Progression of Toxic Epidermal Necrolysis Following Tanning Bed Exposure

Abstract

Background

In addition to recreational tanning bed use, these ultraviolet radiation exposures are sometimes sought to self-treat skin conditions. The ability of tanning bed exposure to trigger toxic epidermal necrolysis has not been reported.

Observations

We describe a case of an unfortunate young woman who attempted to treat a self-limiting drug hypersensitivity reaction by obtaining a tanning bed treatment, which resulted in a systemic toxic epidermal necrolysis-like reaction. Studies with cultured keratinocytes and an epithelial cell line reveal that ultraviolet A radiation can synergize with other stimuli such as phorbol esters or interleukin-1 to produce large amounts of tumor necrosis factor-alpha, providing a potential mechanism for this exaggerated reaction.

Conclusion

This case report suggests that in addition to inducing photodamage and skin cancer, tanning bed exposure can trigger a toxic epidermal necrolysis-like reaction, possibly via the exaggerated production of keratinocyte cytokines like tumor necrosis factor-alpha.

Introduction

Drug hypersensitivity reactions exist as a spectrum from a mild self-limiting skin eruption to the more ominous Stevens-Johnson Syndrome and Toxic Epidermal Necrolysis (TEN)^1,2. The pathogenesis of these reactions involves cytotoxic T cells and cytokines including tumor necrosis factor-alpha (TNF-α)^3,4. Ultraviolet radiation including UVA and UVB have profound effects on human skin and large doses can mimic the epidermal cytotoxic effects seen in a severe drug hypersensitivity reaction⁵. UVB triggers cytokine production, including TNF-α in keratinocytes^6,7. Recent in vitro studies have demonstrated that UVB can synergize with other stimuli such as phorbol esters and interleukin-1α (IL-1α) to trigger large amounts of TNF-α^8,9. In contrast, UVA1 (wavelengths of 340–400nm) treatment does not result in significant TNF-α release in human keratinocytes⁷. The present case report describes a young woman who sought self-treatment for a mild uncomplicated drug hypersensitivity reaction by obtaining a tanning bed treatment, which resulted in the progression of an uncomplicated drug hypersensitivity reaction to TEN. Moreover, the ability of UVA to synergize with other stimuli to induce TNF-α production in human keratinocytes was tested.

Case Report

A 22 year old female with Fitzpatrick type II skin presented to our hospital with a two day history of red burning and blistering skin, including involvement of oral, ocular and vaginal mucosa. Patient was otherwise healthy, on no other medications, though she had a history of abnormal reactions to non-steroidal anti-inflammatory drugs, which included swelling of her lips in response to acetaminophen and ibuprofen. The patient had ingested 200mg of ibuprofen for menstrual discomfort five days previously. Within twenty four hours of the ibuprofen ingestion, she noticed mild lip swelling and a minimally symptomatic eruption consisting of small red macules and papules on her chest, back and proximal arms. Two days after the skin eruption began, in an attempt to self treat what was a stable non-progressing rash, she went to a local tanning salon where she has a history of frequenting 4–5 times in the past year and received an approximately eight minute exposure in a SunDazzler Stand-Up Tanning Bed (Heartland Tanning, Inc., Lee's Summit, MO). It should be noted that the patient wore her blue jeans, brassiere and used an ocular shield during the tanning session.

Within approximately four hours following the tanning bed exposure the patient noted increased itchiness of her tanning bed exposed skin. The next morning the patient experienced severe redness, pain and the beginning of blister formation on her tanning bed exposed abdomen, back, face, and proximal arms. Moreover, she developed redness and pain on her lips and oral cavity, including her throat. Mild vaginal irritation was also noted. The patient was afebrile and was able to take liquids. The patient then went to an emergency room at her local hospital and was felt to have a streptococcal infection or drug eruption and was given intramuscular corticosteroids and oral Amoxicillin. The patient was discharged from the emergency room but returned approximately six hours later for worsening symptoms. At this time the patient was hospitalized and within twenty-four hours she was intubated and given fluid boluses for episodes of hypotension. A vaginal culture was performed which did not reveal Staphylococcus aureus or group A streptococcus (GAS). A throat culture was negative for GAS. Following the cultures, the patient was placed on i.v. Vancomycin. With her worsening symptoms, the patient was transferred to our medical center. Upon arrival, she was placed in our burn unit and dermatology, ophthalmology and gynecology services were consulted. She was afebrile and her blood pressure was stable. As shown in Figure 1A,B, the patient exhibited blister formation on the areas exposed to the tanning bed radiation, especially chest, abdomen, proximal arms and upper back (approximately 35% body surface area). Laboratory tests revealed mild transaminitis (AST = 80 [nl < 45]), but were otherwise normal.

Figure 1. Clinical Photographs Demonstrating Photo-accentuation of Blistering Skin Eruption and Biopsy Specimen of Skin Showing Epidermal Necrosis and Increased TNF-α Immunoreactivity.

A,B-clinical pictures of patient demonstrating blisters in areas of tanning bed exposure and residua of mild morbilliform eruption on covered skin (breasts). C-. Histology of skin biopsy specimen demonstrating necrotic keratinocytes (100×). D,E- Immunohistochemical studies demonstrating increased cytoplasmic TNF-α immunoreactivity in patient (D) over control tissue (E); (200×)

Dermatology suspected TEN and performed skin biopsies which revealed a lymphocytic infiltrate at the dermal-epidermal junction with some apoptotic keratinocytes consistent with TEN (Fig. 1C). Ophthalmology noted ocular involvement also consistent with TEN and the gynecological service found erythema and swelling of her vaginal introitus with several small ulcerations. The patient was started on I.V. Cyclosporine 5mg/kg/day and was treated supportively in the burn unit. After two days, due to lack of response and some extension of the blisters onto groin and proximal thighs, the Cyclosporine was discontinued and the patient received three days of intravenous immunoglobulin G (1 g/kg/day). The patient slowly recovered and was extubated after one week. After a total of almost three weeks, the patient was discharged to home and has done very well, except for mild ocular light sensitivity.

Methods

Immunohistochemistry

Immunohistochemistry with a human anti-TNF-α antibody (Assay Designs, Ann Arbor, MI) was performed on a formalin-fixed, paraffin-embedded tissue from the patient's skin biopsy as previously described¹⁰. Normal skin from a different individual was used as a control for TNF-α immunohistochemistry.

Keratinocyte studies

Primary cultures of human keratinocytes were grown from human neonatal foreskins as previously described, and the human epithelial cell line KB was used and culture conditions as previously reported^10,11. Keratinocytes approximately 80–90% confluent were treated with 10 nM phorbol myristic acetate (PMA; Sigma Chemical Co, St. Louis, MO) or 1 ng/ml IL-1α (Peprotek, Rocky Hill, NJ) for thirty minutes before treatment with UVA. KB cells were approximately 50–70% confluent and were treated similarly with PMA +/− UVA. The broadband UVA source (range 315–400 nm; peak 350 nm) consisted of a bank of four Philips PL-L 36W bulbs (Philips International B.V., Amsterdam, The Netherlands). The intensity of the tanning bed and the UVA source used for the experiments were measured using an IL1700 radiometer and a SED240 UVB detector and an SED033 UVA detector (International Light Technologies, Inc. Peabody, MA) at a distance of 8 cm from the radiation source. The cells were incubated for six hours and TNF-α was measured in supernatants using EIA kit (Assay Designs, Ann Arbor, MI), and cells were then treated with trypsin and counted.

Results

Several studies have described that the combination of ultraviolet B radiation (UVB) and the cytokine IL-1 or the phorbol ester phorbol myristic acetate (PMA) results in synergistic TNF-α production in human keratinocytes^8,9. Given that TNF-α has been implicated in the pathogenesis of TEN^3,12, we hypothesized the experimental finding of exaggerated keratinocyte cytokine production by the combination of UV radiation and IL-1 or PMA could explain how a tanning bed exposure in the setting of a mild uncomplicated drug reaction could result in the patient's presentation. Immunohistochemical staining of the patient's skin biopsy revealed significant cytoplasmic TNF-α immunoreactivity (Fig. 1D) in comparison to control skin from another patient (Fig. 1E). Next, we traveled to the tanning salon and measured the UV output of a SunDazzler Tanning Bed using a radiometer, which was 20 W/m² UVA and 0.13 W/m² UVB. Thus, the patient received approximately 9.6 kJ/m² UVA and 62 J/m² UVB for her eight minute exposure. These findings fit with literature values for a mild tanning bed exposure^13,14.

To test the ability of UVA to trigger a synergistic response as is seen in response to UVB, we exposed primary cultures of human keratinocytes to either vehicle, the phorbol ester PMA, IL-1α alone or 30 min before irradiation with sham or two doses (5, 20 KJ/m²) of UVA. In other experiments, the human epitheliod cell line KB was treated with PMA alone or 30 min before UVA radiation. At six hours post radiation, TNF-α protein released into the supernatants was measured by ELISA. As shown in Fig 2, UVA did not stimulate appreciable TNF-α production in primary cultures of HK or KB cells. However, the combination of UVA with IL-1α or PMA resulted in the synergistic (~ four- to six-fold increased) production of TNF-α protein. It should be noted the synergistic TNF-α production seen with UVA is less than those reported for UVB with these stimuli^8,9. These studies indicate that the combination of UVA and other stimuli such as IL-1α or PMA induce synergistic TNF-α production in human keratinocytes.

Figure 2. UVA Irradiation Augments TNF-α Production in Human Keratinocytes and KB cell line.

Primary cultures of human keratinocytes or KB cell line were treated with vehicle control, 1 ng/ml IL-1α, or 10 nM PMA, followed 30 min later by treatment with sham, 5 or 20 kJ/m² UVA. The TNF-α released into the supernatants was measured 6 h following UVA treatment by EIA. The data depicted are mean +/− SD from triplicate samples of a representative experiment from at least four with similar results.

Discussion

Tanning bed exposure has been linked to skin cancers including malignant melanoma^15,16. Moreover, tanning beds are involved in approximately 700 emergency room visits per year⁵. The present case report describes a patient who apparently developed an uncomplicated mild drug reaction from ingesting ibuprofen, and several days after developing the eruption, attempted to “self-treat” with tanning bed, resulting in TEN. Though a severe sunburn reaction would be in the differential diagnosis, the history of a cutaneous exanthema preceding the tanning bed exposure as well as mucous membrane involvement fit better with TEN. In addition, that almost 72 hours had passed between the ibuprofen ingestion and tanning bed exposure also suggests this was not a phototoxic reaction to this drug. Of note, there is a previous report of photo-accentuated TEN with high levels of TNF-α measured in blister fluid from the affected patient, though tanning bed exposure was not involved¹⁷. The present studies also suggest a mechanism by which this photo-accentuated reaction could have occurred, namely that UVA exposure in the setting of keratinocytes already activated from the pre-existing drug reaction resulted in an exaggerated production of epidermal cytokines such as TNF-α. The current in vitro studies using keratinocytes demonstrate that combining IL-1α or PMA with UVA results in the synergistic production of keratinocyte TNF- α fit with this notion. Of interest, we have found that other keratinocyte-derived cytokines including the T cell chemokine CCL-20 are also highly up-regulated in response to the combination of UV and PMA/IL-1α (data not shown). The present in vitro studies probably would underestimate the exaggerated TNF-α response from a tanning bed as the low amount of UVB found in the tanning radiation was not tested.

There is an increasing trend for patients to seek a tanning bed for self therapy as the general public's perception of tanning beds includes the ability to treat skin rashes. Of note, a recent report of over 1200 subjects indicated that almost 10% of those using tanning salons did so in response to treatment of skin disease, and 5% were tanning due to their physician's recommendation¹⁸. The current report should provide caution to those who recommend tanning beds for the treatment of undiagnosed skin conditions.

In summary, these studies describe a clinical case report of an unfortunate woman who developed TEN from the combination of a cutaneous drug reaction and tanning bed exposure. Treatment of cultured keratinocytes activated by IL-1α or PMA with approximately similar doses of UVA that the patient experienced resulted in an exaggerated production of TNF-α, providing a potential mechanism for this clinical scenario.