The abacavir hypersensitivity reaction is a potentially fatal syndrome occurring in approximately 5% of HIV-infected patients exposed to this nucleoside analogue after a median of 11 days (range, 1–318 days) [1,2]. Systemic manifestations can include fever, rash, fatigue, and gastrointestinal or respiratory symptoms. In the largest retrospective analysis, non-African ethnicity, female sex, Centers for Disease Control and Prevention (CDC) HIV class A or B, and use of a standardized report form were associated with suspected hypersensitivity [1,2]. The presence of HLA-B*5701 has been associated with elevated odds of hypersensitivity in Caucasian and Hispanic populations [3–6].
Nine (18%) of 50 HIV-positive men treated with abacavir at the University of Washington Primary Infection Clinic developed suspected hypersensitivity after 1–60 days on therapy. Previous investigations have focused on persons with chronic HIV infection; this retrospective analysis evaluated risk factors for hypersensitivity among our cohort of individuals with primary HIV infection (PHI).
Materials and methods
From 2000 to 2004, Acute HIV Infection and Early Disease Research Program (AIEDRP) treatment protocol AI-02-001 evaluated effects of antiretroviral treatment in PHI, including individuals with either acute or early HIV infection. Eligible subjects had detectable HIV RNA and negative EIA (acute HIV infection) or a positive EIA with either negative HIV test within 240 days or negative ‘detuned’ antibody test (LS-EIA) (early HIV infection). We estimated the date of HIV infection to be the first day of symptoms or, for asymptomatic individuals, the midpoint between date of presentation and last negative HIV test. This study was approved by the University of Washington Institutional Review Board, and all subjects gave written consent.
AI-02-001 Study procedures
Depending on enrollment date and protocol version, subjects were offered one of three open-label antiretroviral (ARV) regimens including abacavir (Ziagen) and lamivudine (Epivir) plus: (i) efavirenz (Sustiva); (ii) efavirenz + indinavir (Crixivan); or (iii) ritonavir (Norvir)-boosted indinavir. Subjects were instructed to consult with study clinicians for symptoms suggestive of abacavir hypersensitivity. In such cases, subjects discontinued abacavir and were evaluated within 24 h. For this analysis, the ‘follow-up’ visit was the study visit most proximal to the suspected hypersensitivity reaction for cases and to day 9 (the median day of abacavir hypersensitivity) for controls.
Suspected cases of abacavir hypersensitivity were identified prospectively by study clinicians and reviewed retrospectively by two authors without knowledge of the subjects' HLA types. Subjects were classified using standardized case definitions: (i) highly probable abacavir hypersensitivity — clinically compatible syndrome with three or more symptoms, symptom resolution after abacavir discontinuation, no symptoms when ARV was continued without abacavir, and no alternative explanation; (ii) probable abacavir hypersensitivity — compatible syndrome with either two symptoms in subjects who continued ARV (with substitution for abacavir) or three or more symptoms in subjects who discontinued all ARV; (iii) possible abacavir hypersensitivity — compatible syndrome with alternative explanations or symptoms atypical for hypersensitivity.
Rechallenge was not permitted due to reports of fatal reactions caused by repeated administration of abacavir following a hypersensitivity reaction, and therefore no individual was classified as having definite hypersensitivity. Control subjects were participants in AI-02-001 in Seattle who received abacavir without evidence of hypersensitivity. Charts were reviewed for all controls experiencing rash within 90 days after initiating abacavir to confirm classification as controls.
From the study start until June 2002, plasma HIV quantification was performed using bDNA (Chiron Corporation, Emeryville, California, USA). Subsequently, a real-time PCR assay was used (Taqman, Roche Diagnostics, Basel, Switzerland) (Coombs RC, personal communication).
Fluorescent activated cell sorting
Fluorescent activated cell sorting was used to quantify activated (CD45RO+/CD62L−/HLA-DR+), resting (CD45RO+/CD62L−/HLA-DR−), and naive (CD45RA+/CD62L+/HLA-DR−) CD4 T-cells and other cells expressing activation markers including CD25, CD38, and CD95.
Class I and II alleles at A, B, DRB1, and DQB1 loci were identified by PCR using sequence-specific primers (Micro SSP DNA Typing Kit; One Lambda Inc., Canoga Park, California, USA) at the Puget Sound Blood Center Immunogenetics Laboratory (Seattle, Washington, USA) according to manufacturer's instructions.
We examined peripheral blood mononuclear cells (PBMC) from six subjects for abacavir-specific lymphoproliferation, interferon (IFN)-γ producing cells (ELISpot), and IFN-γ and tumor necrosis factor (TNF)-α production in culture supernatant. Testing included four subjects with suspected hypersensitivity (patients #3, 4, 5, and 7) and two who were abacavir-tolerant. Both subjects having HLA-B*5701 (#4 and 7) were tested. In lymphoproliferation assays (LPA) , cryopreserved PBMC were plated at 1 × 105 cells/well in quadruplicate into wells having no antigen (background), abacavir (0.1, 0.5 and 1.0 μg/ml), or phytohemagglutinin (PHA, 1 μg/ml, positive control). Abacavir concentrations were selected to be representative of levels achieved in blood plasma . The stimulation index was the ratio of median counts per minute (c.p.m.) in stimulated wells to median c.p.m. in control wells. An SI ≥ 3 was considered positive.
Cytokine secretion in cell culture supernatants was measured by ELISA (Endogen, Boston, Massachusetts, USA) according to manufacturer's instructions. PBMC were incubated at 2 × 106 cells/ml in 24-well plates in the presence of abacavir (1.0 μg/ml), PHA (2 μg/ml), or medium alone (negative control). IFN-γ and TNF-α were measured in supernatants harvested after 5 and 2 days, respectively. All evaluations were run in duplicate, and lower limits of detection of the assay were 1.2 pg/ml for IFN-γ and 8.0 pg/ml for TNF-α.
ELISpot assays for IFN-γ producing antigen-specific T cells were performed as previously described . Cryopreserved PBMC were thawed and plated at 1 × 105 cells/well in duplicate with abacavir (0.1, 0.5, and 1.0 μg/ml), PHA (2 μg/ml), or medium alone. An automated ELISpot reader (Immunospot; Cellular Technology Cleveland, Ohio, USA) counted colored spot-forming cells (SFC). Responses were considered positive if ≥ 50 SFC were detected per 1 × 106 cells and mean SFC were twofold higher than in negative controls.
Univariate linear logistic regression, t-tests, Wilcoxon rank-sum tests, and Fisher's exact test were performed using STATA software (StataCorp LP, College Station, Texas, USA). The confidence interval for incidence of hypersensitivity was calculated by binomial distribution.
Fifty men were enrolled in AI-02-001 at the University of Washington Primary Infection Clinic. Forty-five (90%) experienced symptoms associated with acute infection; 13 (26%) had fever and rash. On the first day of ARV (day 0), one case and one-third of controls continued to have at least one acute HIV-related symptom.
Day 0 was a median 50 days after the estimated date of HIV acquisition (range, 7–134 days), and subjects had a mean (range) CD4 T-cell count of 467 (138–943) cells/μl and mean log10 plasma HIV RNA of 4.98 (3.74–7.17) copies/ml. By day 0, the Western blot was positive in 38 of 45 (84.4%) subjects tested.
Nine [18%; 95% confidence interval (CI), 8.6–31.4%] subjects experienced symptoms consistent with suspected abacavir hypersensitivity (Table 1). All individuals recovered without sequelae. Five control subjects taking efavirenz and two on protease inhibitor therapy developed rashes without systemic symptoms, and ARV were continued without interruption. One control subject died on day 3 due to an unrelated intracranial hemorrhage.
The presence of HLA-B*5701 was associated with a relative risk of hypersensitivity of 6.9 (95% CI, 3.5–13.6; P = 0.03). In univariate regression analyses (Table 2), there was either association or trend towards association of abacavir hypersensitivity with markers of more robust immunologic status, including higher baseline CD4 percentages and naive T cells (CD45RA/CD62L); lower number of CD8 T cells, percentages, and cells expressing CD38; lower baseline HIV viremia; and greater reduction in HIV viremia at follow up. Variables that were not associated included age, concurrent ARV, medication allergies, hepatitis virus coinfection, baseline complete blood counts and liver function tests, change in CD4 and CD8 T cells from baseline, and other cell surface markers including CD95 (Fas). Race/ethnicity could not be sufficiently evaluated because of the homogeneity of our patient population, and multiple variable regression analyses were not performed due to the small sample size.
We observed no significant abacavir-specific lymphoproliferation or IFN-γ production in PBMC among four cases and two control subjects who had cells available for testing (Table 3). Only small amounts of IFN-γ (10–21 pg/ml, 8–17-fold over background) were detected in abacavir-stimulated culture supernatants from the two cases having HLA-B*5701. We detected small amounts of TNF-α in culture supernatants from all six subjects, but there was no significant difference in TNF-α production in vitro prior to or following abacavir use (Table 3 and data not shown).
We report a rate of suspected abacavir hypersensitivity reaction in a cohort of individuals with PHI (18%; 95% CI, 8.6–31.4%) that is significantly higher than expected for a cohort with chronic infection (5%). It is plausible that initiation of treatment during PHI is associated with an increased frequency of hypersensitivity. Risk factors have been established for abacavir hypersensitivity in chronic HIV infection, but the mechanism remains undetermined. Although our conclusions are dependent on the accuracy of case identification, it is possible that the differential rate of hypersensitivity during PHI may be due to an alternate mechanism responsible for hypersensitivity in early infection compared to chronic HIV infection or that a common mechanism may be associated with factors that peak soon after HIV acquisition.
We observed no difference in cytokine production or lymphocyte proliferation in response to in vitro stimulation with abacavir in individuals with suspected hypersensitivity compared to abacavir-tolerant controls. This contrasts with two previous investigations in which the proportions of TNF-α  and interleukin (IL)-4  positive cells were higher in persons with abacavir hypersensitivity compared to abacavir-tolerant controls. These differences support the possibility that other factors may be associated with abacavir hypersensitivity during PHI. We did not assess IL-4 production in response to abacavir stimulation in our patient population, but it is not unreasonable that our other findings might differ from previous investigations given that, at least in the first study, 94% of cases were HLA-B*5701 positive.
Although HLA-B*5701 was statistically associated with suspected hypersensitivity in our cohort, the polymorphism occurred less frequently than expected based on studies of individuals with chronic infection [3–5]. HLA-B*5701 has been associated with a decreased frequency of symptomatic acute HIV infection and an increased frequency of viral control during primary HIV infection . As a consequence, individuals with the HLA-B*5701 allele would be less likely to be diagnosed with primary infection and enrolled. If hypersensitivity during primary infection is predominantly dependent on HLA-B*5701 status, we would have expected significantly fewer cases of hypersensitivity rather than what we observed. Although HLA-B*5701 screening prior to abacavir initiation is not standard of care, screening could identify patients at risk for hypersensitivity and may be cost-effective in chronic HIV infection . Even so, screening would have failed to prevent seven suspected cases in our cohort and would not eliminate the need for clinical management of abacavir hypersensitivity.
In previous analyses, immune markers such as higher CD4 T-cell count and CDC class A or B were associated with abacavir hypersensitivity [1,2]. Our study suggests that markers of a more intact immune system are also associated with hypersensitivity during PHI, although this may be related to association with HLA-B*5701 .
We did not anticipate the association of hypersensitivity with abacavir initiation later within the period of PHI. Relatively severe immunosuppression occurs immediately following HIV acquisition that is characterized by decreasing CD4 T-cell counts, high plasma HIV RNA levels, and risk for opportunistic infections normally associated with advanced HIV disease . Schacker et al. reported stabilization of CD4 T-cell counts around 120 days into HIV infection that was also associated with an inflection (set) point in RNA levels . The exact mechanism for HIV control may be related to T-cell activation or factors that peak at this time. Our study suggests that abacavir hypersensitivity risk may parallel these responses.
Six other small studies of subjects who received abacavir during PHI reported rates of suspected hypersensitivity between 0% and 29%. ([14–17] and J. Hernandez, personal communication). In two of three studies where data were available, subjects initiated abacavir earlier in the period following HIV acquisition than in our cohort.
The greatest limitation of our study is the absence of objective gold standard for the diagnosis of abacavir hypersensitivity. Potential for case misclassification may be increased by aspects of our study, including lack of blinding to abacavir, concomitant efavirenz use, and association of PHI itself with fever and rash. Although potentially diagnostic, rechallenge with abacavir is contraindicated due to potential for severe or even fatal consequences. Thus management tends to err towards discontinuation of abacavir if hypersensitivity is considered, as shown by the increased diagnosis of abacavir hypersensitivity in studies with standardized reporting  and the diagnosis of abacavir hypersensitivity among 3% of individuals receiving abacavir placebo in a double-blind clinical trial . Possible diagnostic tests include HLA-B*5701 testing, TNF-α, and patch testing ; however, these tests may have differential utility in primary infection, and clinical diagnosis remains the standard.
In conclusion, treatment with abacavir during PHI may be associated with increased risk for hypersensitivity. We identified that risk factors associated with abacavir hypersensitivity in chronic infection, such as HLA-B*5701 and markers of a more intact immune status and viral control, were associated with suspected hypersensitivity in PHI. The temporal association of hypersensitivity with initiation of abacavir later during early infection raises questions for further investigation.
We thank all of the individuals who participated in the study, Terri Smith for data assistance, and Jaime E. Hernandez, MD and Irene Gray from GlaxoSmithKline for sharing their ABC HSR expertise.
Sponsorship: Dr. Stekler was supported by NIH T32 AI07140-26. The study was sponsored by the Division of AIDS National Institute of Allergy and Infectious Diseases (AI41535 and AI57005). Pharmaceutical support was provided by: Bristol Myers Squibb (formerly DuPont Pharmaceuticals Co.), GlaxoSmithKline, and Merck.
1. Brothers C, Cutrell A, Zhao H, Edwards M, Fleming J, Scott T, et al
. Once-daily administration of abacavir is not a clinical risk factor for suspected hypersensitivity reactions in clinical trials, and rash alone is not sufficient to diagnose the reaction
. Twelfth Conference on Retroviruses and Opportunistic Infections
. Boston, MA, February 2005 .
2. Cutrell AG, Hernandez JE, Fleming JW, Edwards MT, Moore MA, Brothers CH, et al
. Updated clinical risk factor analysis of suspected hypersensitivity reactions to abacavir. Ann Pharmacother 2004; 38:2171–2172.
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. Mallal S, Nolan D, Witt C, Masel G, Martin AM, Moore C, et al
. Association between presence of HLA-B*
5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1 reverse transcriptase inhibitor abacavir. Lancet 2002; 359:727–732.
5. Hughes DA, Vilar FJ, Ward CC, Alfirevic A, Park BK, Pirmohamed M. Cost-effectiveness analysis of HLA B*
5701 genotyping in preventing abacavir hypersensitivity. Pharmacogenetics 2004; 14:335–342.
6. Hughes AR, Mosteller M, Bansal AT, Davies K, Haneline SA, Lai EH, et al
. Association of genetic variations in HLA-B region with hypersensitivity to abacavir in some, but not all, populations. Pharmacogenomics 2004; 5:203–211.
7. Malhotra U, Berrey MM, Huang Y, Markee J, Brown DJ, Ap S, et al
. Effect of combination antiretroviral therapy on T-cell immunity in acute human immunodeficiency virus type 1 infection. J Infect Dis 2000; 181:121–131.
8. Donnerer J, Kronawetter M, Kapper A, Haas I, Kessler HH. Therapeutic drug monitoring of the HIV/AIDS drugs abacavir, zidovudine, efavirenz, nevirapine, indinavir, lopinavir, and nelfinavir. Pharmacology 2003; 69:197–204.
9. Russell ND, Hudgens MG, Ha R, Havenar-Daughton C, McElrath MJ. Moving to human immunodeficiency virus type 1 vaccine efficacy trials: defining T cell responses as potential correlates of immunity. J Infect Dis 2003; 187:226–242.
10. 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 USA 2004; 101:4180–4185.
11. King D, Tomkins S, Waters A, Easterbrook PJ, Thurmond LM, Thorborn DE, et al
. Intracellular cytokines may model immunoregulation of abacavir hypersensitivity in HIV-infected subjects. J Allergy Clin Immunol 2005; 115:1081–1087.
12. Altfeld M, Addo MM, Rosenberg ES, Hecht FM, Lee PK, Vogel M, et al
. Influence of HLA-B57
on clinical presentation and viral control during acute HIV-1 infection. AIDS 2003; 17:2581–2591.
13. Schacker TW, Hughes JP, Shea T, Coombs RW, Corey L. Biological and virologic characteristics of primary HIV infection. Ann Intern Med 1998; 128:613–620.
14. Kost RG, Hurley A, Zhang L, Vesanen M, Talal A, Furlan S, et al
. Open-label phase II trial of amprenavir, abacavir, and fixed-dose zidovudine/lamivudine in newly and chronically HIV-1-infected patients. J Acquir Immune Defic Syndr 2001; 26:332–339.
15. Cooper D, Perrin L, Kinloch, Janossy G, Gazzard B, Phillips A, et al
. Intervention with quadruple HAART (Combivir/Abacavir/Amprenavir) during primary HIV-1 infection (PHI) is associated with rapid viraemia clearance and decrease of immune activation
. Seventh Conference on Retroviruses and Opportunistic Infections
. San Francisco, January 2000 [abstract 552].
16. Markowitz M, Hurley A, Ramratnam B, Louie M, Kost R, Captan B, et al
. The safety and efficacy of a Ritonavir-boosted Amprenavir-based regimen after switch from Amprenavir-based HAART
. Eighth Conference on Retroviruses and Opportunistic Infections
. Chicago, February 2001 [abstract 405].
17. Bhaduri S, Miller A. Response to ‘Is antiretroviral treatment of primary HIV infection clinically justified on the basis of current evidence' by Smith et al.
. Acquir Immune Defic Syndr 2004; 18:2448.
18. DeJesus E, Herrera G, Teofilo E, Gerstoft J, Buendia CB, et al
. Abacavir versus zidovudine combined with lamivudine and efavirenz, for the treatment of antiretroviral-naïve HIV-infected adults. Clin Infect Dis 2004; 39:1038–1046.
19. Phillips EJ, Sullivan JR, Knowles SR, Shear NH. Utility of patch testing in patients with hypersensitivity syndromes associated with abacavir. AIDS 2002; 16:2223–2225.