Risk of Breast Cancer With CXCR4-Using HIV Defined by V3 Loop Sequencing. (1st January 2015)

• Record Type: Journal Article
• Authors: Goedert, James J.; Swenson, Luke C.; Napolitano, Laura A.; Haddad, Mojgan; Anastos, Kathryn; Minkoff, Howard; Young, Mary; Levine, Alexandra; Adeyemi, Oluwatoyin; Seaberg, Eric C.; Aouizerat, Bradley; Rabkin, Charles S.; Harrigan, P. Richard; Hessol, Nancy A.
• Is Part Of: Journal of acquired immune deficiency syndromes. Volume 68:Number 1(2015)
• Publication Date: 2015-01-01
• Subjects: chemokine receptors -- HIV -- AIDS -- breast cancer -- parallel sequencing -- women; AIDS (Disease) -- Periodicals; Acquired Immunodeficiency Syndrome -- Periodicals; AIDS (Disease); Periodicals; 616.9792005
• DOI: 10.1097/QAI.0000000000000400
• Languages: English
• ISSNs: 1525-4135

Abstract

Abstract : Objective: Evaluate the risk of female breast cancer associated with HIV-CXCR4 (X4) tropism as determined by various genotypic measures. Methods: A breast cancer case–control study, with pairwise comparisons of tropism determination methods, was conducted. From the Women's Interagency HIV Study repository, one stored plasma specimen was selected from 25 HIV-infected cases near the breast cancer diagnosis date and 75 HIV-infected control women matched for age and calendar date. HIV-gp120 V3 sequences were derived by Sanger population sequencing (PS) and 454-pyro deep sequencing (DS). Sequencing-based HIV-X4 tropism was defined using the geno2pheno algorithm, with both high-stringency DS [false-positive rate (3.5) and 2% X4 cutoff], and lower stringency DS (false-positive rate, 5.75 and 15% X4 cutoff). Concordance of tropism results by PS, DS, and previously performed phenotyping was assessed with kappa (κ) statistics. Case-control comparisons used exact P values and conditional logistic regression. Results: In 74 women (19 cases, 55 controls) with complete results, prevalence of HIV-X4 by PS was 5% in cases vs 29% in controls ( P = 0.06; odds ratio, 0.14; confidence interval: 0.003 to 1.03). Smaller case-control prevalence differences were found with high-stringency DS (21% vs 36%, P = 0.32), lower stringency DS (16% vs 35%, P = 0.18), and phenotyping (11% vs 31%, P = 0.10). HIV-X4 tropism concordance was best between PS and lower stringency DS (93%, κ = 0.83). Other pairwise concordances were 82%–92% (κ = 0.56–0.81). Concordance was similar among cases and controls. Conclusions: HIV-X4 defined by population sequencing (PS) had good agreement with lower stringency DS and was significantly associated with lower odds of breast cancer.