Historically the hypersensitivity reaction (HSR) has been the most frequent reason for the short-term discontinuation of abacavir in HIV-infected patients receiving this agent as part of HAART. Up to 8% of abacavir-treated patients discontinue the agent for suspected HSR . After initial observations that white Caucasian patients were at an increased risk of abacavir HSR it has been proved that the strongest predictor of the reaction is the presence of the human leukocyte antigen (HLA) subtype B*5701 [2–4]. Retrospective analyses showed that individuals with this haplotype have a high risk (approximately 70%) of developing abacavir HSR, whereas those without this allele had a very low risk (less than 1%) of suspected HSR . Subsequently, two prospective cohorts, one in western Australia  and one in the United Kingdom , have demonstrated a statistically significant reduction in the incidence of abacavir HSR with the use of prospective HLA screening and the avoidance of abacavir in those testing HLA-B*5701 positive.
We aimed to study the use of prospective HLA-B*5701 testing and the impact of this test on the rate of abacavir HSR in our clinic cohort.
Routine HLA-B*5701 screening has been utilized at our centre since August 2005, with all treatment-naive individuals starting therapy and treatment-experienced patients for whom a switch to abacavir is being considered undergoing a test. Abacavir was avoided in all individuals testing HLA-B*5701 positive.
We identified all individuals undergoing prospective HLA-B*5701 testing (Delphic Diagnostics Ltd., London, UK) between August 2005 and July 2006. Using our prospectively collected cohort database we recorded ethnicity, sex, treatment history and adverse events; in addition a detailed case note review of all individuals who discontinued abacavir was performed. In order to compare the rate of HSR with the incidence observed before routine HLA screening we calculated the rate of HSR for the period August 2004 to July 2005.
The rates of abacavir HSR before and after the introduction of routine HLA-B*5702 testing were compared using the chi-squared test with Yates' correction. Chi-squared testing was also used to compare the incidence of HSR between groups (according to HLA status) with Yates' or Fisher's exact corrections depending on the subject numbers involved.
A total of 739 tests for HLA-B*5701 status were performed between August 2005 and July 2006. Of these tests, only four failed (failure rate 0.0054%); of the 735 successful tests, 54 yielded a positive HLA-B*5701 result (7.3%): 11 out of 111 (9.9%) women and 43 out of 624 (6.9%) men (P value 0.35). A breakdown of the results by ethnic group is presented in Table 1.
Of the 54 subjects with a positive HLA-B*5701 result, 25 were treatment naive. Fourteen remained off therapy at the time of analysis, nine commenced non-abacavir-containing HAART and two started abacavir-containing therapy (treatment started before a review of results), both of whom developed symptoms consistent with abacavir HSR. Of the 29 treatment-experienced subjects with the B*5701 haplotype, none switched to abacavir but seven had a history of previous abacavir exposure. Four of the seven had experienced abacavir HSR symptoms previously and three out of seven had tolerated abacavir for 9 days (switched off abacavir when HLA result reviewed), 6 weeks and 5 years, respectively.
A total of 681 individuals had a negative HLA-B*5701 result, 285 treatment naive and 396 treatment experienced. A total of 122 out of 285 treatment-naive patients started HAART, 47 abacavir-containing with no abacavir discontinuations. A total of 151 out of 396 treatment-experienced subjects switched to an abacavir-containing regimen; eight (5.3%) discontinued abacavir, four (2.6%) for suspected HSR. Of the four subjects who tested HLA-B*5701 negative yet discontinued abacavir with HSR-like symptoms, two underwent skin patch testing and one had a strong cutaneous reaction with contemporaneous systemic symptoms. The symptoms experienced by these four patients are described in Table 2.
A total of 144 patients commenced abacavir-containing HAART in the period before routine HLA screening and 10 discontinued abacavir for HSR-like symptoms (7.5%). After the introduction of HLA testing six out of 207 (3.0%) subjects starting or switching to abacavir experienced HSR symptoms; the difference in HSR incidence between the two time periods was not statistically significant (P = 0.10). If the two subjects who commenced abacavir before their HLA-B*5701 result was reviewed are excluded, that is the analysis is confined to the subjects who developed HSR symptoms despite a negative result, then the difference between the two time periods does become statistically significant: 7.5 versus 2.0% (P = 0.03).
The use of prospective HLA screening reduced the incidence of abacavir HSR in our cohort, although this was only statistically significant if the two subjects who started abacavir before their HLA result was reviewed are excluded from the analysis. There remain, however, four individuals who developed symptoms suspicious of HSR despite a negative HLA-B*5701 result, and one of these had a strongly positive skin patch test. Skin patch testing has been shown to be a durable test in individuals with a remote history of abacavir HSR, and positive results are significantly more frequent in HLA-B*5701 individuals with a history of abacavir exposure than their HLA-B*5701-negative counterparts . The prognostic value of skin patch testing is, however, uncertain and its role in clarifying abacavir HSR is currently under investigation in a large, international study.
Our results highlight the need to review results carefully before commencing therapy, and that it is essential to maintain clinical vigilance even in the presence of a negative HLA-B*5701 result.
We would like to acknowledge Andrea Gritz for her help in collecting and analysing the data.
1. Lucas A, Nolan D, Mallal S. HLA-B*5701 screening for susceptibility to abacavir hypersensitivity
. J Antimicrob Chemother
2007; E-pub ahead of print, 22 February 2007.
2. Mallal S, Nolan D, Witt C, Masel G, Martin AM, Moore C, et al
. Association between presence of HLAB*
5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet 2002; 359:727–732.
3. Hetherington S, Hughes AR, Mosteller M, Shortino D, Baker KL, Spreen W, et al
. Genetic variations in HLA-B region and hypersensitivity reactions to abacavir. Lancet 2002; 359:1121–1122.
4. Martin AM, Nolan D, Gaudieri S, Almeida CA, Nolan R, James I, et al
. Predisposition to abacavir hypersensitivity conferred by HLA-B*
5701 and a haplotypic Hsp70-Hom variant. Proc Natl Acad Sci U S A 2004; 101:4180–4185.
5. Rauch A, Nolan D, Martin A, McKinnon E, Almeida C, Mallal S. Prospective genetic screening decreases the incidence of abacavir hypersensitivity reactions in the western Australian HIV Cohort Study. Clin Infect Dis 2006; 43:99–102.
6. Reeves I, Churchill D, Fisher M. Screening for HLA-B*
5701 reduces the frequency of abacavir hypersensitivity reactions. In: Program and abstracts of the 8th International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV, San Francisco, CA, USA, 24–26 September 2006. Abstract 14.270.
7. Phillips EJ, Wong GA, Kaul R, Shahabi K, Nolan DA, Knowles SR, et al
. Clinical and immunogenetic correlates of abacavir hypersensitivity. AIDS 2005; 19:979–981.