ctinical and Experimentat Dermatology 1992; 17: 456-457.
Toxic epidermal necrolysis localized to an area of lymphoedema S.M.WILKINSON, A.H.M.HEAGERTY AND A.G.SMITH Department of Dermatology, North Staffordshire Hospital Centre, Stoke-on-Trent ST4 7FA, UK Accepted for publication 31 October 1991
Summary We present a man who, on two occasions, developed toxic epidermal necrolysis initially localized to an area of lymphoedema. To our knowledge, lymphoedema as a site of prediliction has not previously been reported. The cutaneous lesions of toxic epidermal necrolysis usually begin as a symmetrical eruption beginning on the face and upper trunk with subsequent generalization.' Sites of pressure from clothing may be less affected and a predominance in light-exposed areas may be seen.^ Preferential involvement of specific areas ofthe body has not previously been reported.
erosions ofthe mucosal surface of his lips and eyelids with sheered erythema and blistering of his left leg and a small {20 cm in diameter) area of his back. Nikolski's sign was positive on the lefr leg and back. He was afebrile but tachypnoeic with widespread crackles on auscultation of his chest. Physical examination was otherwise unremarkable. A skin biopsy showed a superficial perivascular lymphocytic infiltrate with epidermal necrosis consistent
Case report An 82-year-old man presented in 1991 with a history of blistering of his left leg 3 days after being given ampicillin for a chest infection. He subsequently developed lesions on his back and mouth. His past history included excision of a squamous cell carcinoma from the dorsum of his left foot in 1984. The following year he developed inguinal lymphadenopathy and underwent a block dissection of these nodes with the development of lymphoedema. He was well until 1989 when, following a fall, he sustained a subcapital fracture ofthe left neck ofthe femur which wasfixedwith screws. Post-operatively, the wound failed to heal and a sinus tract formed. One month later, he was readmitted with an infected joint and loosening ofthe screws. Blood cultures grew Streptococcus sanguis and a wound swab Staphylococcus aureus which were treated with flucloxacillin and ampicillin. Three days later, he developed a blistering eruption of his mouth and left leg. He gave a history of a similar reaction to cotrimoxazole and his antibiotic was changed to erythromycin with resolution of his rash. On admission to hospital in 1991 he had crusting and Correspondence: Dr S.M.Wilkinson, Dermatology Department, Central Outpatient Department, Hartshill, Stoke-on-Trent ST4 7PA, UK.
456
Figure 1. Sheeted erythema and liesquammation of skin on a Iymphoedematous limb with uninvolved right leg behind.
TOXIC EPIDERMAL NECROLYSIS
457
with toxic epidermal necrolysis. A full blood count revealed a normochromic normocytic anaemia with a haemoglobin of 10 g/dl and normal white cell count and platelets. His urea was elevated at 32-3 mmol/1 with an elevated creatinine of 817 /imol/1. Renal ultrasound was normal. Chest X-ray, sputum and blood culture were normal. A skin swab from the left hip sinus grew Proteus spp. and one from an area of blistering grew normal flora. He was rreated with intravenous fluids and topical antiseptics and was nursed on a soft surface. His skin gradually re-epithelialized over the ensuing 3 weeks. However, his recovery was complicated by the development of acute urinary retention and serial chest X-rays showed the development of right lower lobe consolidation. The pneumonia was treated with inhaled steam, physiotherapy and erythromycin.
brane in the dermal type and through the epidermis in rhe epidermal form of the disease. It is possible that the dermal oedema present in lymphoedema may hasten the shedding ofthe epidermis when a split develops below the basement membrane. However, this does nor explain the original confinement of the inflammatory process to this area. Ir is interesting to speculate rhat restriction to areas of lymphoedema might occur becau.se of poor clearance of the offending drug from the interstitial fluid by the damaged lymphatics. A similar argument is used to explain rhe development of xanrhoma localized to areas of lymphoedema.""** This would provide some support for rhe idea of using plasmapheresis early in the course of roxic epidermal necrolysis to reduce drug levels and consequently disease severity.^'**
Discussion
References
Both of our patient's episodes of toxic epidermal necroly1. Roujeau JC, Chosidow O, Saiag P, Guillaume JC. Toxic epidermal necroly.sis {Lyell ^yndramt) Journat of the Ameriean sis followed the administration of ampicillin. This is a Academy of Dermatology 1990; 23: 1039-1058. recognized cause with a relatively high incidence of toxic 2. Huff JC, Weston WI,, Tonnesen MG. Erythema multiforme: a epidermal necrolysis when compared with the number of eritieal review of eharaeteristics, diagnostic criteria and causes. days of treatment with the drug.' The differential Journal of the American Academy of Dermatotogy \9Hy,8:l(>?.-115. diagnosis of toxic epidermal necrolysis includes the ,1 Schopf E, Stuhmer A, Rzany B, Victor N, Zentgraf R, Kapp JF. Toxic epidermal necrolysis and Stevens-Johnson syndrome: An staphylococcal scalded skin syndrome and generalized epidemiologie study from West Germany, Archives of Dermatofixed drug eruption as well as the Stevens-Johnson logy 1991; 127:839-842, syndrome, which some consider ro be part of a spectrum 4. Lyell A. Requiem for toxic epidermal necrolysis. British Journat of with toxic epidermal necrolysis.' Despite the unusually Dermatotogy 1990; 122: 837-838, localized area of disease we consider that our patient had 5. Orfanos CE, Schaumberg-Lever G, Lever WF, Dermal and epidermal types of erythema multiforme. A histopathologic study toxic epidermal necrolysis. The staphylococcal scaldedof 24 cases. .'Irehives of Dermatology 1974; 109: 682-688. skin syndrome was excluded by the negative cultures and 6. Duff IS, Everall J. Lymphangioma circumscriptum British subepidermal nature ofthe blister. Lyell^ has suggested Journal of Dermatology 1963; 75: 122-123. generalized bullous fixed-drug eruption as an alternative 7. Coburn JG. Extracellular cholesterolosis and lymphoedema, diagnosis in cases where the rash is reliably reproduced by Britisb Journal of Dermatohgy 1963; 75: 128-129." drugs. This seems unlikely in our patient because ofthe 8. Wooling KR, Jenkins RE, Dolan PA and Evans PV. Localised xanthomas in lymphedema praecox. Journal of the Ameriean extent ofthe mucosal involvemenr and the sheeted nature Medical Association 1970; 211: 1372-1374, ofthe eruption rather than the round lesions more typical 9. Gerard A, Schooneman F, Roche G et al. Lyell's syndrome: of fixed drug eruption. The Stevens-Johnson syndrome treatment by plasma exchange. Plasma Therapy Transfusion may result in mucosal involvement but there were no Technology 1984; 5: 259-260. target lesions of erythema multiforme and there was 10. Kamanabroo D, Schmitz-Landgraf W, Czarnetzki BM. Plasmagreater than 15% epidermal loss. pheresis in severe drug-induced toxic epidermal necrolysis. Archives of Dermatology 1985; 121: 1548-1549. Orfanos et al.^ stated that the site of bulla formation in erythema multiforme occurs below the basement mem-
Toxic epidermal necrolysis localized to an area of lymphoedema S.M.WILKINSON, A.H.M.HEAGERTY AND A.G.SMITH Department of Dermatology, North Staffordshire Hospital Centre, Stoke-on-Trent ST4 7FA, UK Accepted for publication 31 October 1991
Summary We present a man who, on two occasions, developed toxic epidermal necrolysis initially localized to an area of lymphoedema. To our knowledge, lymphoedema as a site of prediliction has not previously been reported. The cutaneous lesions of toxic epidermal necrolysis usually begin as a symmetrical eruption beginning on the face and upper trunk with subsequent generalization.' Sites of pressure from clothing may be less affected and a predominance in light-exposed areas may be seen.^ Preferential involvement of specific areas ofthe body has not previously been reported.
erosions ofthe mucosal surface of his lips and eyelids with sheered erythema and blistering of his left leg and a small {20 cm in diameter) area of his back. Nikolski's sign was positive on the lefr leg and back. He was afebrile but tachypnoeic with widespread crackles on auscultation of his chest. Physical examination was otherwise unremarkable. A skin biopsy showed a superficial perivascular lymphocytic infiltrate with epidermal necrosis consistent
Case report An 82-year-old man presented in 1991 with a history of blistering of his left leg 3 days after being given ampicillin for a chest infection. He subsequently developed lesions on his back and mouth. His past history included excision of a squamous cell carcinoma from the dorsum of his left foot in 1984. The following year he developed inguinal lymphadenopathy and underwent a block dissection of these nodes with the development of lymphoedema. He was well until 1989 when, following a fall, he sustained a subcapital fracture ofthe left neck ofthe femur which wasfixedwith screws. Post-operatively, the wound failed to heal and a sinus tract formed. One month later, he was readmitted with an infected joint and loosening ofthe screws. Blood cultures grew Streptococcus sanguis and a wound swab Staphylococcus aureus which were treated with flucloxacillin and ampicillin. Three days later, he developed a blistering eruption of his mouth and left leg. He gave a history of a similar reaction to cotrimoxazole and his antibiotic was changed to erythromycin with resolution of his rash. On admission to hospital in 1991 he had crusting and Correspondence: Dr S.M.Wilkinson, Dermatology Department, Central Outpatient Department, Hartshill, Stoke-on-Trent ST4 7PA, UK.
456
Figure 1. Sheeted erythema and liesquammation of skin on a Iymphoedematous limb with uninvolved right leg behind.
TOXIC EPIDERMAL NECROLYSIS
457
with toxic epidermal necrolysis. A full blood count revealed a normochromic normocytic anaemia with a haemoglobin of 10 g/dl and normal white cell count and platelets. His urea was elevated at 32-3 mmol/1 with an elevated creatinine of 817 /imol/1. Renal ultrasound was normal. Chest X-ray, sputum and blood culture were normal. A skin swab from the left hip sinus grew Proteus spp. and one from an area of blistering grew normal flora. He was rreated with intravenous fluids and topical antiseptics and was nursed on a soft surface. His skin gradually re-epithelialized over the ensuing 3 weeks. However, his recovery was complicated by the development of acute urinary retention and serial chest X-rays showed the development of right lower lobe consolidation. The pneumonia was treated with inhaled steam, physiotherapy and erythromycin.
brane in the dermal type and through the epidermis in rhe epidermal form of the disease. It is possible that the dermal oedema present in lymphoedema may hasten the shedding ofthe epidermis when a split develops below the basement membrane. However, this does nor explain the original confinement of the inflammatory process to this area. Ir is interesting to speculate rhat restriction to areas of lymphoedema might occur becau.se of poor clearance of the offending drug from the interstitial fluid by the damaged lymphatics. A similar argument is used to explain rhe development of xanrhoma localized to areas of lymphoedema.""** This would provide some support for rhe idea of using plasmapheresis early in the course of roxic epidermal necrolysis to reduce drug levels and consequently disease severity.^'**
Discussion
References
Both of our patient's episodes of toxic epidermal necroly1. Roujeau JC, Chosidow O, Saiag P, Guillaume JC. Toxic epidermal necroly.sis {Lyell ^yndramt) Journat of the Ameriean sis followed the administration of ampicillin. This is a Academy of Dermatology 1990; 23: 1039-1058. recognized cause with a relatively high incidence of toxic 2. Huff JC, Weston WI,, Tonnesen MG. Erythema multiforme: a epidermal necrolysis when compared with the number of eritieal review of eharaeteristics, diagnostic criteria and causes. days of treatment with the drug.' The differential Journal of the American Academy of Dermatotogy \9Hy,8:l(>?.-115. diagnosis of toxic epidermal necrolysis includes the ,1 Schopf E, Stuhmer A, Rzany B, Victor N, Zentgraf R, Kapp JF. Toxic epidermal necrolysis and Stevens-Johnson syndrome: An staphylococcal scalded skin syndrome and generalized epidemiologie study from West Germany, Archives of Dermatofixed drug eruption as well as the Stevens-Johnson logy 1991; 127:839-842, syndrome, which some consider ro be part of a spectrum 4. Lyell A. Requiem for toxic epidermal necrolysis. British Journat of with toxic epidermal necrolysis.' Despite the unusually Dermatotogy 1990; 122: 837-838, localized area of disease we consider that our patient had 5. Orfanos CE, Schaumberg-Lever G, Lever WF, Dermal and epidermal types of erythema multiforme. A histopathologic study toxic epidermal necrolysis. The staphylococcal scaldedof 24 cases. .'Irehives of Dermatology 1974; 109: 682-688. skin syndrome was excluded by the negative cultures and 6. Duff IS, Everall J. Lymphangioma circumscriptum British subepidermal nature ofthe blister. Lyell^ has suggested Journal of Dermatology 1963; 75: 122-123. generalized bullous fixed-drug eruption as an alternative 7. Coburn JG. Extracellular cholesterolosis and lymphoedema, diagnosis in cases where the rash is reliably reproduced by Britisb Journal of Dermatohgy 1963; 75: 128-129." drugs. This seems unlikely in our patient because ofthe 8. Wooling KR, Jenkins RE, Dolan PA and Evans PV. Localised xanthomas in lymphedema praecox. Journal of the Ameriean extent ofthe mucosal involvemenr and the sheeted nature Medical Association 1970; 211: 1372-1374, ofthe eruption rather than the round lesions more typical 9. Gerard A, Schooneman F, Roche G et al. Lyell's syndrome: of fixed drug eruption. The Stevens-Johnson syndrome treatment by plasma exchange. Plasma Therapy Transfusion may result in mucosal involvement but there were no Technology 1984; 5: 259-260. target lesions of erythema multiforme and there was 10. Kamanabroo D, Schmitz-Landgraf W, Czarnetzki BM. Plasmagreater than 15% epidermal loss. pheresis in severe drug-induced toxic epidermal necrolysis. Archives of Dermatology 1985; 121: 1548-1549. Orfanos et al.^ stated that the site of bulla formation in erythema multiforme occurs below the basement mem-
