Abstract
Abacavir is one of the most widely used drugs in the treatment of HIV. However, 4% of patients taking abacavir suffer from a life-threatening hypersensitivity reaction. Multiple research groups have confirmed the association between hypersensitivity and a particular genotype of a component of the major human histocompatibility complex type I: the HLA-B*57:01 allele. Predictive tests are, therefore, performed on all HIV-positive patients prior to abacavir treatment, as recommended by the drug label. The procedure used at Hospital Universitario de La Princesa to determine the HLA-B*57:01 genotype consists of inverse hybridization prior to sequencing. This technique determines the HLA-B allele but lacks the resolution necessary to discriminate the HLA-B*57:01 subvariant. Thus, the rest of the HLA-B genotypes other than HLA-B*57:01 are discriminated, for which abacavir can be administered. In patients carrying the HLA-B*57 allele, a second determination is made by PCR and Sanger sequencing to confirm the existence of HLA-B*57:01. Therefore, the current procedure is technically complex, costly and involves a moderate response time; therefore, substitution by another method, such as genotyping using hybridization probes, will shorten response times and reduce costs. The objective of this study is to validate a method based on the analysis of a theoretically perfect linkage disequilibrium between HLA-B*57:01 and the G allele of the rs2395029 polymorphism in the HLA complex P5 gene (HCP5), to be used as a biomarker predictive for abacavir hypersensitivity. From 1225 patients genotyped for HLA-B since 2008, 49 patients positive and 177 negatives for HLA-B*57:01 were genotyped by quantitative PCR with allele-specific hybridization probes for polymorphism rs239529. Specificity and sensitivity values were 100%, which 95% confidence intervals were 93−100% and 98−100% respectively. Positive predictive value was estimated as 84.4%, which 95% confidence interval was 48.1−93.9%. The negative predictive value was estimated as 99.9%; likewise, when setting the confidence level at 95%, the confidence level was 99.4−100%. A quantitative study of HCP5 was also performed, since this gene is a copy number variation zone, so we studied whether 5 patients apparently homozygous for the allele G of HCP5 but heterozygous for HLA-B had the most frequent deletions described for this zone. The results showed that they do not have such deletions. Also, economic and practical comparison between both methods resulted in an improvement in time and cost if the HCP5 rs2395029 method is incorporated. In conclusion, HCP5 genotyping is a viable alternative method for predicting hypersensitivity to abacavir, being able to replace the conventional HLA-B*57:01 typing.
