Objective: Investigation of a possible involvement of differences in human leukocyte antigens (HLA) in the risk of nevirapine (NVP)-induced skin rash among HIV-infected patients.
Methods: A step-wise case-control association study was conducted. The first set of samples consisted of 80 samples from patients with NVP-induced skin rash and 80 samples from NVP-tolerant patients. These patients were genotyped for the HLA-A, HLA-B, HLA-C, HLA-DRB1, HLA-DQB1, and HLA-DPB1 by a sequence-based HLA typing method. Subsequently, we verified HLA alleles that showed a possible association in the first screening using an additional set of samples consisting of 67 cases with NVP-induced skin rash and 105 controls.
Results: An HLA-B*3505 allele revealed a significant association with NVP-induced skin rash in the first and second screenings. In the combined data set, the HLA-B*3505 allele was observed in 17.5% of the patients with NVP-induced skin rash compared with only 1.1% observed in NVP-tolerant patients [odds ratio (OR)=18.96; 95% confidence interval (CI)=4.87-73.44, Pc=4.6x10] and 0.7% in general Thai population (OR=29.87; 95% CI=5.04-175.86, Pc=2.6x10). The logistic regression analysis also indicated HLA-B*3505 to be significantly associated with skin rash with OR of 49.15 (95% CI=6.45-374.41, P=0.00017).
Conclusion: A strong association between the HLA-B*3505 and NVP-induced skin rash provides a novel insight into the pathogenesis of drug-induced rash in the HIV-infected population. On account of its high specificity (98.9%) in identifying NVP-induced rash, it is possible to utilize the HLA-B*3505 as a marker to avoid a subset of NVP-induced rash, at least in Thai population.