Antineutrophil Cytoplasmic Antibody–Positive Cutaneous Leukocytoclastic Vasculitis Associated With Propylthiouracil Therapy FREE

Administration of propylthiouracil therapy has been associated with a hypersensitivity syndrome that typically manifests as vasculitis.¹ Most cases of propylthiouracil therapy–induced antineutrophil cytoplasmic antibody (ANCA) positivity react to perinuclear ANCA (p-ANCA).¹ We describe a patient who presented with cutaneous manifestations of propylthiouracil therapy hypersensitivity vasculitis with ANCA positivity.

A 43-year-old woman with hyperthyroidism presented to our institution with generalized eruptions. The patient's hyperthyroidism had been controlled with propylthiouracil, 100 mg/d, for the previous 3 years. Cutaneous examination revealed many irregularly shaped, purpuric plaques with necrotic centers and surrounding rims of erythema on the face, ears, trunk, and extremities (Figure 1). The p-ANCA level was 901 U/mL (reference range, <7 U/mL). The cytoplasmic ANCA (c-ANCA) level was within the reference range (<4.4 U/mL) (Fluoro ANCA test; MBL, Nagoya, Japan). A skin biopsy sample demonstrated leukocytoclastic vasculitis (Figure 2A). The luminal walls of the vascular endothelial cells were positive for myeloperoxidase (Figure 2B). To further confirm that the myeloperoxidase on the luminal walls of cutaneous small vessels was the target antigen of the p-ANCA, we performed a direct immunofluorescence (DIF) study of the lesional skin with the c-ANCA– and p-ANCA–positive control serum samples of the Fluoro ANCA test at a dilution of 1:20. The results showed granular deposits of p-ANCA target antigen along the vessel wall (Figure 2C), and the results of DIF for c-ANCA target antigen were negative (Figure 2D). Propylthiouracil therapy was discontinued because of laboratory and histopathologic evidence of propylthiouracil therapy–induced ANCA-positive vasculitis. The serum p-ANCA titer then decreased to 9.2 U/mL, and the eruptions resolved in the following 2 weeks.

The target antigen of the most common p-ANCA pattern is specific to myeloperoxidase. It has been suggested that myeloperoxidase, which is involved in the formation of metabolites from propylthiouracil therapy, may bind covalently to one of these metabolites.¹ This can then form an immunogenic conjugate and behave as a hapten.¹ The changed enzymes could initiate production of antimyeloperoxidase autoantibodies. A study by Xiao et al² provided evidence that ANCAs are sufficient to cause systemic pauci-immune vasculitis in vivo and offers direct evidence that p-ANCA is pathogenic. Tumor necrosis factor α–primed neutrophils result in mitogen-activated protein kinase–dependent translocation of ANCA antigens to the cell surface.³ Stimulation of primed neutrophils by p-ANCA results in neutrophil activation, with reactive oxygen species generation, degranulation, and adhesion to and killing of endothelial cells.⁴ Also, degranulated myeloperoxidase can opsonize unprimed neutrophils, rendering them directly amenable to activation by p-ANCA.⁴ Endothelial cells can bind to and internalize this degranulated myeloperoxidase, resulting in an increase in intracellular oxidant radicals.⁵ Circulating p-ANCA can bind to degranulated myeloperoxidase, forming immune complexes that dimerize Fcγ receptors on neutrophils, resulting in activation.⁴ Therefore, the vessels are damaged. In our case, the lumen side of the endothelial cells of vasculitic vessels was stained positive for myeloperoxidase. This provided evidence that endothelial cells can bind with degranulated myeloperoxidase. A DIF study also confirmed the presence of p-ANCA target antigen. Our observation offers direct clinical evidence of the presence of myeloperoxidase on the cutaneous vascular endothelial surface as the target antigen in propylthiouracil therapy–induced p-ANCA–positive vasculitis.

Correspondence: Chia-Yu Chu, MD, Department of Dermatology, National Taiwan University Hospital, 7, Chung-Shan South Road, Taipei 100, Taiwan (chiayu.chu@gmail.com).

Financial Disclosure: None reported.

Accepted for Publication: January 30, 2006.

Author Contributions:Study concept and design: Sheen, Chu, and Yu. Acquisition of data: Sheen and Chu. Analysis and interpretation of data: Sheen and Chu. Drafting of the manuscript: Sheen and Chu. Critical revision of the manuscript for important intellectual content: Sheen, Chu, and Yu. Obtained funding: Sheen. Administrative, technical, and material support: Sheen. Study supervision: Sheen, Chu, and Yu.