1217
Association of Antibody to Human Immunodeficiency Virus Type 1 Core Protein (p24), CD4+ Lymphocyte Number, and AIDS-Free Time Homayoon Farzadegan, Joan S. Chmiel, Nancy Odaka, Linda Ward, Linda Poggensee, Alfred Saah, and John P. Phair
Department of Epidemiology. Johns Hopkins University School of Hygiene and Public Health. Baltimore. Maryland; Division oflnfectious Diseases. Department of Medicine. Comprehensive AIDS Center. and Cancer Center Biometry Section. Northwestern University Medical School. and Howard Brown Memorial Clinic. Chicago. lllinois
The diagnosis of infection with human immunodeficiency virus type I (HIV - I) can be achieved by using serologic tests for antibody, such as ELISA or Western blot [I, 2], and for the HIV p24 core protein, by using antigen capture assays with good sensitivity and specificity [3, 4]. In addition, use of serologic and cellular assays as markers or predictors of progression of HIV-I infection and disease status has been suggested [5-8]. We have investigated the dynamics of the appearance, persistence, and disappearance ofantibody to HIV p24 and viral HIV p24 as markers of the natural history of HIV-I infection and immunostatus of infected individuals. Specifically, the associations among these serologic markers, immunostatus of infected individuals, and HIV -I disease stage were studied cross-sectionally in homo- or bisexual men who were infected with HIV-I at entry and longitudinally in men who seroconverted during follow-up in the Multicenter AIDS Cohort Study (MACS).
Materials and Methods Study population. MACS is a prospective study of the natural history of HIV-I infection among 4954 homo- or bisexual
Received 25 March 1992; revised 22 July 1992. Presented in part: VI International Conference on AIDS, San Francisco. June 1990 (abstract Th.A.73). Informed consent was obtained from all study participants. The Multicenter AIDS Cohort Study (MACS) was reviewed and approved by the Institutional Review Boards of each MACS center (Baltimore and Chicago). Financial support: National Institutes of Health (AI-72634. Baltimore. and AI-32535. Chicago; National Cancer Institute). Reprints or correspondence: Dr. Homayoon Farzadegan, Department of Epidemiology, Rm. 6029. Johns Hopkins University School of Hygiene and Public Health. 615 N. Wolfe sr., Baltimore. MD 21205-2179. The Journal of Infectious Diseases 1992;166:1217-22 © 1992 by The University of Chicago. All rights reserved. 0022-1 899{92/6606-0003$0 1.00
men without AIDS enrolled between April 1984 and March 1985 in four centers [9]. About 36% of participants were infected with HIV- I at entry. All participants are evaluated semiannually. At each visit, blood is collected, a physical examination done, and a detailed questionnaire administered. Outcomes are identified and reported on a continuous basis by review of hospital records, death certificates, patient reports, and other relevant material. A subset ofMACS participants at Johns Hopkins University School of Hygiene and Public Health (Baltimore) and Northwestern University and Howard Brown Memorial Clinic (Chicago) was studied. The total number of MACS participants at these two centers was 2255 (I 153 in Baltimore, I 102 in Chicago). Participants were selected for this study on the basis of Centers for Disease Control (CDC) class [10] and development of HlV- I infection during the study and include a convenience sample of 206 (26%)of79 I baseline seropositive participants in the Baltimore and Chicago cohorts whose serum samples were analyzed cross-sectionallyand 84 (52%)of 163 participants with incident HIV- I infection (seroconverters) as of January 1990 who were studied longitudinally. Availability of quantitative anti-p24 kits determined the size of the cross-sectional sample. Seroconverters were selected for study if they had at least 4 samples available within the 36-month interval after seroconversion. CDC classes are defined as follows: class II/III, asymptomatic with or without generalized lymphadenopathy; class IV-A, at least one of four constitutional symptoms (fever, diarrhea, weight loss, or night sweats-for this study, self-reported symptoms in the 6 months before the visit were used); class IV-C, presence of thrush, herpes infection (facial, genital, or oral), or both; and AIDS, as defined in 1987 by CDC [10]. The cross-sectional study included 92 men in class II/III, 35 in class IV-A, 60 in class IV-C, and 19 with AIDS. Sixty-three participants in the cross-sectional study who were AIDS-free at the time blood was obtained subsequently progressed to AIDS as of November 199I. Eighteen ofthe 84 seroconverters had developed AIDS by the analysis date.
Downloaded from http://jid.oxfordjournals.org/ at Yale University on July 7, 2015
Serum antibody to p24 (anti-p24) and p24 antigen, alone and in combination with CD4+ lymphocyte number, were evaluated as predictors of progression of human immunodeficiency virus type 1 (HIV-1) infection. Two hundred six HIV-I-prevalent seropositive men in the Multicenter AIDS Cohort Study since 1984-1985 were studied cross-sectionally and 84 seroconverters were evaluated longitudinally. Cross-sectional analyses revealed significant associations among titer of anti-p24, CD4+ cell count, disease status (Centers for Disease Control class), and progression to AIDS. A high titer of anti-p24 was associated with lack of p24 antigenemia. Longitudinal studies of seroconverters demonstrated that a low titer of anti-p24, low CD4+ cell count, and detection of HIV-I p24 antigen are individually strong predictors of AIDS, but only low CD4+ cell count retains its independent predictive value in multivariate analysis of the three markers during the period immediately after infection with HIV-1.
Table 1. Cross-sectional study ofgeometric mean titer (GMT) of anti-p24 and mean CD4+ cell number for homosexual men at different stages of HlV-I infection.
CD4+ CDCclass (n) II/III (92)
IV-A (35) IV-C (60) AIDS (19)
lID 1992; 166 (December)
Farzadegan et al.
1218
Anti-p24 GMT' 95.58 ± 10.33 ± 16.78 ± 1.10 ±
6.95 12.86 9.75 3.08
(cells/mrrr')"
561 ± 33 479 ± 54 415 ± 38 77 ± 14
Laboratory tests. For p24 antigen testing, a solid-phase antigen-capture ELISA that used polyclonal HlV-l antibodies (Abbott Laboratories, Abbott Park, lL) was used according to the manufacturer's protocol. For anti-p24 titration, a competitive ELISA that used recombinant HlV-l p24 antigen was used (Abbott). Serum samples were diluted five times at fivefold for determining the titer of antibody to p24 (anti-p24). These titers were expressed as the reciprocal of the end-point dilution. A single value was obtained for each sample at each time point. For T cell phenotyping, flow cytometric analysis of T cells in peripheral blood was done as previously described [11]. Statistical analysis. Analysis of variance was used to compare CDC class means for continuous variables in the cross-sectional study. Pairwise comparisons were evaluated subsequently by two-sample t tests and Bonferroni's multiple comparison procedure. Reported Pvalues are two-sided and unadjusted for multiple comparisons unless otherwise indicated. x 2 tests were used to assess the significance ofassociations between categorical variables. Kaplan-Meier survival curves, log-rank tests, and Cox proportional hazards regression methods were used to evaluate the relative utility of single and multiple laboratory measures as markers of disease progression. Diagnosis of AIDS after entry but within the 72-month follow-up interval was the outcome variable in the cross-sectional evaluation of the predictive value of a single measurement of anti-p24. Analysis of the serial antip24 values during the first 36 months after seroconversion included all reported AIDS events as of the analysis date (March 1990). For the latter analyses, the three possible predictor variables were defined according to whether the marker was ever detected (p24 antigen) or ever not detected (anti-p24) during follow-up or ever fell below a threshold (CD4+,
Association of Antibody to Human Immunodeficiency Virus Type 1 Core Protein (p24), CD4+ Lymphocyte Number, and AIDS-Free Time Homayoon Farzadegan, Joan S. Chmiel, Nancy Odaka, Linda Ward, Linda Poggensee, Alfred Saah, and John P. Phair
Department of Epidemiology. Johns Hopkins University School of Hygiene and Public Health. Baltimore. Maryland; Division oflnfectious Diseases. Department of Medicine. Comprehensive AIDS Center. and Cancer Center Biometry Section. Northwestern University Medical School. and Howard Brown Memorial Clinic. Chicago. lllinois
The diagnosis of infection with human immunodeficiency virus type I (HIV - I) can be achieved by using serologic tests for antibody, such as ELISA or Western blot [I, 2], and for the HIV p24 core protein, by using antigen capture assays with good sensitivity and specificity [3, 4]. In addition, use of serologic and cellular assays as markers or predictors of progression of HIV-I infection and disease status has been suggested [5-8]. We have investigated the dynamics of the appearance, persistence, and disappearance ofantibody to HIV p24 and viral HIV p24 as markers of the natural history of HIV-I infection and immunostatus of infected individuals. Specifically, the associations among these serologic markers, immunostatus of infected individuals, and HIV -I disease stage were studied cross-sectionally in homo- or bisexual men who were infected with HIV-I at entry and longitudinally in men who seroconverted during follow-up in the Multicenter AIDS Cohort Study (MACS).
Materials and Methods Study population. MACS is a prospective study of the natural history of HIV-I infection among 4954 homo- or bisexual
Received 25 March 1992; revised 22 July 1992. Presented in part: VI International Conference on AIDS, San Francisco. June 1990 (abstract Th.A.73). Informed consent was obtained from all study participants. The Multicenter AIDS Cohort Study (MACS) was reviewed and approved by the Institutional Review Boards of each MACS center (Baltimore and Chicago). Financial support: National Institutes of Health (AI-72634. Baltimore. and AI-32535. Chicago; National Cancer Institute). Reprints or correspondence: Dr. Homayoon Farzadegan, Department of Epidemiology, Rm. 6029. Johns Hopkins University School of Hygiene and Public Health. 615 N. Wolfe sr., Baltimore. MD 21205-2179. The Journal of Infectious Diseases 1992;166:1217-22 © 1992 by The University of Chicago. All rights reserved. 0022-1 899{92/6606-0003$0 1.00
men without AIDS enrolled between April 1984 and March 1985 in four centers [9]. About 36% of participants were infected with HIV- I at entry. All participants are evaluated semiannually. At each visit, blood is collected, a physical examination done, and a detailed questionnaire administered. Outcomes are identified and reported on a continuous basis by review of hospital records, death certificates, patient reports, and other relevant material. A subset ofMACS participants at Johns Hopkins University School of Hygiene and Public Health (Baltimore) and Northwestern University and Howard Brown Memorial Clinic (Chicago) was studied. The total number of MACS participants at these two centers was 2255 (I 153 in Baltimore, I 102 in Chicago). Participants were selected for this study on the basis of Centers for Disease Control (CDC) class [10] and development of HlV- I infection during the study and include a convenience sample of 206 (26%)of79 I baseline seropositive participants in the Baltimore and Chicago cohorts whose serum samples were analyzed cross-sectionallyand 84 (52%)of 163 participants with incident HIV- I infection (seroconverters) as of January 1990 who were studied longitudinally. Availability of quantitative anti-p24 kits determined the size of the cross-sectional sample. Seroconverters were selected for study if they had at least 4 samples available within the 36-month interval after seroconversion. CDC classes are defined as follows: class II/III, asymptomatic with or without generalized lymphadenopathy; class IV-A, at least one of four constitutional symptoms (fever, diarrhea, weight loss, or night sweats-for this study, self-reported symptoms in the 6 months before the visit were used); class IV-C, presence of thrush, herpes infection (facial, genital, or oral), or both; and AIDS, as defined in 1987 by CDC [10]. The cross-sectional study included 92 men in class II/III, 35 in class IV-A, 60 in class IV-C, and 19 with AIDS. Sixty-three participants in the cross-sectional study who were AIDS-free at the time blood was obtained subsequently progressed to AIDS as of November 199I. Eighteen ofthe 84 seroconverters had developed AIDS by the analysis date.
Downloaded from http://jid.oxfordjournals.org/ at Yale University on July 7, 2015
Serum antibody to p24 (anti-p24) and p24 antigen, alone and in combination with CD4+ lymphocyte number, were evaluated as predictors of progression of human immunodeficiency virus type 1 (HIV-1) infection. Two hundred six HIV-I-prevalent seropositive men in the Multicenter AIDS Cohort Study since 1984-1985 were studied cross-sectionally and 84 seroconverters were evaluated longitudinally. Cross-sectional analyses revealed significant associations among titer of anti-p24, CD4+ cell count, disease status (Centers for Disease Control class), and progression to AIDS. A high titer of anti-p24 was associated with lack of p24 antigenemia. Longitudinal studies of seroconverters demonstrated that a low titer of anti-p24, low CD4+ cell count, and detection of HIV-I p24 antigen are individually strong predictors of AIDS, but only low CD4+ cell count retains its independent predictive value in multivariate analysis of the three markers during the period immediately after infection with HIV-1.
Table 1. Cross-sectional study ofgeometric mean titer (GMT) of anti-p24 and mean CD4+ cell number for homosexual men at different stages of HlV-I infection.
CD4+ CDCclass (n) II/III (92)
IV-A (35) IV-C (60) AIDS (19)
lID 1992; 166 (December)
Farzadegan et al.
1218
Anti-p24 GMT' 95.58 ± 10.33 ± 16.78 ± 1.10 ±
6.95 12.86 9.75 3.08
(cells/mrrr')"
561 ± 33 479 ± 54 415 ± 38 77 ± 14
Laboratory tests. For p24 antigen testing, a solid-phase antigen-capture ELISA that used polyclonal HlV-l antibodies (Abbott Laboratories, Abbott Park, lL) was used according to the manufacturer's protocol. For anti-p24 titration, a competitive ELISA that used recombinant HlV-l p24 antigen was used (Abbott). Serum samples were diluted five times at fivefold for determining the titer of antibody to p24 (anti-p24). These titers were expressed as the reciprocal of the end-point dilution. A single value was obtained for each sample at each time point. For T cell phenotyping, flow cytometric analysis of T cells in peripheral blood was done as previously described [11]. Statistical analysis. Analysis of variance was used to compare CDC class means for continuous variables in the cross-sectional study. Pairwise comparisons were evaluated subsequently by two-sample t tests and Bonferroni's multiple comparison procedure. Reported Pvalues are two-sided and unadjusted for multiple comparisons unless otherwise indicated. x 2 tests were used to assess the significance ofassociations between categorical variables. Kaplan-Meier survival curves, log-rank tests, and Cox proportional hazards regression methods were used to evaluate the relative utility of single and multiple laboratory measures as markers of disease progression. Diagnosis of AIDS after entry but within the 72-month follow-up interval was the outcome variable in the cross-sectional evaluation of the predictive value of a single measurement of anti-p24. Analysis of the serial antip24 values during the first 36 months after seroconversion included all reported AIDS events as of the analysis date (March 1990). For the latter analyses, the three possible predictor variables were defined according to whether the marker was ever detected (p24 antigen) or ever not detected (anti-p24) during follow-up or ever fell below a threshold (CD4+,
