The Effect of Influenza Vaccination on Human Immunodeficiency Virus Type 1 Load: A Randomized, Double-Blind, Placebo-Controlled Study

1332 The Effect of Influenza Vaccination on Human Immunodeficiency Virus Type 1 Load: A Randomized, Double-Blind, Placebo-Controlled Study Marshall J. Glesby, Donald R. Hoover, Homayoon Farzadegan, Joseph B. Margolick, and Alfred J. Saah Division of Infectious Diseases, Department of Medicine, Johns Hopkins University School of Medicine; Department of Epidemiology and Department of Molecular Microbiology and Immunology, Johns Hopkins University School of Hygiene and Public Health, Baltimore, Maryland Since antigenic stimulation causes CD4 cell activation and up-regulation of human immunodeficiency virus type 1 (HIV1) in vitro [1], concern has been raised about the effects of vaccinating persons with HIV-1 infection [2, 3]. Vaccination with a T cell-dependent antigen, such as inactivated influenza virus vaccine, might be harmful if it led to CD4 cell activation, increased HIV-1 replication, and subsequent loss of CD4 cells. Recent studies have reported conflicting findings as to whether HIV-1 load increases after influenza vaccination [4- 8]. Absence of placebo controls has limited the interpretation of published reports, since factors other than vaccination could result in transient changes in HIV-1 viremia [9]. To determine if influenza vaccination adversely affects HIV1 load, we conducted a randomized, double-blind, placebocontrolled trial of influenza vaccination in HIV-1 - infected subjects with CD4 cell counts in the intermediate range (200-500 cells/j.LL). Methods Study population and protocol. Eligible subjects were ~ 18 years of age and HIV-1- seropositive with absolute CD4 cell Received 19 March 1996; revised 1 July 1996. Presented in part: Third Conference on Retroviruses and Opportunistic Infections, Washington, DC, 28 January-1 February 1996 (abstract no. 98). Informed consent was obtained from all study participants, and the guidelines of the US Department of Health and Human Services and those of the Joint Committee on Clinical Investigation of the Johns Hopkins Medical Institutions were followed. Grant support: NIH (cooperative agreement IUOl-35042-03, grant RR00722). Reprints or correspondence (present address): Dr. Marshall J. Glesby, Community Research Initiative on AIDS, 275 7th Ave., New York, NY 10001. The Journal of Infectious Diseases 1996; 174:1332-6 © 1996 by The University of Chicago. All rights reserved. 0022-1899/96/7406-0027$01.00 counts between 200 and 500 cells/J.1,L or CD4 cell percentage between 14% and 29%. Subjects were randomized to receive influenza vaccine or a saline placebo injection using permuted blocks of size 3, such that for every three treatment assignments, 2 subjects received vaccine and 1 subject placebo. The randomization was stratified on use of antiretroviral therapy to achieve a balance in this factor between vaccine and placebo groups. Subjects, investigators, and laboratory, clinic, and data entry personnel were blinded to treatment assignment until completion of the initial 90 days of follow-up. Vaccine recipients received the 1994-1995 trivalent split-virus influenza vaccine (Wyeth-Ayerst, Marietta, PA) containing the following antigens: A/Texas/36/91 (H 1N 1), A/ShangDong/9/93 (H3N2), and BlPanama/45/90. Placebo recipients received a saline injection from an identical-appearing syringe. The day of injection was considered to be day 0 of the study. HIV-l load assays. HIV-1 load was measured blindly on coded specimens by two different assays at days - 7, 0, 7, 10, 14, and 30. Plasma HIV-1 viremia was measured using the branched DNA (bDNA) amplification method [10] (assay kits provided by Chiron, Emeryville, CA). For subjects with plasma viremia that was undetectable by the standard bDNA assay (lower level of detection, 10 kEq/mL), an ultrasensitive bDNA assay [11] (lower level of detection, 0.5 kEq/ mL) was done by Chiron on all of that subject's specimens when possible. If sufficient plasma was not available to run the uItrasensitive assay, a value of 5 kEq/mL (half of the minimal detectable level of 10 kEq/mL) was assigned for that visit. If virus load was undetectable by the uItrasensitive assay at a given visit, a value of 0.25 kEq/mL (half of the minimal detectable level of 0.5 kEq/ mL) was assigned for that visit. Infectious cell-associated HIV-1 viremia was assayed by peripheral blood mononuclear cell microculture of fresh blood samples using the standardized consensus protocol of the AIDS Clinical Trial Group [12]. Standardized computer software was used to ascertain infectious units of HIV-1 per million peripheral blood To determine if vaccination induces replication of human immunodeficiency virus type 1 (HIV1),42 HIV-1-infected subjects with CD4 cell counts of200-500 cellS/ILL were randomized to receive influenza vaccine or saline placebo. Infectious cell-associated and plasma HIV-1 RNA virus load were measured twice at baseline and then at 7, 10, 14, and 30 days after injection by quantitative microculture and branched DNA amplification. The ratios of the geometric mean plasma HIV-1 load of the four follow-up visits compared with baseline in vaccine (n = 28) and placebo (n = 14) recipients were similar (1.05 [95% confidence interval, 0.80-1.37] for vaccine; 0.96 [95% confidence interval, 0.68-1.33] for placebo; P = .90). The geometric mean ratios of plasma virus load at each follow-up visit to baseline did not differ significantly from 1.0 for each group. Infectious cellassociated virus load measures yielded similar results. CD4 cell counts declined similarly in both groups at 6 months. Influenza vaccination did not increase HIV-1 load in this controlled clinical trial. JID 1996; 174 (December) Concise Communications Results Patient characteristics and follow-up. Twenty-eight subjects were randomized to the influenza vaccine group and 14 subjects to the placebo group. Baseline characteristics of these subjects are summarized in table 1. The groups were similar with respect to demographic characteristics, prior history of influenza vaccination, and use of antiretroviral drugs. Although absolute CD4 cell counts and CD4 cell percentages were similar between groups, the vaccine group had higher geometric mean plasma and infectious cell-associated HIV-1 load at baseline, but the differences between groups were not statistically significant (P = .09 for plasma viremia and P = .27 for infectious cell-associated viremia). HIV-l viremia. Both plasma (figure 1) and infectious cellassociated viremia were relatively constant over the initial 30 days after injection in both vaccine and placebo groups. Although relatively minor fluctuations in mean virus load occurred over time in each group, there were no statistically significant changes from baseline. For each group, the ratio of the geometric mean plasma viremia of the four follow-up visits to the geometric mean baseline plasma viremia was close to 1.0 (1.05 [95% CI, 0.80-1.37] for vaccine; 0.96 [95% CI, Table 1. Baseline characteristics of subj (...truncated)