study. QJM 2012; 105: 839–846. Zhu QY, Ma L, Luo XQ, et al. Toxic epidermal necrolysis: performance of SCORTEN and the score-based comparison of the efficacy of corticosteroid therapy and intravenous immunoglobulin combined therapy in China. J Burn Care Res 2012; 33: 295–308. Avakian R, Flowers FP, Araujo OE, et al. Toxic epidermal necrolysis: a review. J Am Acad Dermatol 1991; 25: 69–79. Kaur S, Nanda A, Sharma KV. Elucidation and management of 30 patients of drug-induced toxic epidermal necrolysis (DTEN). Indian J Dermatol Venereol Leprol 1990; 56: 196–199. Hu CH, Chang NJ, Liu EK, et al. SCORTEN and impaired renal function related to mortality of toxic epidermal necrolysis syndrome patients in the Asian population. J Eur Acad Dermatol Venereol 2013; 27: 628–633. Mockenhaupt M. Severe cutaneous adverse reactions. In: Burgdorf WHC, Plewig G, Wolff HH, Landthaler M, eds. Braun-Falco’s Dermatology, 3rd edn. Heidelberg: Springer Medizin Verlag, 2008: 473–484. Ting HC, Adam BA. Stevens–Johnson syndrome: a review of 34 cases. Int J Dermatol 1985; 24: 587–591. Guillaume J, Roujeau JC, Revuz J, et al. The culprit drugs in 87 cases of toxic epidermal necrolysis (Lyells syndrome). Arch Dermatol 1987; 123: 1166–1170.
International Journal of Dermatology 2015, 54, e18–e26
e25
e26
Report
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
31 Halebian PH, Corder VJ, Madden MR, et al. Improved burn center survival of patients with toxic epidermal necrolysis managed without corticosteroids. Ann Surg 1986; 204: 503–512. 32 Bachot N, Revuz J, Roujeau JC. Intravenous immunoglobulin treatment for Stevens–Johnson syndrome and toxic epidermal necrolysis: a prospective non-comparative study showing no benefit on mortality or progression. Arch Dermatol 2003; 139: 33–36. 33 Prins C, Vittorio C, Padilla RS, et al. Effect of high-dose intravenous immunoglobulin therapy in Stevens–Johnson syndrome: a retrospective, multicenter study. Dermatology 2003; 207: 96–99. 34 Kardaun SH, Jonkman MF. Dexamethasone pulse therapy for Stevens–Johnson syndrome/toxic epidermal necrolysis. Acta Derm Venereol 2007; 87: 144–148. 35 Schneck J, Fagot JP, Sekula P, et al. 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. 36 Huang YC, Li YC, Chen TJ. The efficacy of intravenous immunoglobulin for the treatment of toxic epidermal
International Journal of Dermatology 2015, 54, e18–e26
37
38
39
40
41
necrolysis: a systematic review and meta-analysis. Br J Dermatol 2012; 167: 424–432. Lee HY, Dunant A, Sekula P, et al. The role of prior corticosteroid use on the clinical course of Stevens– Johnson syndrome and toxic epidermal necrolysis: a case–control analysis of patients selected from the multinational EuroSCAR and RegiSCAR studies. Br J Dermatol 2012; 167: 555–562. Tan SK, Tay YK. Profile and pattern of Stevens–Johnson syndrome and toxic epidermal necrolysis in a general hospital in Singapore: treatment outcomes. Acta Derm Venereol 2012; 92: 62–66. Yang Y, Xu J, Li F, et al. Combination therapy of intravenous immunoglobulin and corticosteroid in the treatment of toxic epidermal necrolysis and Stevens–Johnson syndrome: a retrospective comparative study in China. Int J Dermatol 2009; 48: 1122–1128. Wolkenstein P, Latarjet J, Roujeau JC, et al. Randomized comparison of thalidomide versus placebo in toxic epidermal necrolysis. Lancet 1998; 352: 1586–1589. Wolkenstein P, Revuz J. Toxic epidermal necrolysis. Dermatol Clin 2000; 18: 485–495.
ª 2013 The International Society of Dermatology
A clinicotherapeutic analysis of Stevens–Johnson syndrome and toxic epidermal necrolysis with an emphasis on the predictive value and accuracy of SCORe of Toxic Epidermal Necrolysis Shuchi Bansal, MD, Vijay K. Garg, MD, MNAMS, Kabir Sardana, MD, DNB, MNAMS, and Rashmi Sarkar, MD, DNB, MNAMS
Department of Dermatology, Maulana Azad Medical College and Lok Nayak Hospital, New Delhi, India Correspondence Kabir Sardana, MD, DNB, MNAMS 466, Sector 28 Noida Uttar Pradesh 201303 India E-mail: [email protected] Funding: None.
Abstract Background The SCORTEN scale (SCORe of Toxic Epidermal Necrolysis) is widely used to predict mortality in patients with Stevens–Johnson syndrome (SJS) or toxic epidermal necrolysis (TEN). Although it is largely a useful tool, the predictive ability of the scale is variable, and modifications to the existing scale have been suggested. In addition, there is controversy regarding the roles of active therapy and conservative management. Objectives This study was conducted to assess outcomes in SJS/TEN patients managed conservatively and to analyze the predictive performance of SCORTEN. Methods Sixty patients were studied prospectively from October 2008 to September 2011. The accuracy of SCORTEN in predicting mortality was analyzed on days 1, 3, and 5 of admission. All patients were managed conservatively. Results The discriminative power of SCORTEN was acceptable, being best on day 5. The
Conflicts of interest: None.
agreement between actual and predicted mortality was poor, as indicated by a low P-value of the Hosmer–Lemeshow statistic. Only three parameters (heart rate, blood urea, and serum bicarbonate) were found to be significant on multivariate analysis, and all of these represented components within the original SCORTEN scale. The mortality rate was 16.7%. All deaths occurred within the first two weeks (8–12 days). Conclusions Serial analysis using SCORTEN is better than analysis performed only on day 1. Although the scale is largely useful, its performance is influenced by the demographic profile of the study population; minor modifications based on the population studied may increase the predictive accuracy of the original SCORTEN. In addition, conservative management is a valid therapeutic option, is preferable to treatment with steroids and immunosuppressants, and is highly recommended.
Introduction
e18
Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are life-threatening drug-induced reactions characterized by the occurrence of distinctive skin lesions, severe mucosal involvement, and systemic symptoms; severity is categorized based on the percentage of epidermal detachment.1 Drugs have been implicated as the most important etiological factors.2–5 Mortality ranges from 5% in SJS to 10–15% in the SJS–TEN overlap and 30–40% in TEN.1–5 Substantial endeavor has been made to determine the predictors, both clinical and laboratory-based, of prognosis in patients with SJS and TEN.6 The SCORTEN scale (SCORe of Toxic Epidermal Necrolysis), which provides International Journal of Dermatology 2015, 54, e18–e26
a specific score indicating the severity of illness, was developed and validated by Bastuji-Garin et al.,7 in 2000 to determine disease severity and prognosis in patients with SJS or TEN. The scale is based on seven parameters, each of which is given a score of 1 and calculated on day 1 of admission. A score of 1–7 predicts a probability of mortality ranging from 3.9 to 98.5% as calculated using the equation to predict death proposed by the authors.7 Since its development, SCORTEN has been widely used to predict mortality in patients with SJS or TEN, as well as to help determine the efficacy of therapeutic interventions.8–10 Although the scale performs well, there are issues regarding its predictive power as well as its accuracy.11–13 In a study of 144 patients conducted by Guegan et al.,11 SCORTEN was found to perform best on day 3, rather ª 2013 The International Society of Dermatology
Bansal et al.
than on day 1 as originally proposed by the authors. Vaishampayan et al.14 and Sekula et al.13 have called for a reevaluation of the existing SCORTEN parameters and have also suggested modifications to the original scale. Spornraft-Ragaller et al.15 observed a considerable overestimation and concluded that SCORTEN did not perform well in severely affected patients. Imahara et al.16 found that the predictive performance of SCORTEN is influenced by the treatment protocol used, which has changed significantly since the introduction of the scale. Notwithstanding the pioneering work of its original authors7 and its validity,8,9 there is a need to fine-tune the use of the scale. As no studies have as yet attempted to confirm the accuracy of SCORTEN in a large cohort of Indian patients, we aimed to analyze the predictive performance of SCORTEN in SJS and TEN patients using analysis on days 1, 3, and 5 of admission and to analyze other prognostic factors that have been suggested by other authors but not included in SCORTEN. The clinical course and final outcomes in patients were assessed using a conservative line of management with the aim of comparing our therapeutic outcome with those reported by other studies to date. Materials and methods Study design and settings This prospective observational study was conducted in the Department of Dermatology, Venereology, and Leprology in a tertiary care hospital from October 2008 to September 2011. Patients Sixty-four consecutive patients with SJS or TEN were admitted to this department from October 2008 to September 2011. The diagnosis of disease was based upon the widely accepted classification system suggested by Bastuji-Garin et al.1 Because some of these patients had been referred from primary centers, only those patients who had not been treated with any modality other than conservative therapy were included in the study. Four patients did not fulfill this criterion and were thus excluded. Study protocol Patients were subjected to a detailed assessment of their clinical history and a thorough general physical as well as mucocutaneous examination. In addition to routine hematological and biochemical investigations, all the necessary investigations required for analysis of SCORTEN were carried out. A plain skiagram of the chest along with culture samples from skin, blood, and urine were obtained in all patients. Patients were managed using a conservative protocol that included the withdrawal of the suspected causative drug, fluid replacement, nutritional support, temperature regulation, and the provision of daily dressings and prophylactic antibiotics. ª 2013 The International Society of Dermatology
Therapeutic and SCORTEN analysis in SJS/TEN
Report
SCORTEN scores were calculated for all patients on days 1, 3, and 5 of admission. Patients were hospitalized until complete resolution of skin and mucosal lesions and were subsequently followed for a further three months after discharge.
Endpoint The primary endpoint for evaluation was mortality rate at 1 month from the onset of SJS or TEN. This was compared with SCORTEN-predicted mortality as calculated using the predictive death equation: PðdeathÞ ¼ e logit 1 þ e logit where logit = 4.448 + 1.237 (SCORTEN).
Statistical analysis The discriminative power of SCORTEN was assessed by the area under the receiver operator characteristic (ROC) curve. The area under the curve (AUC) ranges from 0 to 1; a value of 1 indicates perfect discrimination by the test, whereas a value of 30% Delay in hospitalization (days)
suggested
to
augment
Vaishampayan et al.14 Palmieri et al.17 Revuz et al.6 de Prost et al.18 Revuz et al.6 Revuz et al.6 Hague et al.12 Namdar et al.19 Sekula et al.13 Palmieri et al.17
International Journal of Dermatology 2015, 54, e18–e26
e19
e20
Report
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
three years. Twenty patients had been referred from various primary centers, and the remaining 40 had been admitted directly to our center. Final diagnoses were SJS in 30 patients, overlapping SJS and TEN in 10 patients, and TEN in 20 patients. The mean standard deviation age of the patients was 23.3 4.3 years (range: 7–54 years). A positive history of drug intake was present in 96.6% of patients. Anticonvulsants represented the most common causative drug (43.3% of patients), followed by analgesics (26.6% of patients). Most of the patients had deranged hematological and biochemical results. Bacterial skin colonization was present in 26 (43.3%) patients, nine of whom also had systemic infection. Four patients showed pulmonary infiltrates on chest x-ray. Coexistent morbidities were present in 11 (18.3%) patients at the time of disease onset in the form of tuberculosis, cardiovascular disease, leprosy, diabetes, and hepatitis. Malignancy workup was negative in all patients.
respectively, showed that although the SCORTEN values for deceased patients were, as expected, significantly higher than those for surviving patients at all three timepoints, the difference was most significant at day 5 (P < 0.001; Table 3). A comparison of observed and predicted mortalities for each score value showed poor agreement, as indicated by the significantly low P-values (30% also had a better correlation with mortality than did epidermal detachment of >10% (P = 0.070 and P = 0.299, respectively). Deranged electrolytes, skin colonization, and delayed hospitalization did not show any significant association with mortality. Multivariate analysis
All of the parameters found to be significant on univariate analysis (both the original SCORTEN parameters, as well as the additional parameters) were included in multiple regression analysis. For the SCORTEN parameters, only day 5 values were entered in the logistic regression because the discriminative power of SCORTEN was found to be best at that time-point. Respiratory involvement, although found to be significantly important on univariate analysis, could not be included in the logistic regression because none of the surviving subjects had respiratory involvement. After backward elimination, only three parameters remained significant on multivariate analysis: blood urea, serum bicarbonate, and heart rate (Table 5). Discussion Stevens–Johnson syndrome and TEN are life-threatening conditions that contribute to a large percentage of the ª 2013 The International Society of Dermatology
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
Report
Table 3 SCORTEN values and corresponding mortality (days 1, 3 and 5) Day 1
Day 3 Deaths, actual/ predicted, n
SCORTEN score
Patients, n
0 1 2 3 4 5 6 Total Mean SCORTEN (P-value) SMR (95% CI) AUC H–L test
0 – 26 1/1.04 20 2/2.40 8 6/2.60 3 0/1.86 3 1/2.60 0 – 60 10/10.50 1.78a/2.80b (0.002) 0.952 (0.362–1.54) 0.787 0.001
Patients, n
Day 5 Deaths, actual/ predicted, n
0 – 28 1/1.12 19 2/2.28 7 5/2.24 3 1/1.86 3 1/2.60 0 – 60 10/10.10 1.70a/2.90b (0.001) 0.990 (0.376–1.60) 0.811 0.013
Patients, n
Deaths, actual/ predicted, n
0 – 30 1/1.20 17 1/2.04 6 4/1.92 4 2/2.48 3 2/2.55 0 – 60 10/10.19 1.60a/3.30b ( 75%), it was best at day 5, which is contrary to the findings of the original study.7 A similar observation in a sample of Indian patients was made by Vaishampayan et al.,14 who found the performance of SCORTEN to be best on day 5, although the small sample size of their study precluded any statistical analysis. Guegan et al.11 found the accuracy of SCORTEN for predicting mortality to be best on day 3 of admission, whereas a study in Taiwanese patients, conducted by Ho et al.,22 found it to be best on day 1 of admission. We hypothesize that this variation is associated with the fact that the acute phase of SJS/TEN lasts 8–12 days, during which the disease progresses,25 and that the SCORTEN scale will perform best when the disease is at its peak. The mean delay in admission in our patients was 2.5 days, whereas mean delays of 5.2 and 6.75 days, respectively, were reported by Guegan et al.11 and Ho et al.22 Moreover, the majority (n = 40/60) of patients in the present study presented to us directly and were not referred from other centers. Hence, the disease reached its peak in our patients a few days further into hospital admission than in the two International Journal of Dermatology 2015, 54, e18–e26
e21
Report
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
90 SCORTEN-predicted mortality
80
Mortality on day 1 70 Mortality on day 3 Mortality (%)
e22
60 Mortality on day 5 50 40 30 20 10 0 1
2
3 SCORTEN score
4
5
Figure 2 Line diagram comparing mortality rates on days 1, 3 and 5 with SCORTEN-predicted mortality and showing that mortality is underestimated for scores of 2 and 3, and overestimated for scores of 4 and 5
Table 4 Univariate analysis of factors prognostic for mortality
Prognostic factor
Surviving patients(n = 50) (%)
Deceased patients(n = 10) (%)
P-value
Age >40 yearsa Presence of malignancya Heart rate >120 bpma Epidermal detachment >10%a Blood urea >28 mg/dla Serum glucose >252 mg/dla Serum bicarbonate 30%b Delay in hospitalization (days)b
15/50 0/50 12/50 23/50 13/50 4/50 13/50 6/50 20/50 4/50 5/50 0/50 8/50 14/50 Mean
2/10 (20) 0/10 (0) 7/10 (70) 7/10 (70) 8/10 (80) 1/10 (10) 8/10 (80) 5/10 (50) 6/10 (60) 5/10 (50) 1/10 (10) 4/10 (40) 2/10 (20) 6/10 (60) Mean 2.40
0.709 _ 0.008 0.299 0.002 1.0 0.002 0.013 0.305 0.004 1.0 120 bpm
44.133 25.445 6.823
0.000 0.001 0.050
3.787 3.237 1.920
a
After applying backward elimination.
International Journal of Dermatology 2015, 54, e18–e26
earlier studies,11,22 in which the disease was already at its peak at admission. Thus, the duration of disease from the time of onset, rather than the duration of hospital stay, is crucial in influencing the predictive performance of SCORTEN. Consequently, a serial analysis of SCORTEN would serve as a better indicator of mortality, especially in patients who present in the early stages of the reaction. Analyses of individual SCORTEN parameters showed that only three parameters (heart rate, blood urea, and serum bicarbonate) had a significant association with ª 2013 The International Society of Dermatology
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
Report
Table 6 Seminal studies on the efficacy of SCORTEN in patients with Stevens–Johnson syndrome (SJS) or toxic epidermal necrolysis (TEN) Results Authors (year)
Patients, n
SMR (95% CI)
AUC
H–L statistic
Conclusions
Trent et al.8 (2004)
24 TEN patients
0.912 (0.393–1.800)
–
–
Guegan et al.11 (2006)
144 Patients
–
>80%
>0.30 Highest on day 3
Vaishampayan et al.14 (2008)
10 SJS–TEN/TEN
Statistical analysis not done
Cartotto et al.9 (2008)
61 Patients 7% SJS, 39% overlap, 54% TEN
1.17 (0.695–1.853)
0.82
1.381
George et al.20 (2008)
145 Patients (including erythema multiforme)
–
0.762
–
Morales et al.21 (2010) Ho et al.22 (2010)
82 Patients
Correlation coefficient 0.97, P = 0.005
51 Patients 37 SJS, 8 overlap, 6 TEN
1.12 (0.48–2.2)
>80% on days 1–5
Highest on day 1 (0.88)
Sekula et al.13 (2011)
166 Patients 72 SJS, 59 overlap, 35 TEN
–
0.75
–
Hu et al.27 (2013)
101 TEN patients
1.625 (1.229–2.109)
–
–
Present study (2013)
60 Patients 30 SJS, 10 overlap, 20 TEN
Day 1: 0.952 (0.362–1.54) Day 3: 0.990 (0.376–1.60) Day 5: 0.990 (0.376–1.60)
Day 1: 0.787 Day 3: 0.811 Day 5: 0.861
Day 1: 0.001 Day 3: 0.013 Day 5: 0.095
No significant difference between actual and SCORTEN-predicted mortality SCORTEN performance during the first 5 days of hospitalization was excellent and best on day 3 Systemic diseases should be included in SCORTEN BSA >30% should be given greater weight SCORTEN is an accurate scoring system for estimation of mortality among SJS–TEN patients Discriminative power of SCORTEN was better than that of the APACHE II score (AUC = 0.737), but not as good as that of ICNARC (AUC = 0.795) SCORTEN value correlates well with mortality rate Discrimination and calibration of SCORTEN were good and best on day 1 SCORTEN remains the tool of choice, but the auxillary score might be an alternative in retrospective settings with missing laboratory data Mortality rate was directly proportional to the SCORTEN score Renal insufficiency and failure significantly affected outcome Discriminative power of SCORTEN is good and best on day 5, but the calibration between observed and predicted mortalities is poor Performance of SCORTEN is influenced by the demographic profile of the study population
AUC, area under the receiver operator characteristic curve; SMR, standardized mortality ratio; 95% CI, 95% confidence interval; H–L test, P-value of the Hosmer–Lemeshow goodness-of-fit test; BSA, body surface area; APACHE, Acute Physiology and Chronic Health Evaluation; ICNARC, Intensive Care National Audit and Research Center.
ª 2013 The International Society of Dermatology
International Journal of Dermatology 2015, 54, e18–e26
e23
e24
Report
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
mortality, whereas age, blood glucose, and epidermal detachment of >10% did not show significant associations (Table 4). Interestingly, we noticed that the mean age of patients in our study was 23.3 years, which is much lower than the mean age of 42.3 years reported in the original study on SCORTEN.7 This explains why age, which has been regarded as an important prognostic factor in most Western studies, was not found to have significant influence in the present study (Table 4). This tendency to affect younger patients has also been noted in previous Indian studies on SJS and TEN.26 Similarly, malignancy, which is an important parameter in Western populations, lacks significance in the current setting as a result of its low incidence in the study population. This also has important bearing on how ethnicity influences outcomes in TEN patients. Factors such as increasing age and malignancy, which are known to have negative effects on outcome, are less commonly seen in Asian populations, which may indicate a favorable outcome in comparison with those in Western populations. The inclusion of blood glucose as one of the parameters in SCORTEN was not found to be relevant, which echoes the findings of Hu et al.27 A re-evaluation of the existing SCORTEN parameters was suggested by Sekula et al.13 and Vaishampayan et al.14 in their respective studies. Analyses of other possible risk factors found that hypernatremia, delayed admission, and skin colonization did not show any impact on mortality. Hague et al.12 suggested respiratory involvement to be a poor prognostic factor that is not reflected in SCORTEN. In the present study, we were unable to assess this factor on multivariate analysis, but because there is a strong correlation between serum bicarbonate and respiratory involvement, it is not necessary to add respiratory involvement as a factor. The inclusion of coexistent morbidities has also been suggested, but this factor did not show a significant association when other prognostic factors were controlled for. Moreover, although it is difficult to define the actual cause of death in SJS/TEN, death within the first six weeks is considered to be related to the reaction itself.28 This claim is supported by events in the current study, in which all patient deaths occurred within the first two weeks. Thus the inclusion of coexistent morbidities in the calculation of SCORTEN scores is probably unnecessary. Of the 60 patients included in the present study, 10 succumbed to death, giving a mortality rate of 16.7%. Mortality rates of 10–66% have been reported in various previous studies,29–31 including a population-based registry in which mortality was as high as 48%.28 That we observed a low mortality rate of 16.7% in the present study while using a conservative line of treatment has important therapeutic implications for patients with SJS or TEN. International Journal of Dermatology 2015, 54, e18–e26
As well as conservative management strategies, various therapies such as corticosteroids, intravenous immunoglobulins, cyclosporine, thalidomide, and plasmapheresis have been tried in various studies, with conflicting results. However, most of these studies represented either small case series or retrospective non-randomized studies.10,32–39 A Cochrane review on interventions in TEN reported that, to date, only a single, double-blind, randomized, placebo-controlled study had used thalidomide in SJS and TEN.40 However, the study had to be stopped midway because of excessive mortality in the thalidomide group. Retrospective studies do not show any benefit of the use of steroids or other immunosuppressants in TEN, and higher rates of morbidity and mortality have been reported in patients treated with steroids.28,31,41 Many instances of SJS and TEN have involved patients who were using high-dose steroids for pre-existing diseases, which is eloquent proof that steroids do not alter the pathogenesis or course of TEN.28,41 Although SCORTEN predicted mortality reasonably well in the current study, only three parameters (heart rate, serum bicarbonate level, and blood urea) were found to be significant. However, we admit that the testing of 15 variables (some of which were evaluated at three timepoints) may have led to an unavoidable margin of error. In addition, we did not attempt to mirror the rigorous revalidation of SCORTEN in three populations performed previously by other authors.11 The key messages of the present study relate to the value of using serial analysis of SCORTEN values obtained on a daily or alternate-day basis, which is better than using analysis on day 1 of admission only. In addition, the performance of the scale is influenced by the demographic profile and characteristics of the study population, and minor modifications based on the population under study may increase the predictive accuracy of the original scale. Moreover, different standards of care should be factored into the use of SCORTEN. It is probably time to review SCORTEN. Just as change is the only constant in life, so it is in SCORTEN. The benefits and risks of steroid therapy are subject to international controversy, as are those of other immunomodulatory therapies. In view of the low mortality rate observed in the present study, the conservative line of treatment followed in our set-up appears to represent a valid therapeutic option that is preferable to the administration of high-dose steroids and immunosuppressants and is highly recommended. References 1 Bastuji-Garin S, Rzany B, Stern RS, et al. Clinical classification of cases of toxic epidermal necrolysis, ª 2013 The International Society of Dermatology
Bansal et al.
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Stevens–Johnson syndrome, and erythema multiforme. Arch Dermatol 1993; 129: 92–96. Roujeau JC, Guillaume JC, Fabre JP, et al. Toxic epidermal necrolysis (Lyell syndrome): incidence and drug etiology in France, 1981–1985. Arch Dermatol 1990; 126: 37–42. Schopf E, Stuhmer A, Rzany B, et al. Toxic epidermal necrolysis and Stevens–Johnson syndrome: an epidemiological study from West Germany. Arch Dermatol 1991; 127: 839–842. Roujeau JC, Kelly JP, Naldi L, et al. Medication use and the risk of Stevens–Johnson syndrome or toxic epidermal necrolysis. N Engl J Med 1995; 333: 1600–1608. Devi K, George S, Criton S, et al. Carbamazepine – the commonest cause of toxic epidermal necrolysis and Stevens–Johnson syndrome: a study of 7 years. Indian J Dermatol Venereol Leprol 2005; 71: 325–328. Revuz J, Penso D, Roujeau JC, et al. Toxic epidermal necrolysis: clinical findings and prognosis factors in 87 patients. Arch Dermatol 1987; 123: 1160–1165. Bastuji-Garin S, Fouchard N, Bertocchi M, et al. SCORTEN: a severity of illness score for toxic epidermal necrolysis. J Invest Dermatol 2000; 115: 149–153. Trent JT, Kirnser RS, Romanelli P, et al. Use of SCORTEN to accurately predict mortality in patients with toxic epidermal necrolysis in the United States. Arch Dermatol 2004; 140: 890–892. Cartotto R, Mayich M, Nickerson D, et al. SCORTEN accurately predicts mortality among toxic epidermal necrolysis patients treated in a burn center. J Burn Care Res 2008; 29: 141–146. Trent JT, Kirsner RS, Romanelli P, et al. Analysis of intravenous immunoglobulin for the treatment of toxic epidermal necrolysis using SCORTEN. Arch Dermatol 2003; 139: 39–43. Guegan S, Bastuji-Garin S, Poszepczynska-Guigne E, et al. 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. Hague JS, Goulding JM, Long TM, et al. Respiratory involvement in toxic epidermal necrolysis portends a poor prognosis that may not be reflected in SCORTEN. Br J Dermatol 2007; 157: 1294–1296. Sekula P, Liss Y, Davidovici B, et al. Evaluation of SCORTEN on a cohort of patients with Stevens– Johnson syndrome and toxic epidermal necrolysis included in the RegiSCAR study. J Burn Care Res 2011; 32: 237–245. Vaishampayan SS, Das AL, Verma R. SCORTEN: does it need modification? Indian J Dermatol Venereol Leprol 2008; 74: 35–37. Spornraft-Ragaller P, Theilen H, Gottschlich G, et al. Treatment of toxic epidermal necrolysis. Experience with nine patients with consideration of intravenous immunoglobulin. Hautarzt 2006; 57: 185–194. Imahara SD, Holmes JHIV, Heimbach DM, et al. SCORTEN overestimates mortality in the setting of a
ª 2013 The International Society of Dermatology
Therapeutic and SCORTEN analysis in SJS/TEN
17
18
19
20
21
22
23
24
25
26
27
28
29 30
Report
standardized treatment protocol. J Burn Care Res 2006; 27: 270–275. Palmieri TL, Greenhalgh DG, Saffle JR, et al. A multicenter review of toxic epidermal necrolysis in US burn centers at the end of twentieth century. J Burn Care Rehabil 2002; 23: 87–96. de Prost N, Ingen-Housz-Oro S, Duong T, et al. Bacteremia in Stevens–Johnson syndrome and toxic epidermal necrolysis: epidemiology, risk factors, and predictive value of skin cultures. Medicine (Baltimore) 2010; 89: 28–36. Namdar T, von Wild T, Siemers F, et al. Does hypernatremia impact mortality in toxic epidermal necrolysis? Ger Med Sci 2010; 8: 30. George SMC, Harrison DA, Welch CA, et al. Dermatological conditions in intensive care: a secondary analysis of the Intensive Care National Audit and Research Centre (ICNARC) case mix program database. Crit Care 2008; 12: S1. Morales ME, Purdue GF, Verity SM, et al. Ophthalmic manifestations of Stevens–Johnson syndrome and toxic epidermal necrolysis and relation to SCORTEN. Am J Ophthalmol 2010; 150: 505–510. Ho YL, Chang YT, Chu YT, et al. Performance of the SCORTEN in Taiwanese patients with Stevens–Johnson syndrome and toxic epidermal necrolysis. Dermatologica Sinica 2010; 28: 15–20. Kannenberg SM, Jordaan HF, Koegelenberg CF, et al. Toxic epidermal necrolysis and Stevens–Johnson syndrome in South Africa: a 3-year prospective study. QJM 2012; 105: 839–846. Zhu QY, Ma L, Luo XQ, et al. Toxic epidermal necrolysis: performance of SCORTEN and the score-based comparison of the efficacy of corticosteroid therapy and intravenous immunoglobulin combined therapy in China. J Burn Care Res 2012; 33: 295–308. Avakian R, Flowers FP, Araujo OE, et al. Toxic epidermal necrolysis: a review. J Am Acad Dermatol 1991; 25: 69–79. Kaur S, Nanda A, Sharma KV. Elucidation and management of 30 patients of drug-induced toxic epidermal necrolysis (DTEN). Indian J Dermatol Venereol Leprol 1990; 56: 196–199. Hu CH, Chang NJ, Liu EK, et al. SCORTEN and impaired renal function related to mortality of toxic epidermal necrolysis syndrome patients in the Asian population. J Eur Acad Dermatol Venereol 2013; 27: 628–633. Mockenhaupt M. Severe cutaneous adverse reactions. In: Burgdorf WHC, Plewig G, Wolff HH, Landthaler M, eds. Braun-Falco’s Dermatology, 3rd edn. Heidelberg: Springer Medizin Verlag, 2008: 473–484. Ting HC, Adam BA. Stevens–Johnson syndrome: a review of 34 cases. Int J Dermatol 1985; 24: 587–591. Guillaume J, Roujeau JC, Revuz J, et al. The culprit drugs in 87 cases of toxic epidermal necrolysis (Lyells syndrome). Arch Dermatol 1987; 123: 1166–1170.
International Journal of Dermatology 2015, 54, e18–e26
e25
e26
Report
Bansal et al.
Therapeutic and SCORTEN analysis in SJS/TEN
31 Halebian PH, Corder VJ, Madden MR, et al. Improved burn center survival of patients with toxic epidermal necrolysis managed without corticosteroids. Ann Surg 1986; 204: 503–512. 32 Bachot N, Revuz J, Roujeau JC. Intravenous immunoglobulin treatment for Stevens–Johnson syndrome and toxic epidermal necrolysis: a prospective non-comparative study showing no benefit on mortality or progression. Arch Dermatol 2003; 139: 33–36. 33 Prins C, Vittorio C, Padilla RS, et al. Effect of high-dose intravenous immunoglobulin therapy in Stevens–Johnson syndrome: a retrospective, multicenter study. Dermatology 2003; 207: 96–99. 34 Kardaun SH, Jonkman MF. Dexamethasone pulse therapy for Stevens–Johnson syndrome/toxic epidermal necrolysis. Acta Derm Venereol 2007; 87: 144–148. 35 Schneck J, Fagot JP, Sekula P, et al. 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. 36 Huang YC, Li YC, Chen TJ. The efficacy of intravenous immunoglobulin for the treatment of toxic epidermal
International Journal of Dermatology 2015, 54, e18–e26
37
38
39
40
41
necrolysis: a systematic review and meta-analysis. Br J Dermatol 2012; 167: 424–432. Lee HY, Dunant A, Sekula P, et al. The role of prior corticosteroid use on the clinical course of Stevens– Johnson syndrome and toxic epidermal necrolysis: a case–control analysis of patients selected from the multinational EuroSCAR and RegiSCAR studies. Br J Dermatol 2012; 167: 555–562. Tan SK, Tay YK. Profile and pattern of Stevens–Johnson syndrome and toxic epidermal necrolysis in a general hospital in Singapore: treatment outcomes. Acta Derm Venereol 2012; 92: 62–66. Yang Y, Xu J, Li F, et al. Combination therapy of intravenous immunoglobulin and corticosteroid in the treatment of toxic epidermal necrolysis and Stevens–Johnson syndrome: a retrospective comparative study in China. Int J Dermatol 2009; 48: 1122–1128. Wolkenstein P, Latarjet J, Roujeau JC, et al. Randomized comparison of thalidomide versus placebo in toxic epidermal necrolysis. Lancet 1998; 352: 1586–1589. Wolkenstein P, Revuz J. Toxic epidermal necrolysis. Dermatol Clin 2000; 18: 485–495.
ª 2013 The International Society of Dermatology
