Pruritus and Leucocytoclastic Vasculitis due to Azathioprine

Abstract

Azathioprine is known for a wide array of cutaneous adverse events such as contact dermatitis and exanthems [1]. These type IV allergic reactions occur fairly frequently and can be proven with patch testing [2]. Vasculitis, however, is rare and has only been described in 7 patients [3–5], whereas pruritus due to azathioprine has not been reported before. A 55-year-old, non-atopic woman, with a history of Crohn’s disease and rheumatic complaints, was admitted with a 1-year history of pruritus localised on both arms. Anamnestically, these complaints had started when her daily dose of azathioprine, which she had been using for 2 years, had been doubled. Concurrent medication was mesalazine, cholestyramine, acetaminophen/codeine and vitamin B12. Clinical examination revealed only a few pinpoint excoriations on the extensor surfaces of both upper arms, but no other lesions like papules or vesicles. Blood screening showed no abnormalities, and a skin biopsy of an excoriation revealed no pathological changes. Our initial diagnosis was pruritus sine materia, possibly related to azathioprine. As azathioprine could not be stopped, she was treated with hydroxyzine 10 mg and 0.1% triamcinolone cream once daily. During the following months, pruritus responded well to this therapy, although exacerbations occurred during stressful events (e.g. deaths in the family). During a few months, she did not consult us but then returned with the same complaints of pruritus on both arms, this time not related to stress. Apart from some excoriations, a single, pinpoint, erythematous papule was seen on the right upper arm. Histologically, this papule showed a leucocytoclastic vasculitis with some eosinophils, consistent with either a drug reaction (vasculitis allergica) or urticarial vasculitis. Anamnestically, azathioprine was the most likely cause, although mesalazine and acetaminophen have also been reported to cause vasculitis [1]. As there is no standard method to investigate a type III allergic reaction such as vasculitis, patch tests and intradermal tests can be carried out to detect a possible concurrent type IV or type I allergy to the putative culprit drug. Thus, patch tests were performed with the European standard series and the patient’s own medication, but no reactions were seen on days 2 and 3. Because mesalazine is water insoluble, scratch tests were carried out with azathioprine and mesalazine (both 1% in saline), neither of which induced any reaction. Next, intradermal tests were performed with a dilution series of azathioprine (0.01, 0.1 and 1% in normal saline) to determine the threshold for the allergic reaction [6] and normal saline as negative control. This induced only a dosedependent wheal-and-flare reaction to 0.1 and 1% azathioprine after 20 min, which was discounted as an irritant reaction, which azathioprine is known to produce. After 24 h however, these two injection sites showed a non-blanchable erythema with some induration, with the most pronounced reaction at the 1% injection site. A skin biopsy from this site showed a leucocytoclastic vasculitis, similar to the histology from the pruritic papule. This reaction is unlikely to be a toxic one, as these usually do not persist after 24 h [7] and are usually not of vasculitic nature. In addition, intradermal tests in a healthy control were negative: a wheal-and-flare reaction was present at the 1% injection site after 20 min, gradually disappeared, and after 24 h no skin reaction was visible or palpable. Taken together, these observations suggest vasculitis allergica with localised pruritus due to azathioprine. Unfortunately, the gastro-enterologist does not have a suitable alternative, but the azathioprine dose is kept as low as possible. Our patient uses hydroxyzine and fluticasone cream when necessary. In contrast to our case, all previously reported patients presented with an acute, erythematous rash or purpuric lesions on their legs. Although the diagnosis was confirmed by histology, unfortunately no patch or intradermal tests were performed. Intradermal tests have only been carried out in 1 patient with azathioprine hypersensitivity (urticarial vasculitis) [8]. Unfortunately, as no details of the intradermal tests were provided, comparison with our method was not possible. In conclusion, we report a rare case of localised pruritus most probably due to azathioprine-induced vasculitis allergica, which was supported by histology and intradermal tests. Similar to a patient without therapeutic alternatives described by Schmitt et al. [8], it may be worth trying 6-mercaptopurine, an active metabolite of azathioprine, to prevent a relapse.