Zidovudine-induced pure red cell aplasia presenting after 4 years of therapy

Abstract

Reversible pure red cell aplasia is a recognized complication of both zidovudine and lamivudine, typically occurring within the first 3 months of therapy [1,2]. We report the case of a 29-year-old man with HIV, severe haemophilia A and hepatitis C, who developed reversible pure red cell aplasia 4 years after commencing zidovudine. The patient presented to our centre in September 2004 with a 2-week history of exertional dyspnoea and palpitations. He had severe haemophilia A and chronic hepatitis C (genotype 3B), which had never been treated. HIV was diagnosed in 1984; highly active antiretroviral therapy (HAART) was commenced in February 1999 with zidovudine, lamivudine and indinavir. In 2002, his HAART regimen was changed to Trizivir (zidovudine, abacavir and lamivudine). As he had previously developed Stevens–Johnson syndrome on co-trimoxazole, monthly nebulized pentamidine was used as Pneumocystis carinii prophylaxis. His haemophilia A was managed with plasma-derived factor VIII on demand (Fandhi 3000 units), which he used two to three times a month. He was taking no other medications. On examination, he appeared pale but otherwise well. There was conjunctival pallor, but no jaundice. Respiratory and abdominal examination was normal. He had a hyperdynamic circulation with a grade 3/6 ejection systolic flow murmur confined to the left sternal edge. There were no symptoms or signs of bleeding. Initial laboratory investigations showed a severe macrocytic anaemia with a haemoglobin level of 3.5 g/dl and mean corpuscular volume of 115 fl. The reticulocyte count was low at 3 × 109/l. Platelet and white cell counts were normal. The blood film showed an oval macrocytosis. On review of previous blood counts, the patient had had a macrocytosis (range 117–122 fl) without anaemia since commencing HAART in 1999. At the time of presentation, his HIV-RNA level was less than 50 copies/ml and the CD4 cell count was 391 cells/μl. Serum ferritin, B12 and folate were normal, as were plasma methylmalonic acid and homocysteine (functional markers of B12 and folate deficiency, respectively). Renal and liver function tests were normal. The serum erythropoietin level was appropriately raised at 584 U/l. Bone marrow aspirate and trephine biopsy were performed with factor VIII cover. The aspirate showed a megaloblastic picture with maturation arrest of the erythroid series at the early erythroblast stage. The trephine biopsy was hypercellular and megaloblastic with disorganized haematopoeisis. Parvovirus IgG was positive but IgM was negative. Mycoplasma serology was negative. Bone marrow trephine stains for acid-fast bacilli, fungi, parvovirus, human herpes virus 8 and Epstein–Barr virus were negative. Polymerase chain reaction testing of bone marrow aspirate for parvovirus B19 and HHV-6 DNA was negative. Peripheral blood flow cytometry for paroxysmal nocturnal haemoglobinuria was negative. Four units of packed red cells were transfused, and the patient was monitored for 10 days. As he failed to mount a reticulocyte response and the virology findings proved negative, his HAART regimen was changed to lamivudine, tenofovir and efavirenz. Eight days after stopping zidovudine, he mounted a reticulocyte response of 120 × 109/l. The full blood count normalized 20 days after the cessation of zidovudine. Six months later, he remains well with a normal full blood count and normal mean corpuscular volume. Pure red cell aplasia is a well-documented adverse reaction of HAART regimens containing both zidovudine and lamivudine. In this instance, the discontinuation of zidovudine, but not of lamivudine, led to the prompt resolution of the anaemia. This is the first published case of zidovudine-induced pure red cell aplasia occurring so late in zidovudine therapy, and illustrates that this complication is not restricted to the first 3 months of treatment.