OBJECTIVES

To determine the frequencies of HLA-B alleles in Ugandan patients in the NORA substudy of the DART trial and to compare HLA-B allele frequencies in those with and without clinically diagnosed hypersensitivity reaction (HSR).

METHODS

DNA-based HLA-B genotyping was used to determine HLA alleles in 247 participants who received abacavir, including all six participants ('cases') with clinically diagnosed abacavir HSR.

RESULTS

The incidence of clinical abacavir HSR in this double-blinded study was 2.0% (6/300) in the abacavir group. As HLA-B5701 was absent throughout the entire cohort, including the six HSR 'cases', an association could not be established between HLA-B5701 and clinically diagnosed abacavir HSR. No other HLA-B57 alleles were present among the six 'cases'. HLA-B5703 was the most frequent HLA-B*57 allele among the abacavir-tolerant participants.

CONCLUSION

The rate of clinical HSR was low, which may reflect the expected 2-3% clinical false-positive rate seen in previous double-blind randomized studies. The presumption that these cases may be false-positive abacavir HSR is supported by the fact that no HLA-B5701 alleles were found in the abacavir group. Implementation of prospective HLA-B5701 screening must be based on benefit/risk considerations within local practice. Clinical risk management remains paramount.