Table of Content

Archive

Autumn 2015, Vol. 23 No. 3

Hong Kong J. Dermatol. Venereol. (2015) 23, 121-125


Case Report

Topical sulphamethoxazole ophthalmic solution induced toxic epidermal necrolysis

由外用磺胺甲噁唑眼藥水引發的中毒性表皮壞死鬆解症

LC Wu 吳隆基 and CC Chen 陳志強

Abstract

Toxic epidermal necrolysis (TEN) is a severe adverse drug reaction that leads to widespread mucocutaneous necrosis. Medications are the most common cause of TEN. However, some cases develop following infection or vaccination. TEN is very rarely induced by topical ophthalmic solution or topical agent. We report a case of a 31-year-old female who developed TEN after using topical sulphamethoxazole ophthalmic solution. Skin biopsy was performed and revealed upper epidermal necrosis with cell-poor inflammatory infiltrate. After admission, the patient received systemic and topical corticosteroid, as well as supportive treatment. The prognosis was good at follow-up 21 days later.

中毒性表皮壞死鬆解症是一種嚴重的藥物不良反應,會導致皮膚粘膜廣泛地壞死。固而言之,藥物是常見的成因。然而,一些病例是在感染後或疫苗接種後發生的,而外用眼藥水或外用藥劑引起的個案則少之又少。我們報告一例一名三十一歲女仕在使用外用磺胺甲噁唑眼藥水後發生的中毒性表皮壞死鬆解症個案。患處的皮膚活檢組織發現上表皮壞死伴隨不多的炎性細胞浸潤。患者入院後,接受了全身和外用的皮質類固醇,並輔以支持治療。在二十一天後的覆診,可見良好的預後。

Keywords: Erythema multiforme, ocular solutions, Stevens-Johnson syndrome, sulphamethoxazole, toxic epidermal necrolysis

關鍵詞: 多形性紅斑、眼部溶液、史蒂芬強森症候群、磺胺甲噁唑、中毒性表皮壞死鬆解症

Introduction

Toxic epidermal necrolysis (TEN), also called Lyell's disease, is a severe adverse drug reaction that leads to widespread mucocutaneous necrosis. Drugs such as sulphonamides, pyrazolones, barbiturates, and antiepileptics or their metabolites are the most common trigger of host inappropriate immune activation and attack on keratinocytes. TEN is preceded by a prodrome that includes fever, symptoms of upper respiratory infection, anorexia and malaise. Subsequently, generalised dusky, erythematous skin rash, skin sloughing and blisters appeared symmetrically over the face, neck and extremities. Purulent conjunctivitis and mucositis over the mouth and genital area may occur at the same time or a few days later.1 Ocular manifestations vary in severity and include conjunctivitis, eyelid swelling, discharge, corneal ulcer, cicatrising symblepharon, fornix foreshortening, and panophthalmitis. Although medications are the most common cause of TEN, some cases develop the condition after an infection, such as post-mycoplasma or herpes simplex virus infection, or post-vaccination. TEN is very rarely induced by topical ophthalmic solution or topical agent. We report a case of TEN after using topical sulphamethoxazole eye drops.

Case report

A 31-year-old female was admitted to our emergency department due to severe eye pain and extensive skin eruptions for three days. According to her medical records, there were no significant past medical problems and had no history of food or drug allergy. Four days prior to admission, she had visited an ophthalmology clinic due to persistent bilateral ocular itching for several days. She denied taking any medication or dietary supplement before the ocular manifestation. The ophthalmologist had prescribed two kinds of eye drop solutions for the provisional diagnosis of conjunctivitis. The trade name of one of the solutions was Ming-Mei, of which the major components are neomycin and dexamethasone. The trade name of the other solution was Saloon, of which the major component is sulphamethoxazole. After using both solutions for one day, the patient suffered from skin rash, headache, fever, cough, and rhinorrhoea. She continued to use the solutions and the skin lesions progressed diffusely over the face, trunk and limbs.

Physical examination revealed a heart rate of 85 beats/min, marked eyelid oedema, conjunctival chemosis and discharge, multiple symblephara with bilateral corneal ulceration (Figure 1A, Right eye; Figure1B, Left eye), intense facial oedema, erythema, and bullae formation with detachment (Figure 1F) involving up to 40% of the body surface area (Figure 1C), especially in the oral mucosa (Figure 1D) and genital area (Figure 1E). Laboratory results included normal complete blood count, serum urea: 20 mg/dl, serum glucose (A.C.): 118 mg/dl, serum bicarbonates: 24 meq/L; mildly elevated C-reactive protein (0.58 mg/dL); normal procalcitonin; normal urine routine microscopy; negative sputum culture; negative serology for mycoplasma and toxoplasma; and negative serology for viruses (cytomegalovirus and Epstein Barr virus). Wound culture and peripheral blood culture were taken 48 hours after admission. Wound culture was positive for Enterococcus species, while peripheral blood culture was negative. Skin biopsy revealed confluent necrosis of keratinocytes in the upper part of the epidermis and detachment with cell-poor inflammatory infiltrate in the epidermis. There were vacuolisation changes at the basement membrane and a few melanophages in the papillary dermis (Figures 2A-D). The patient was diagnosed as TEN induced by topical sulphamethoxazole eyedrops according to her clinical appearance and histopathological findings. Skin eruptions were observed over 40% of total body surface area. Thus, the SCORTEN scale score was one with an associated mortality rate of about 3.2%.2

Figure 1 Corneal ulceration (A: Right eye, B: Left eye); (C) Skin eruption involving up to 40% of the body surface area; (D) Erythema, oedematous swelling, blister formation in the oral mucosa; (E) Genital area; (F) Intense facial oedema, erythema, bullae formation with detachment.

Figure 2 Skin biopsy showing confluent necrosis of keratinocytes in the upper part of the epidermis and detachment with cell-poor inflammatory infiltrate. There are vacuolation changes at the basement membrane and a few melanophages in the papillary dermis. (A) Haematoxylin and eosin stain, original magnification x40; (B) Haematoxylin and eosin stain, original magnification x400; (C) Haematoxylin and eosin stain, original magnification x400; (D) Haematoxylin and eosin stain, original magnification x100.

After admission, 80 mg of hydrocortisone was administered intravenously every eight hours accompanied by intensive wound care and life support treatment. Ophthalmologist was consulted for bilateral corneal ulceration and symblephara. A glass rod was used to lyse the symblephara. The patient was also treated with fluorometholone eye drops and artificial tears every two hours. During admission, no new lesions occurred and the mucocutaneous erosion healed and re-epithelialised. After using intravenous hydrocortisone for five days, treatment was changed to oral prednisolone 20 mg three times a day and the dosage was subsequently tapered. At day 14, oral medications were discontinued and only topical agents were continued. At 21 day follow-up, total resolution of the mucocutaneous lesions without sequelae was observed.

Discussion

TEN, Stevens-Johnson syndrome (SJS) and erythema multiforme (EM) are a spectrum of mucocutaneous diseases. According to the classification of Bastuji-Garin et al,2 this spectrum can be divided into five categories: (1) Bullous EM: detachment over <10% of body surface area (BSA) plus typical or atypical targets (2) SJS: detachment over <10% of BSA plus widespread purpuric or erythematous macules, or atypical targets (3) Overlap SJS/TEN: detachment between 10% and 30% of BSA plus widespread purpuric or erythematous macules, or atypical targets (4) TEN with spots: detachment over >30% of BSA plus widespread purpuric or erythematous macules, or atypical targets (5) TEN without spots: detachment over >10% of BSA plus large epidermal sheets without purpuric macules or targets.

EM is a mild clinical condition with frequent recurrence that is associated with herpes simplex virus or mycoplasma infection. SJS and TEN have a more serious clinical course and higher mortality rate. The incidence of TEN is 2.7 times higher among elderly patients than among younger patients. The mortality rate of TEN is twice as high in the elderly population (about 51%) than in the younger population (about 25%). We used a severity-of-illness scale, SCORTEN, to evaluate the severity and predict the mortality.

The pathogenesis of TEN is thought to involve either a Fas or a granulysin mediated apoptotic pathomechanism. However, reactive oxygen species (ROS) have been recognised for their involvement in intracellular damage and precede the activation of pro-apoptotic systems.1 Many medications have been reported to cause TEN, including sulphonamide antibiotics, anticonvulsants, allopurinol and oxicam nonsteroidal anti-inflammatory drugs.3

There are two well established principles to improve the survival rate of TEN. One is early withdrawal of the offending drug and the other is prompt referral to a burns unit. Systemic corticosteroids were considered first-line treatment for TEN in the past. However, this concept has become debatable nowadays. Some studies showed that steroids are associated with an increased mortality, higher rates of sepsis and prolonged hospital admission. In contrast, other studies demonstrated that early use of intravenous corticosteroid can reduce mortality and prevent ocular complications of cicatrisation as well as preserving the visual acuity.4 To date, intravenous immunoglobulin has been demonstrated to be effective in several studies.1 In addition, plasmapheresis, ciclosporin, anti-tumour necrosis factor therapy and N-acetylcysteine are also found to be helpful.5 However, more randomised control studies are required to prove their benefits.

In addition to EM and SJS/TEN, generalised bullous fixed drug eruption (GBFDE) should be considered in the differential diagnosis. There are several characteristic features of GBFDE that can help to distinguish it from SJS/TEN even when mucosal lesions appear to be quite similar:

  1. During the clinical course, GBFDE has a shorter latent period with less mucosal involvement.
  2. In terms of histology, GBFDE results in more eosinophil infiltration and dermal melanophages.
  3. Immunohistopathologically, there are more dermal CD4+ cells and Foxp3+ cells and fewer intraepidermal CD56+ and granulysin+ cells in the lesional sites.6

Due to several similar and overlapping clinical and pathological findings between SJS/TEN and GBFDE, it is sometimes difficult to distinguish between them. In this case, confluent necrosis of the upper half of epidermis without a dermal eosinophil infiltrate favours SJS/TEN. The basal layer vacuolisation change and focal dermal melanophages could be noted both in SJS/TEN and GBFDE, although the dermal melanophages might be more prominent in GBFDE in some reports.6 As for the time of onset of skin manifestations, wide variation has been noted in various reports, ranging from 24 hours to 9 days in both GBFDE and SJS/TEN.7-11 Serious involvement of ocular, oral and genital mucosae in our case provided crucial evidence for the diagnosis of SJS/TEN. Therefore, with the combination of histopathological findings and clinical manifestations, SJS/TEN is more likely to be the diagnosis in our case.

SJS or TEN is rarely induced by topical ophthalmic agent when compared with oral or intravenous medications. Only five case reports have been published over the past 20 years. Byrom et al presented a case of a 15-year-old boy who developed TEN due to sulphacetamide eye drops.8 Florez et al reported a 60-year-old woman who suffered from TEN after using an ophthalmic agent that contained dorzolamide, timolol, and latanoprost.9 Howard et al described a case of a 9-year-old boy who developed TEN after being prescribed sulphonamide eye drops.7 Two other cases have been published in French and Spanish, respectively.10,11 Although our patient denied any family or personal history of sulpha drug allergy, the rapid clinical progression of TEN and severe side effects of topical preparations with sulphur derivatives in our patient may indicate that this is not the first-time exposure. We postulated that our patient probably has been exposed to sulphur derivatives before, during which the symptoms were subclinical. The symptoms might have been too mild to be recognised by the patient. For this episode, the topical preparations with sulphur derivatives could be the exposure for the second-time, thus explaining the rapid immune sensitisation and clinical progression. Nevertheless, topical neomycin cannot be totally excluded as a cause in this case. According to the review article of Sasseville,12 neomycin can cause urticaria, allergic contact dermatitis, exfoliative dermatitis and erythroderma. However, there are no case reports of SJS or TEN induced by neomycin. Clinicians should be aware of the signs of the prodrome of TEN to avoid this life-threatening condition.

References

1. Downey A, Jackson C, Harun N, Cooper A. Toxic epidermal necrolysis: review of pathogenesis and management. J Am Acad Dermatol 2012;66:995-1003.

2. Bastuji-Garin S, Rzany B, Stern RS, Shear NH, Naldi L, Roujeau JC. Clinical classification of cases of toxic epidermal necrolysis, Stevens-Johnson syndrome, and erythema multiforme. Arch Dermatol 1993;129:92-6.

3. Mockenhaupt M, Viboud C, Dunant A, Naldi L, Halevy S, Bouwes Bavinck JN, et al. Stevens-Johnson syndrome and toxic epidermal necrolysis: assessment of medication risks with emphasis on recently marketed drugs. The EuroSCAR-study. J Invest Dermatol 2008;128:35-44.

4. Araki Y, Sotozono C, Inatomi T, Ueta M, Yokoi N, Ueda E, et al. Successful treatment of Stevens-Johnson syndrome with steroid pulse therapy at disease onset. Am J Ophthalmol 2009;147:1004-11.

5. Fernando SL. The management of toxic epidermal necrolysis. Australas J Dermatol 2012;53:165-71.

6. Cho YT, Lin JW, Chen YC, Chang CY, Hsiao CH, Chung WH, et al. Generalized bullous fixed drug eruption is distinct from Stevens-Johnson syndrome/toxic epidermal necrolysis by immunohistopathological features. J Am Acad Dermatol 2014;70:539-48.

7. Fine HF, Kim E, Eichenbaum KD, Antoniades K, Williams CA. Toxic epidermal necrolysis induced by sulfonamide eyedrops. Cornea 2008;27:1068-9.

8. Byrom L, Zappala T, Muir J. Toxic epidermal necrolysis caused by over the counter eyedrops. Australas J Dermatol 2013;54:144-6.

9. Florez A, Roson E, Conde A, Gonzalez B, Garcia-Doval I, de la Torre C, et al. Toxic epidermal necrolysis secondary to timolol, dorzolamide, and latanoprost eyedrops. J Am Acad Dermatol 2005;53:909-11.

10. Asensio-Sanchez VM. Toxic epidermal necrolysis following dorzolamide eyedrops. An Med Interna 2008;25:47-8.

11. Schmutz JL, Barbaud A, Trechot P. Toxic epidermal necrolysis following eyedrop treatment for glaucoma. Ann Dermatol Venereol 2007;134:417.

12. Sasseville D. Neomycin. Dermatitis 2010;21:3-7.