August 16, 2006
Although abacavir (ABC, Ziagen) is a highly effective drug for the treatment of HIV, a drawback to its use is the risk of hypersensitivity reaction (HSR). To date, 8% of patients in clinical studies have experienced HSR, and some cases have proven fatal.1 The implementation of a clinical risk management program has successfully reduced the incidence of serious outcomes among patients on abacavir, but a highly predictive genetic marker could significantly improve the risk-benefit ratio even more. Multiple studies from Simon Mallal's group in Australia have shown that the human leukocyte antigen (HLA) allele HLA-B*5701 is highly associated with abacavir HSR in whites.2,3 Complementing Mallal's work, Michael Mosteller and colleagues from GlaxoSmithKline (Glaxo) presented a fascinating study4 on the pharmacogenetics behind abacavir hypersensitivity that aimed to answer several key questions:
A total of 595 cases were retrospectively identified among individuals who were suspected of having abacavir HSR. These individuals were compared with 744 abacavir-treated control patients who had no evidence of abacavir HSR. Of the 1,339 individuals, 73% were white, 9% black, 10% Hispanic and 8% Asian (primarily Thai). Their mean age was 40.9 years, and 24% were female.
The predictive value of single and paired genetic markers was examined for both cases and controls. Racial groups were analyzed separately. Although several markers located throughout the genome were significantly associated with suspected abacavir HSR, the strongest associations involved markers located on chromosome 6 in the region surrounding HLA-B. In white individuals, HLA-B*5701 was most strongly associated with clinically suspected abacavir HSR, as shown in the table below, and no combination of markers exhibited sensitivities and specificities greater than those of HLA-B*5701 alone. The sensitivity and positive predictive value of HLA-B*5701 varied in non-whites, being highest in Asian individuals and lowest in blacks.
|Performance Characteristics of HLA-B*5701, by Racial Group|
|Racial Group||Whites||Blacks||Hispanics||Asians (Thai)|
|Genotypic P Value||7.5x10-73||0.16||1.2x10-5||6.3x10-6|
|OR (95% CI)||42.1 (22.8-77.8)||4.3 (0.7-28.3)||41.3 (2.4-708.7)||263.6 (11.8-5909.1)|
CI, confidence interval; NPV, negative predictive value; OR, odds ratio; PPV, positive predictive value.
Although the results in non-whites must be considered exploratory given the small sample sizes of the analyzed groups, HLA-B*5701 may be a useful marker for predicting abacavir HSR in Latinos and Thais -- more so than in blacks. Studies involving larger populations of non-white individuals are needed to verify these preliminary findings. Studies assessing different white subpopulations are also required to determine whether the predictive value of HLA-B*5701 is equally as strong across the entire racial group.
Clinical diagnosis of abacavir HSR is required for optimal management of the condition, but is quite difficult. HLA testing is a relatively inexpensive test that may improve the diagnostic accuracy of predicting abacavir HSR, especially when combined with skin patch testing. Two Glaxo studies incorporating skin patch testing have begun. PREDICT-1 is a prospective trial comparing the abacavir HSR rate between two groups of patients receiving standard abacavir treatment: a diverse group of patients and a group of patients who do not carry the HLA-B*5701 allele. The SHAPE study is a retrospective trial evaluating the performance characteristics of HLA-B*5701 in whites and blacks.
Abacavir has proven to be a potent and useful nucleoside analog for the treatment of HIV. Hopefully clinicians will soon be able to take better advantage of this effective agent by using HLA typing and/or skin testing to tailor abacavir use to the patients least likely to develop HSR.