Abacavir-induced reaction with fever and severe skin rash in a patient tested human leukocyte antigen-B*5701 negative

INTRODUCTION

The most frequent adverse events associated with abacavir are gastrointestinal and systemic symptoms, which usually appear in the first weeks of treatment and tend to resolve spontaneously and be of mild-to-moderate severity. ^1,2

The most serious adverse event caused by abacavir is the hypersensitivity reaction, a severe allergic reaction that occurs in 5–8% of persons who begin abacavir therapy, usually within the first six weeks of treatment. Clinical manifestations include fever, rash, gastrointestinal or respiratory symptoms, and may lead to life-threatening hypotension, renal failure and death in case of re-challenge with abacavir after the occurrence of hypersensitivity. Historically, the hypersensitivity reaction has been the most common reason for the short-term discontinuation of abacavir in patients receiving this agent as part of antiretroviral treatment. ^3,4

In recent years, a statistically significant correlation between the abacavir hypersensitivity reaction and the presence of the human leukocyte antigen (HLA) subtype B*5701 has been proved. Retrospective analyses showed that patients with this haplotype had a high risk (approximately 70%) of developing the allergic reaction, whereas those without this allele had a very low risk (lower than 1%) of presenting this adverse event. ^5,6 Moreover, two recent cohort studies have demonstrated that hypersensitivity to abacavir is strongly associated with the presence of the HLA-B*5701 allele, and routine screening for this allele has significantly reduced the incidence of abacavir-associated allergic reaction in both black and white patients. ^7,8 A recently published prospective randomized trial has shown that the absence of HLA-B*5701 allele had a negative predictive value of 100% for skin patch test confirmed hypersensitivity reactions, and of 96% for clinically suspected allergic reactions. As a result, a routine genotyping test for the HLA-B*5701 allele is now recommended before the start of abacavir therapy, and abacavir should be avoided in all subjects who have tested positive for this haplotype. ⁹

CASE REPORT

A 41-year-old Caucasian female patient with asymptomatic HIV-1 infection and taking antiretroviral therapy for nine years switched from an antiretroviral therapy including tenofovir/emtricitabine (300/200 mg daily) and efavirenz (600 mg daily) to zidovudine/lamivudine/abacavir (300/150/300 mg twice daily) owing to the persistence of insomnia, dizziness and anxiety. The patient had no history of known drug hypersensitivity, did not abuse alcohol or illicit drugs, and did not take dietary, herbal or natural product supplements. Sequence-based HLA-B genotyping confirmed that she did not carry the HLA-B*5701 allele. Immuno-virological tests at the time of switching showed a total lymphocyte count of 2530 cells/mm³, a CD4 lymphocyte count of 1139 cells/mm³ (45%) and an HIV plasma viral load lower than 50 copies/mL.

After 10 weeks, she suddenly developed fever (with a temperature of 38°C) accompanied by chills, weakness and reduced appetite. Two days later, fever persisted and the patient developed a generalized rash on the face, trunk and arms. Skin eruption was characterized by multiple maculopapular and orticarioid lesions, and was associated with severe enanthema including painful mucosal ulcerations of the oropharynx. Blood white cell count, red cell count and haemoglobin level tested normal by liver and kidney function tests. Chest X-ray, echocardiographic study, blood cultures, urine cultures and serological tests (including cryptococcal antigen, Widal-Wright reaction, and search of IgM for Epstein–Barr virus, cytomegalovirus, measles virus, rubella virus, toxoplasma, leishmania and borrelia) proved negative. Abdominal ultrasonography disclosed no significant abnormalities.

Abacavir was immediately discontinued and a new antiretroviral treatment including zidovudine/lamivudine (300/150 mg twice daily) and tenofovir (300 mg daily) was started, in association with oral antihistamine (loratadine, 10 mg daily) and corticosteroid (prednisone, 10 mg daily) therapy for five days. Within two days, the temperature became normal, whereas the skin rash evolved into a desquamative phase and disappeared completely 10 days later. The patient refused to take the abacavir skin patch test, but the association of abacavir with this adverse event is probable on the basis of the Naranjo Causality Scale. ¹⁰

Few severe abacavir-associated adverse events occurring in HLA-B*5701-negative patients have been reported previously. ^7,8,11 Recently, Bonta et al. ¹² described a case of severe hypersensitivity reaction with skin rash and severe rhabdomyolysis in a 61-year-old man, who tested HLA-B*5701 negative. In this case, fever above 40°C, muscle aches, watery diarrhoea and skin rash appeared 10 days after the replacement of zidovudine with abacavir, and were associated with a severe increase in creatinine phosphate kinase levels and myoglobinuria. The hypersensitivity reaction slowly resolved after abacavir discontinuation. Sequence-based HLA-B typing confirmed that he did not carry the HLA-B*5701 allele; the abacavir skin patch test was also negative.

Fox et al. ¹³ observed a 45-year-old, HLA-B*5701-negative subject with AIDS and pulmonary tuberculosis, who developed fever exceeding 39°C, nausea and abdominal pain within five hours of receiving the first dose of abacavir. The gastrointestinal symptoms subsequently resolved but the fever persisted during the next 13 days. Laboratory and radiological investigations tested negative for infectious diseases or malignancies, and the temperature normalized within hours after the interruption of abacavir. The clinical and laboratory parameters of these patients described previously are compared with those of our case in Table 1.

In our case, although the patient tested negative for the HLA-B*5701 allele, severe side-effects, including fever, generalized skin rash and ulcerative mucosal lesion of the oropharynx, were observed 10 weeks after the start of new antiretroviral therapy. Even though the patient refused the skin patch test, the diagnosis of abacavir-associated allergic reaction was very probable because of its clinical presentation and quick resolution after abacavir discontinuation, as confirmed by the Naranjo estimation. ¹⁰

Finally, these cases show that not all severe abacavir-induced adverse events occur as a result of classic hypersensitivity reactions, and can present also in HLA-B*5701-negative patients. Although the HLA-B*5701 allele is strongly related to the hypersensitivity reaction, a second pathogenic mechanism is probably involved irrespective of HLA-B*5701 status. Even if a subject tests negative for this haplotype or for the abacavir patch test or both, severe drug reactions may still occur in patients receiving abacavir and a careful and continuous clinical vigilance is required.