Purpose
To evaluate the severity of ocular involvement of patients with Stevens-Johnson syndrome (SJS), toxic epidermal necrolysis (TEN), and SJS/TEN overlap, and to investigate the relationship of the SCORTEN (a severity-of-illness score for SJS and TEN based on a minimal set of well-defined variables calculated within 24 hours of admission) with eye disease in this patient population.
Design
Retrospective observational case series.
Methods
Charts of all patients admitted to the Parkland Memorial Hospital Burn Center with a preliminary diagnosis of SJS, SJS/TEN overlap, or TEN between 1998 and 2008 were reviewed. Patients were included for study if they met clinical criteria, had positive diagnostic skin biopsy, and had dermatologic and ophthalmologic consultations. Eighty-two patients with a diagnosis of SJS, SJS/TEN overlap, or TEN met inclusion criteria. Ocular manifestations were classified as mild, moderate, or severe. Admission data were used to calculate the SCORTEN. Main outcome measure was the severity of ocular involvement with respect to diagnosis and SCORTEN.
Results
Overall, 84% of patients had ocular involvement (71% SJS, 90% TEN, 100% SJS/TEN overlap). There was no difference in the severity of acute ocular complications among groups. While the SCORTEN value did correlate well with mortality rate (correlation coefficient 0.97, P = .005), there was no correlation between the SCORTEN value and severity of eye involvement in the acute setting. There was also no association of any individual diagnosis of SJS/overlap/TEN with the severity of eye involvement, although eye findings are more common in TEN ( P = .03).
Conclusions
Ocular damage in the acute setting was more frequent in patients with epidermal detachment >10% of the total body surface area. The SCORTEN value did not correlate with the severity of eye involvement in the acute setting.
Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are severe, rare skin reactions, usually attributable to drugs. These conditions can be life-threatening and involve erosions of mucous membranes as well as widespread destruction and detachment of the epidermis. SJS and TEN are closely related illnesses and can be considered a spectrum of increasing severity and mortality. SJS is characterized by widespread macules or flat atypical target lesions with epidermal detachment involving <10% body surface area (BSA) as well as mucous membrane erosions (ie, oral, genital). In TEN, epidermal detachment involves >30% BSA and resembles superficial burns because of the confluence of blisters and erosions on a greater body surface area. Overlapping cases are defined by an intermediate area of the skin lesions and are termed SJS/TEN overlap. The incidence of SJS and TEN is approximately 2 cases per million persons per year and the mortality rates for SJS and TEN are high: 1% to 5% and 25% to 35%, respectively.
Ocular complications are quite common in SJS and TEN and usually involve the cornea, conjunctiva, and eyelids. These eye problems may occur acutely in conjunction with skin involvement or after the skin eruption. Ocular complications may be severe and result in permanent visual loss because of corneal scarring or vascularization.
Previous studies have not found a difference in the severity of acute ocular complications between the SJS patient group and the TEN patient groups, nor have they found a significant difference in the frequency of long-term ocular complications between patients with SJS, overlap syndrome, and TEN. This is in contrast to the finding that percentage of body surface area with detachment of the skin is a major prognostic factor in terms of general complications and death. In the acute phase, mucous membrane involvement other than the eye (ie, mouth, genitalia) can be as high as 81%. For this reason it seems that ocular involvement could also be just as frequent. Few laboratory tests have been correlated with eye findings and those that have been examined have not been found to be significant risk factors of late ocular disease. It is not clearly known if systemic findings in these diseases have a strong correlation to ocular manifestations.
The SCORTEN is a severity-of-illness score for SJS and TEN based on a minimal set of well-defined variables that is calculated within 24 hours of admission. It is used to predict mortality in these patients. Few studies have evaluated these prognostic factors of systemic disease and their relationship with the development of acute ocular sequelae and there has not been a direct comparison of the SCORTEN with acute eye disease. In this study we aim to evaluate the relationship between the SCORTEN score and ocular/mucocutaneous involvement to identify possible clinical predictors for the development of acute ocular complications. To the best of our knowledge, this is the first description of the association between acute eye findings and the SCORTEN.
Methods
We reviewed the charts of all patients admitted to the Burn Center at Parkland Memorial Hospital (PMH) between January 1, 1998, and December 31, 2008, who were given a tentative diagnosis of SJS, overlap syndrome (SJS/TEN), or TEN on admission. Of 140 patients meeting these criteria, 96 were definitively diagnosed by the dermatology service as having SJS, TEN, or overlap syndrome using the classification criteria of Bastuji-Garin and associates as well as by biopsy. Patients were excluded if skin biopsy resulted in a histopathologic diagnosis that was inconsistent with the clinical diagnosis or if an ophthalmology consultation was not obtained. Charts with incomplete data were also excluded from the study. Fourteen of the 96 cases were excluded because of lack of appropriate consultation or lack of complete data, leaving 82 patients in the sample group.
Records were reviewed for age, sex, race, cause of the disease process, acute ocular complications, acute symptoms, visual acuity, other affected mucosal sites, and length of hospital stay. Length of stay was defined as the duration of hospitalization during the acute phase. Admission data recorded within the first 24 hours of admission including age, presence of malignancy, total body surface area detached, tachycardia, serum urea, serum glucose, and serum bicarbonate were also reviewed to calculate the SCORTEN. The SCORTEN value can range from 0 to 7 and is determined by giving 1 point for each of the following risk factors present: age greater than 40 years, heart rate on admission greater than 120 beats/ minute, initial BSA epidermal detachment greater than 10%, history of malignancy, blood urea nitrogen greater than 28 mg/dL, HCO 3 less than 20 mg/dL, and glucose greater than 252 mg/dL.
Ocular manifestations documented by ophthalmologic bedside consultation during hospitalization were classified as mild, moderate, or severe. The examinations included in the study were the initial examinations performed at the time of consultation as well as ophthalmologic examinations until the patient’s discharge or death. Bedside examination consisted of a penlight examination and/or portable slit-lamp examination. Visual acuity was recorded using a near vision testing card with a +3.00 diopter lens for individuals over the age of 40. The criteria used were adapted from those used by Power and associates. Mild ocular involvement consisted of eyelid edema, eyelid skin involvement including denudation and desquamation, mild conjunctival injection, mucous discharge, and/or chemosis. Moderate involvement consisted of membranous conjunctivitis, corneal epithelial defects, more than 30% healing with medical treatment, corneal ulceration, and/or corneal infiltrates. Severe involvement consisted of acquired eyelid malpositions (ie, ectropion or entropion), symblepharon formation, and/or nonhealing corneal epithelial defects, and/or visual loss, and/or conjunctival fornix foreshortening.
Statistical analysis was performed by χ 2 where appropriate. Analysis of variance was used to compare age as well as SCORTEN values between groups. Pearson product moment correlation was used to compare actual mortality with mortality predicted by SCORTEN value. A probability level of P < .05 was considered statistically significant.
Results
A total of 82 patients had biopsy-proven disease and met all inclusion criteria ( Table 1 ). Mean age was 42.6 years (range 2 to 83 years), and there were 31 cases of SJS, 42 cases of TEN, and 9 cases of SJS/TEN overlap. Patients with SJS were significantly more likely to have a greater chance of survival than those with SJS/TEN overlap or TEN.
| Characteristic | SJS | SJS/TEN Overlap | TEN | Total | P |
|---|---|---|---|---|---|
| Patients (n) | 31 | 9 | 42 | 82 | |
| Age (years), mean ± SD | 41.3 ± 25.2 | 55.0 ± 15.2 | 40.9 ± 22.4 | 42.6 ± 23 | .23 a |
| Age (years), range | 5–83 | 29–71 | 2–82 | 2–83 | |
| Male sex, n (%) | 12 (38) | 5 (55) | 23 (55) | 40 (49) | .36 b |
| Race, n (%) | .75 b | ||||
| White | 15 (48) | 6 (67) | 17 (41) | 38 (46) | |
| Black | 8 (26) | 3 (33) | 15 (36) | 26 (32) | |
| Hispanic | 6 (19) | 0 (0) | 6 (14) | 12 (15) | |
| Asian | 2 (7) | 0 (0) | 3 (7) | 5 (6) | |
| Other | 0 (0) | 0 (0) | 1 (2) | 1 (1) | |
| SCORTEN, mean ± SD | 2.3 ± 1.0 | 3.2 ± 1.4 | 2.9 ± 0.9 | 2.7 ± 1.1 | .014 a |
| Mortality, n (%) | 3 (9) | 3 (33) | 16 (38) | 22 (26) | .023 b |
| Days from admission to death, mean ± SD | 12.3 ± 7.6 | 45.3 ± 46.4 | 11.1 ± 10.8 | 16 ± 21.0 | |
| Days from admission to death, range | 4–19 | 3–95 | 2–35 | 2–95 | |
| Days of hospitalization, mean ± SD | 16 ± 12 | 27 ± 27 | 15 ± 23 | 17 ± 19 | |
| Days of hospitalization, range | 2–60 | 3–95 | 2–132 | 2–132 |
The most frequent cause of disease among all 3 groups was medication (94% [77/82]). There was only 1 case of SJS/TEN overlap that was the result of a viral infection. There were no identified cases of Mycoplasma as the cause, although this was tested for. In only 4 of the 82 patients (5%) were there no identifiable causes. Twelve patients were listed as having multiple drug triggers. Antibiotics were the most common drug to incite all diseases and anticonvulsants were second.
The eyes were affected in 69 of the 82 patients (84%). Overall, the eye findings were mild in 44%, moderate in 20%, and severe in 20%. Most manifestations were mild in both the SJS and TEN groups (39% and 50%, respectively) ( Table 2 ). In the SJS/TEN overlap group, 100% of patients had some type of eye involvement and had an even distribution of mild, moderate, and severe ocular manifestations. While the TEN group exhibited eye involvement in 90% of patients, the SJS group was significantly less likely to have eye involvement (71%) ( P = .03). Of those with eye involvement, most (73%) experienced conjunctivitis. Eyelid skin desquamation was the second most encountered ocular finding (65%).
| None, n (%) | Mild, n (%) | Moderate, n (%) | Severe, n (%) | Total With Eye Involvement, n (%) | |
|---|---|---|---|---|---|
| SJS | 9 (29) | 12 (39) | 4 (13) | 6 (19) | 22 (71) |
| SJS/TEN overlap | 0 (0) | 3 (33) | 3 (33) | 3 (33) | 9 (100) |
| TEN | 4 (10) | 21 (50) | 10 (24) | 7 (16) | 38 (90) |
| Total | 13 (16) | 36 (44) | 17 (20) | 16 (20) | 69 (84) |
A SCORTEN value was calculated for each patient. The overall mean SCORTEN was 2.7 ± 1.1. Twenty-two of the 82 (27%) patients died during hospitalization. The causes of death were cardiovascular collapse (64% [14/22]), sepsis (27% [6/22]), hypovolemic shock (14% [3/22]), and 18% (4/22) succumbing to their pre-existing systemic disease (ie, metastatic cancer). Some patients had multiple causes of death. The mean time to death from admission was longest for the SJS/TEN overlap group (mean 45.3 days) and shortest for the TEN group (mean 11.1 days). The longer time from admission to death in the SJS/TEN overlap group was skewed by 1 patient, whose time from admission until death was 95 days. The SCORTEN value was a good predictor of mortality ( Table 3 , Figure ), with a correlation coefficient of 0.97 ( P = .005). There was not a statistically significant correlation between the SCORTEN value and severity of eye involvement ( P = .65) ( Table 4 ).
| SCORTEN | No. of Patients | Predicted Mortality (%) | No. of Deaths | ||
|---|---|---|---|---|---|
| Actual | Actual (%) | Predicted | |||
| 0 | — | 1.2 | — | — | — |
| 1 | 10 | 3.9 | 1 | 10.0 | 0.4 |
| 2 | 24 | 12.2 | 4 | 16.7 | 2.8 |
| 3 | 30 | 32.4 | 10 | 33.3 | 9.7 |
| 4 | 14 | 62.2 | 6 | 42.9 | 8.7 |
| 5 | 4 | 85.0 | 3 | 75.0 | 3.4 |
| 6 | — | 95.1 | — | — | — |
| 7 | — | 98.5 | — | — | — |
| Total | 82 | 24 | 25 | ||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
