Risk factor analysis of hypersensitivity reactions to abacavir

• Published in: Clinical therapeutics 2002-04, Vol.24 (4), p.565-573
• Main Authors: Symonds, William, Cutrell, Amy, Edwards, Mark, Steel, Helen, Spreen, Bill, Powell, Gwendolyn, McGuirk, Sue, Hetherington, Seth
• Format: Article
• Language: English
• Subjects: abacavir; Adult; adverse effects; African Continental Ancestry Group; Analysis of Variance; Anti-HIV Agents - adverse effects; Antibiotics. Antiinfectious agents. Antiparasitic agents; Antiviral agents; Biological and medical sciences; Dideoxynucleosides - adverse effects; drug hypersensitivity; Drug Hypersensitivity - epidemiology; Female; Humans; Logistic Models; Male; Medical sciences; nucleoside reverse-transcriptase inhibitor; Odds Ratio; Pharmacology. Drug treatments; Prognosis; Regression Analysis; Risk Factors
• ISSN: 0149-2918; 1879-114X
• DOI: 10.1016/S0149-2918(02)85132-3

Background: A hypersensitivity reaction to abacavir has been known to occur in ∼4% of patients treated with the drug. Objective: The goal of this study was to determine risk factors associated with the development of hypersensitivity reactions to abacavir. Methods: We constructed an analysis population from all protocols conducted by GlaxoSmithKline that involved at least 24 weeks of abacavir exposure with a quality-assured or validated clinical database by June 30, 2000. Demographic, clinical, and laboratory characteristics of patients who developed a hypersensitivity reaction to abacavir were compared with those of patients who did not. Univariate and multivariate logistic regression models were fit to understand the predictive ability of each potential risk factor. Odds ratios (ORs) and 95% Wald CIs were computed. Results: A total of 5332 patients exposed to abacavir were included in this analysis; 197 cases of hypersensitivity reaction were reported (3.7%). In univariate models, several associations were noted, but most subgroups produced rates within the expected range of 3% to 6%. The multivariate model using all demographic data available (Model 1) indicated that the risk of hypersensivitiy reaction among black patients (3% hypersensitivity reaction) was lower (OR = 0.59; 95% CI, 0.38–0.91) compared with other ethnic groups. In addition, a lower rate of hypersensitivity reaction was observed in patients who received previous therapy for HIV-1 infection with other antiretroviral agents (3% hypersensitivity reaction) compared with those receiving therapy for the first time (OR = 0.58; 95% CI, 0.44–0.78). Conclusions: The only characteristics identified as prognostic factors for hypersensitivity reaction to abacavir were antiretroviral treatment status (ie, treatment experience) and black race, each resulting in a lower rate compared with the expected incidence.