Piroxicam-induced agranulocytosis.

TO THE EDITOR: The nonsteroidal antiinflammatory drugs are common causes of agranulocytosis. However, agranulocytosis has not been reported as a complication of piroxicam therapy.':" We report a case of piroxicam-induced agranulocytosis. Our patient was a 65-year-old woman, with a several-year history of arthrosis. Three months before admission, she presented with tonsillitis and was treated with penicillin and erythromycin for one month. After that, she remained asymptomatic, and drugs were discontinued. Tendays before admission, she suffered painful arthrosis, and piroxicam 20 rng/d was administered for three days. She was not taking any other drugs. Twodays before admission, she developed fever and a sore throat due to necrotic ulcerative lesions of oral mucous membranes. On admission, her white cel1 count was 800/mm' (2"70 neutrophils, 90% lymphocytes, 8% monocytes), hemoglobin was 13.4 g/dl., and platelet count was 255 OOO/mm'. Chest X-ray, electrocardiogram, urinalysis, serum electrolytes, creatinine, and liver enzymes were normal. Urine, blood, and throat cultures obtained on admission were negative. Bone marrow aspiration showed normal red cel1 and platelet production, but only myeloblasts and promyelocytes were found, in the absence of mature myeloidcel1s. Treatment was instituted with cefoxitin, amikacin, and acetaminophen. Within 12days the granulocyte count was 4800/mm' (50"70 neutrophils, 48% lymphocytes, 2% monocytes), and she wasdischarged. Over two months follow-up, her blood count remained stable and she had no newcomplaints or problems. The risks of agranulocytosis in relation to analgesic drug use were evaluated in the International Agranulocytosis and Aplastic Anemia Study' and revised by Laporte." However, piroxicam could not be evaluated because of the small number of cases treated with this drug. This case suggests that piroxicam can cause reversible myelosuppression similar to that seen with other nonnarcotic analgesics.