In the past few years, beta-lactam-resistant gram-positive

Early Diagnosis Is Key in Vancomycin-Induced Linear IgA Bullous Dermatosis and Stevens-Johnson Syndrome Douglas H. Jones, MS Michael Todd, MD Timothy J. Craig, DO Background: With the emergence of highly resistant betalactam gram-positive organisms, vancomycin hydrochloride usage has increased considerably. Consequently, adverse drug reactions, including unfavorable cutaneous events, have also increased. Important adverse skin reactions are linear IgA bullous dermatosis (LABD) and Stevens-Johnson syndrome (SJS). These blistering disorders can have clinical manifestations that are difficult to distinguish; however, it is important to make the distinction because treatment and prognosis are different. Objective: The purpose of this study is to review the literature on LABD and SJS and compare important differentiating characteristics to assist physicians in making the correct diagnosis. Methods: The authors used MEDLINE to search for all published studies on vancomycin adverse events, and combined LABD, SJS, exanthema, and skin rashes each separately with vancomycin adverse events. Furthermore, the authors searched PubMed for all meta-analyses and randomized controlled clinical trials relating to treatment of patients with SJS. Results: Clinically, LABD and SJS both present similarly with bullae. Diagnosis is made by use of perilesional skin biopsy and direct immunofluorescence. Direct immunofluorescence shows linear IgA deposition along the basement membrane zone in LABD, whereas this is absent in SJS. The treatment for both vancomycin-induced SJS and vancomycin-induced LABD is prompt discontinuation of the drug. However, if SJS is diagnosed early, systemic corticosteroids appear to decrease morbidity. From the Milton S. Hershey Medical Center at Pennsylvania State University in Hershey, Pa. Address correspondence to Timothy J. Craig, DO, Pennsylvania State University, The Milton S. Hershey Medical Center, Pulmonary, Allergy, and Critical Care Medicine, HO41, 500 University Dr., PO Box 850, Hershey, PA 17033-0850. E-mail: tcraig@psu.edu Conclusions: In cases of SJS or LABD that are difficult to distinguish clinically, the authors recommend performing a skin biopsy and direct immunofluorescence early to confirm the diagnosis so that effective treatment can be instituted. In the past few years, beta-lactam-resistant gram-positive organisms, including methicillin-resistant Staphylococcus aureus, have become prevalent nosocomial pathogens in the United States. Vancomycin hydrochloride has been the most effective and reliable drug in treating methicillin-resistant Staphylococcus aureus. Consequently, vancomycin usage has increased considerably.1,2 Widespread use of vancomycin has also led to increased reports of adverse events. Among these, unfavorable skin reactions are the most prevalent reason for discontinuation of the medication.2 Vancomycin-induced Stevens-Johnson syndrome (SJS) and linear IgA bullous dermatosis (LABD) are adverse cutaneous reactions that can be difficult to distinguish clinically. It is important to make a correct diagnosis, though, because the prognosis and therapy are dissimilar.3,4 The objective of this study is to review SJS and LABD and discuss important differentiating characteristics so appropriate treatment is rendered, unnecessary use of corticosteroids is prevented, and appropriate prognosis is discussed with the patient. Methods We researched MEDLINE 1966 2002 for all published studies on vancomycin adverse events. We searched LABD, SJS, exanthema, and skin rashes each separately with vancomycin, specifically surveying the literature for articles pertinent to the clinical presentation, diagnosis, treatment, and prognosis of vancomycin-induced LABD and SJS. In addition, we searched on PubMed and OVID for all randomized controlled clinical trials for the treatment of SJS and LABD. We included all publications in English and non-english publications with English abstracts. Results SJS and LABD Clinical Presentation Toxic epidermal necrolysis, SJS, and erythema multiforme are JAOA Vol 104 No 4 April 2004 157

Figure 1. Stevens- Johnson syndrome involvement of oral mucosal membranes. Figure 2. Stevens- Johnson syndrome targets lesions on the back. three disorders that have similar features (Figures 1 and 2). However, researchers are still divided as to whether these disorders exist along a single spectrum or whether the spectrum includes only SJS and toxic epidermal necrolysis, with erythema multiforme being a separate entity.5 There have been several classification systems proposed for defining these disorders. Tables 1 and 2 summarizes these proposed methods of categorization.5,6 In addition to the above classification systems, biopsy specimens can be examined to aid in diagnosis. A frozen section biopsy of denuded epidermis will show primarily infiltrates of macrophages and activated lymphocytes, mainly CD8 cells (Figure 3). Keratinocyte necrosis is also typical of SJS and toxic epidermal necrolysis.7,8 Results of immunofluorescence studies are negative. Linear IgA bullous dermatosis is a subepidermal bullous disease that is defined by the presence of homogeneous linear deposits of IgA at the dermal-epidermal basement membrane.9 The clinical presentation of LABD is presented in Table 3 and Figures 4 and 5.10-13 Even with a detailed history and physical examination, the diagnosis of LABD is challenging because of its heterogenous clinical presentations. Skin biopsy is essential for a definitive diagnosis.9 Histologically, LABD is characterized by subepidermal bullae with a predominantly neutrophilic infiltrate (Figure 6) and basal cell vacuolization. Linear IgA bullous dermatosis can be distinguished from erythema multiforme, SJS, and toxic epidermal necrolysis by the absence of lymphocytic accumulations at the dermal-epidermal interface, subepidermal cleft formation, and keratinocyte necrosis.9 Direct immunofluorescence of perilesional skin is the standard for diagnosis of LABD. Direct immunofluorescence demonstrates a unique linear deposition of IgA along the basement membrane zone at the dermal-epidermal junction (Figure 7).7,9,10,13 Direct immunofluorescence results are negative in SJS. Figure 3. Section of biopsy specimen of a lesion from a patient with Stevens- Johnson syndrome shows the predominantly lymphocytic infiltrate (hematoxylin-eosin stain). SJS and LABD Treatment There is a lack of consensus on the treatment of patients with SJS. On a review of the literature, only two randomized studies were cited.14,15 The first was a prospective study of 16 children with erythema multiforme major (diagnosed based on criteria proposed by Huff et al6). Ten patients received bolus infusions of methylprednisolone (4 mg/kg/d) for 3 to 7 days, while six patients received only supportive treatment. The early use of corticosteroids compared to supportive treatment resulted in significant reduction of fever, acute eruption, and prostration.14 158 JAOA Vol 104 No 4 April 2004

Table 1 Preliminary Diagnostic Criteria for Erythema Multiforme Parameter Erythema Multiforme Minor Erythema Multiforme Major (SJS) Course Acute, self-limited, Acute, self-limited, or episodic or episodic Duration, wk 1-4 1-6 Skin lesions Symmetrically distributed, Symmetrically distributed, fixed fixed (lasting 7 days), discrete, (lasting 7 days), discrete, round round erythematous skin lesions erythematous skin lesions Evolution of skin lesions Concentric color change in Concentric color change in at least some lesions at least some lesions Mucous involvement Absent or limited to one surface Severe, involving at least two (usually mouth) surfaces Histopathology Primarily mononuclear cell infiltrate, Primarily mononuclear cell infiltrate, no leukocytoclastic vasculitis no leukocytoclastic vasculitis SJS indicates Stevens-Johnson syndrome. Table 2 Proposed Classification of Cases in the Spectrum of Severe Bullous Erythema Multiforme (from Bastuji-Garin5) Overlap TEN with TEN without Classification Bullous EM SJS SJS-TEN spots spots Detachment, % 10 10 10-30 30 10 Typical targets Yes NA NA NA NA Atypical targets Raised Flat Flat Flat NA Spots NA Yes Yes Yes NA EM indicates erythema multiforme; SJS, Stevens-Johnson syndrome; TEN, toxic epidermal necrolysis; NA, not applicable. The second study was a prospective analysis of 67 consecutive patients with SJS who were promptly treated with corticosteroids and had a favorable outcome.15 In contrast, retrospective studies demonstrated increased morbidity (higher rates of infections, gastrointestinal bleeding, longer hospital stay, and overall complications) and mortality in corticosteroidtreated patients with erythema multiforme major than in those treated symptomatically.15,16 Nonetheless, controlled studies suggest that early treatment with systemic corticosteroids is beneficial. Intravenous immune globulin therapy,17,18 ulinastatin (urinary trypsin inhibitor),19 plasmapheresis, cyclosporine, and cyclophosphamide may be beneficial for SJS.20-22 But, the data are insufficient and limited to case reports. The treatment for patients with drug-induced LABD dif- fers from treatment for patients with idiopathic LABD. Patients with drug-induced LABD have spontaneous resolution of skin lesions within 1 to 3 weeks after removal of the offending agent. Usually no other treatment is necessary. New lesions cease to appear 1 to 3 days after discontinuation of vancomycin.7,9,11,13 Corticosteroids and dapsone do not appear to influence the prognosis. Sequential biopsy specimens demonstrate resolution of the deposition of IgA without therapy.9 In contrast, it is recommended that patients with idiopathic LABD receive dapsone or sulfapyridine. With an inadequate response to dapsone, systemic corticosteroids are suggested.9,13 In refractory cases, interferon-alpha,2 colchicine,23 or co-trimoxazole may be tried.24 Therapy for idiopathic LABD is indicated, as only 10% to 50% of these cases have a spontaneous remission.9 JAOA Vol 104 No 4 April 2004 159

Table 3 Summary of Clinical Manifestations and Biopsy Results of Linear IgA Bullous Dermatosis and Stevens-Johnson Syndrome Clinical Linear IgA Stevens-Johnson Manifestation Bullous Dermatosis Syndrome References Rash Heterogeneous Confluent 5-8, 10-13, grouped papules, erythema, widely 26 vesicles, or bullae; disseminated can also be purpuric macules urticarial and papules, typical or atypical target lesions Mucous membrane 45% of the time Yes mostly 5, 8, 11 involvement in drug-induced on the oral LABD mucosa, lips, and conjunctiva Pruritus Yes Mainly nonpruritic 26 Bullae Can have EM-like May have bullae 11 lesions, flaccid or with epidermal tense bullae detachment in more severe cases Nikolsky s sign Negative, though Positive 8, 27 may be positive when pustules are confluent Pathology Subepidermal Lymphocytic 2, 7-9, blistering with accumulations 26, 27 predominantly at the dermoneutrophilic and epidermal eosinophilic dermal junction; will see infiltrate; can also keratinocyte see basal cell necrosis vacuolization Direct immunofluorescence Linear IgA Usually negative 2, 5, 7, deposition along 8, 11, 13, the BMZ, IgG 26, 27 and C3 IgA indicates immunoglobulin A; LABD, linear IgA bullous dermatosis; EM; erythema multiforme; BMZ, basement membrane zone. SJS and LABD Prognosis Sepsis, generally caused by Staphylococcus aureus and Pseudomonas aeruginosa, is the principal cause of death among patients with SJS. Other important causes of death include gastrointestinal bleeding and pulmonary embolism.25 The estimated mortality rate is less than 5% for SJS and approximately 30% for toxic epidermal necrolysis.8 The patient s age, percentage of skin detachment, serum urea nitrogen level, and level of neutropenia are the most important variables in determining prognosis.25 The prognosis for patients with vancomycin-induced LABD is excellent. Once vancomycin is discontinued, the patient should be expected to have prompt improvement without residual skin lesions.9,12,23 The skin lesions usually 160 JAOA Vol 104 No 4 April 2004

Figure 4. Linear IgA bullous dermatosis bullous lesions on the hands. Figure 6. Section of biopsy specimen shows the predominantly neutrophilic infiltrate in a patient with linear IgA bullous dermatosis (hematoxylin-eosin stain). Figure 5. Linear IgA bullous dermatosis bullous lesions on the face. Figure 7. Direct Immunofluorescence of perilesional skin showing linear IgA deposits at the basement membrane zone. This is the defining characteristic in a patient with linear IgA bullous dermatosis. resolve in 1 to 3 weeks, and no new lesions appear 1 to 3 days after the drug is stopped.7,9 If patients are rechallenged with vancomycin after having vancomycin-induced LABD, they may have a more severe recurrence, including a shorter latency and a longer course.7 Table 4 summarizes the treatment and prognosis for SJS and LABD. Discussion In cases that are difficult to distinguish based on clinical manifestations, a perilesional skin biopsy and direct immunofluorescence are indicated to make an early and correct diagnosis. Histologically, SJS shows lymphocytic accumulations at the dermoepidermal junction, along with subepidermal cleft forma- tion and keratinocyte necrosis. These features are absent in LABD. The defining test is direct immunofluorescence that demonstrates linear IgA deposition along the basement membrane zone in LABD, and this is absent in SJS. By performing an early biopsy, the correct diagnosis can be made. If SJS is diagnosed early, it seems that a brief course of systemic corticosteroids improves the outcome of the patients by reducing mortality, fever, and the period of an acute eruption. If the diagnosis is delayed and steroids are initiated late in the course of SJS, the prognosis may be worsened. Thus, a rapid diagnosis and therapeutic intervention are necessary in SJS.15,17 In contrast, for vancomycin-induced LABD, discontinuation of vancomycin is the only treatment necessary. The JAOA Vol 104 No 4 April 2004 161

Table 4 Summary of the Treatment and Prognosis of Linear IgA Bullous Dermatosis and Stevens-Johnson Syndrome Variable Linear IgA Bullous Dermatosis Stevens-Johnson Syndrome References Treatment Discontinue vancomycin Discontinue vancomycin 9, 14, 15, promptly. In idiopathic LABD, promptly. If diagnosed 23, 24, dapsone and sulfapyridine early, short course of have shown some benefit. corticosteroids has shown benefit. If given too late, or for too long, corticosteroids can lead to sepsis and other complications. Prognosis Prompt improvement and Depends on person s age, 7-9, 11, 25, lesions resolve spontaneously percent detachment of 26 without scar after skin, neutropenia, and discontinuation of serum urea nitrogen. vancomycin. Little associated Sepsis is the leading cause morbidity or mortality. of mortality. Overall mortality 5% in SJS and approximately 30% in TEN. IgA indicates immunoglobulin A; LABD, linear IgA bullous dermatosis; SJS, Stevens-Johnson Syndrome; TEN, toxic epidermal necrolysis. needless addition of corticosteroids increases the risk of adverse events such as psychosis, femoral head necrosis, hyperglycemia, and hypertension. In either case, treatment of both diseases requires supportive therapy that includes pain control, fluid replacement, avoidance of any adhesive material, fluid and electrolyte management, nutritional support, and protection against infection and sepsis.8 Empiric, prophylactic antibiotics are not generally recommended.25 Prognosis is guarded in SJS and depends on the severity of the disorder and how prompt appropriate treatment is given. Conversely, the prognosis for LABD is excellent as long as the vancomycin is discontinued and not reintroduced. With the varied prognosis, early definitive diagnosis is essential. In summary, early biopsy with immunofluorescence is necessary to ensure appropriate therapy is initiated and that appropriate prognosis is discussed with the patient when LABD and SJS are both potential diagnoses. References 1. VISA/VRSA Vancomycin Intermediate/Resistant Staphylococcus aureus fact sheet. Available at: http://www.cdc.gov/ncidod/hip/aresist/visa.htm. Accessed March 15, 2004. 2. Korman TM, Turnidge JD, Grayson ML. Risk factors for adverse cutaneous reactions associated with intravenous vancomycin. J Antimicrob Chemother. 1997;39:315-318. 3. Schneck B, Termeer C, Mockenhaupt M, Augustin M, Schopt E. Linear IgA dermatosis in an adult with clinical signs of Stevens-Johnson syndrome. Hautarzt. 1999;50:288-291. 4. Hughes AP, Callen JP. Drug-induced linear IgA bullous dermatosis mimicking toxic epidermal necrolysis. Dermatology. 2001;202:138-139. 5. Bastuji-Garin S, Rzany B, Stern RS, Shear NH, Naldi L, Roujeau JC. Clinical classification of toxic epidermal necrolysis, Stevens-Johnson syndrome, and erythema multiforme. Arch Dermatol. 1993;129:92-96. 6. Huff JC, Weston WL, Tonnesen MG. Erythema multiforme: A critical review of characteristics, diagnostic criteria, and causes. J Am Acad Dermatol. 1983;8: 763-775. 7. Nousari HC, Kimyai-Asadi A, Caeiro JP, Anhalt GJ. Clinical, demographic, and immunohistologic features of vancomycin-induced linear IgA bullous disease of the skin: Report of 2 cases and review of the literature. Medicine. 1999;78:1-8. 8. Roujeau JC, Stern RS. Severe adverse cutaneous reactions to drugs. N Engl J Med. 1994;331:1272-1285. 9. Bogdan I, Neughebauer BI, Negron G, Pelton S, Plunkett RW, Beutner EH, et al. Bullous skin disease: An unusual allergic reaction to vancomycin. Am J Med Sci. 2002;323:273-278. 10. Kuechle MK, Stegemeir E, Maynard B, Gibson LE. Leiferman KM. Peters MS. Drug-induced linear IgA bullous dermatosis: Report of six cases and review of the literature. J Am Acad Dermatol. 1994;30:187-192. 11. Palmer RA, Ogg G, Allen J, Banerjee A, Ryatt KS, Ratnavel R, et al. Vancomycin-induced linear IgA disease with autoantibodies to BP 180 and LAD 285. Br J Dermatol. 2001;145:816-820. 162 JAOA Vol 104 No 4 April 2004

12. Whitworth JM, Thomas I, Peltz SA, Sullivan BC, Wolf AH, Cytryn AS. Vancomycin-induced linear IgA bullous dermatosis (LABD). J Am Acad Dermatol. 1996;34:890-891. 13. Klein PA, Callen JP. Drug-induced linear IgA bullous dermatosis after vancomycin discontinuance in a patient with renal insufficiency. J Am Acad Dermatol. 2000;42:316-323. 14. Kakourou T, Klontza D, Soteropoulou F, Kattamis C. Corticosteroid treatment of erythema multiforme major (Stevens-Johnson syndrome) in children. Eur J Pediatr. 1997;156:90-93. 15. Tripathi A, Ditto AM, Grammer LC, Greenberger PA, Mc Grath KG, Zeiss CR, et al. Corticosteroid therapy in an additional 13 cases of Stevens-Johnson syndrome: a total series of 67 cases. Allergy Asthma Proc. 2000;21:101-105. 16. Assier H, Bastuji-Garin S, Revus J, Roujeau JC. Erythema multiforme with mucous membrane involvement and Stevens-Johnson syndrome are clinically different disorders with distinct causes. Arch Dermatol. 1995;131:539-543. 17. Brett AS, Philips D, Lynn AW. Intravenous immunoglobulin therapy for Stevens-Johnson syndrome. South Med J. 2001;94:342. 18. Straussberg R, Harel L, Ben-Amitai D, Cohen D. Amir J. Carbamazepineinduced Stevens-Johnson syndrome treated with IV steroids and IVIG. Pediatr Neurol. 2000;22:231-233. 19. Inamo Y, Okubo T, Wada M, Fuchigami S, Hashimoto K, Fuchigami T, et al. Intravenous ulinastatin therapy for Stevens-Johnson syndrome and toxic epidermal necrolysis in pediatric patients. Int Arch Allergy Immunol. 2002;127: 89-94. 20. Kamanabroo D, Schmitz-Landgraf W, Czarnetzki BM. Plasmaperesis in severe drug-induced toxic epidermal necrolysis. Arch Dermatol. 1985;121:1548-1549. 21. Renfro L, Grant-Kels JM, Daman LA. Drug-induced toxic epidermal necrolysis treated with cyclosporine. Int J Dermatol. 1989; 28:441-444. 22. Heng MC, Allen SG. Efficacy of cyclophosphamide in toxic epidermal necrolysis: Clinical and pathophysiologic aspects. J Acad Dermatol. 1991;25:778-786. 23. Ang P, Tay YK. Treatment of linear IgA bullous dermatosis of childhood with colchicines. Pediatr Dermatol 1999;16:50-52. 24. Pulimood S, Ajithkumar K, Jacob M, George S, Chandi SM. Linear IgA bullous dermatosis of childhood: Treatment with dapsone and co-trimoxazole. Clin Exp Dermatol. 1997;22:90-91. 25. Revuz J, Penso D, Roujeau JC, Guillaume JC, Payne CR, Wechsler J, et al. Toxic epidermal necrolysis: Clinical findings and prognosis factors in 87 patients. Arch Dermatol. 1987;123:1160. 26. Bickle KM, Roark TR, Hsu S. Autoimmune bullous dermatosis: A review. Am Fam Physician. 2002;65:1861-1869. 27. Yeung CK. Toxic epidermal necrolysis and Stevens-Johnson syndrome. Hong Kong Dermatology and Venereology Bulletin. 2002;10:57-66. JAOA Vol 104 No 4 April 2004 163