We read with interest the recent article by Culver et al (1) in which the authors described a rare case of minocycline-induced cutaneous polyarteritis nodosa (PAN), a clinical entity that overlaps with cutaneous granulomatous vasculitides (2). The authors observed fibrinoid necrosis within the walls of medium- or small-sized dermal vessels surrounded by dense granulomatous perivascular infiltrates, a finding that has been reported in classic or cutaneous PAN (3, 4). Based on a review of 6 previously published cases, Culver et al proposed diagnostic guidelines for minocycline-induced cutaneous PAN. Patients should fulfill 6 for the following 7 diagnostic criteria: 1) minocycline use for >12 months, 2) skin manifestations including livedo reticularis and/or subcutaneous nodules, 3) arthritis and/or myalgias and/or neuropathy in the distribution of the rash, 4) lack of systemic organ involvement, 5) skin biopsy sample showing necrotizing vasculitis of small- and/or medium-sized vessels, 6) evidence of perinuclear antineutrophilic cytoplasmic antibodies (pANCA), and 7) improvement after discontinuation of minocycline (1).
The article by Culver and colleagues prompted us to report a case that fulfills their criteria and confirms that minocycline may induce a variant of cutaneous PAN. Our patient, a 40-year-old woman, was admitted for a nodular rash on the lower extremities and associated ankle pain. Minocycline had previously been prescribed for recurrent acne, and she was currently receiving a second cycle of minocycline (100 mg/day for the past 6 months). She also reported taking oral contraceptives (estroprogesterone), but denied taking any additional prescriptions.
Physical examination revealed slightly tender, violaceous, subcutaneous nodules extending from the legs to the thighs. Ankle stiffness and swelling without limitation of joint movement were also noted. Results of the physical examination were otherwise unremarkable, and a chest radiograph was normal.
The laboratory findings included a C-reactive protein level of 1.7 mg/dl, positive pANCA, negative enzyme-linked immunosorbent assay to myeloperoxydase and proteinase-3, and creatinine levels and hepatic tests that were within normal limits. No immunologic or serologic findings known to be associated with cutaneous vasculitis were seen. Results of a skin biopsy showed a granulomatous necrotizing vasculitis involving medium-sized dermal arteria (Figure 1). The minocycline was discontinued, and the skin lesions progressively resolved without the use of corticosteroids. The patient continued oral contraceptive therapy.
This eighth reported case of minocycline-induced PAN closely resembles the seventh case reported by Culver et al. Our case occurred in a woman taking an estroprogestative while taking a long-term course of minocycline for acne. The patient had positive pANCA that lacked specificity for myeloperoxydase and proteinase-3. The skin lesions disappeared after minocycline was discontinued, but the oral contraceptives were continued.
Our patient fulfilled all of the clinical and immunologic criteria proposed by Culver et al for minocycline-induced cutaneous PAN and had a biopsy-proven vasculitis with a perivascular granulomatous inflammation. Our findings confirm the value of combining the 6 nonhistologic criteria proposed by Culver et al in the diagnosis of minocycline-induced PAN.