Although prior studies have causally linked AIDS to HIV and behavioral factors associated with primary HIV infection, we have a limited understanding about behavioral factors that influence HIV disease progression among those already infected [1–3]. Recent studies, variable clinical courses and variable patterns of CD4 cell loss suggest that other factors affect HIV disease progression [4–9]. Although we know that HIV is transmitted through exposure to blood and body fluids, including ejaculate, it remains unclear whether ejaculate exposure through receptive anal intercourse (RAI) influences the rate of immunologic deterioration among those already infected [10,11]. It is likely that infectious diseases, including those transmitted in ejaculate, may cause endogenous extension of HIV infection and CD4 cell loss [12–14].
HIV infection causes impaired immune responses early in disease that get worse with time over the course of the infection [15–19]. CD4 cells are selectively infected and depleted during the course of disease; prior studies suggest that factors which increase the cell-to-cell transmission of HIV infection may enhance overall disease progression [12,20–23]. HIV replication in CD4 cells appears to be dependent upon the differentiation stage of the cells that are infected, the patient's clinical disease stage and the type of activation signal used to stimulate the cells [12,13,23]. Infectious disease agents can activate CD4 cells in vitro and may extend infection in vivo [12,13].
The relations between CD4 cell count, HIV virus replication rates and immunogen exposures are closely intertwined. CD4 cell counts are inversely related to HIV viral replication rates and viral burden in vivo [24–28]; HIV virus replication rates increase with exposure to infectious agents or other general immunogens in vitro [12,13]. While not all immunogens that have been studied and shown to induce HIV replication are sexually transmitted, at least one, HTLV-I, can be sexually transmitted in humans. Other investigators have reported animal and human data that suggest that cytomegalovirus, a sexually transmissible virus, may influence the extent of HIV replication in vivo and, possibly, HIV disease progression [29,30]. Additionally, while sexually transmitted diseases can be conveyed in the absence of exposure to ejaculate, men who report being exposed to ejaculate were more certain to have been exposed to their partner's body fluids than men who reported unprotected RAI alone. Therefore, exposure to infectious diseases and other immunogens through exposure to a partner's ejaculate may increase (HIV) viral replication rates and serve as one pathway for rapid CD4 cell depletion and overall disease progression. We postulate that infectious diseases that are associated with ejaculate exposure may be a causal factor for rapid CD4 cell loss, although these were not measured in the data presented here. To investigate whether RAI with ejaculate exposure is associated with rapid loss of CD4 T lymphocytes, we studied 1808 HIV-seropositive men from the Multicenter AIDS Cohort Study (MACS).
Study population and data collection
The MACS evaluates the natural history of HIV disease and has been described extensively elsewhere . Briefly, between April 1984 and March 1985, men who identified themselves as having sex with men and had not yet developed an AIDS-defining illness were recruited at four study centers. A total of 4955 men were enrolled in the cohort and when tests for the antibody later became available in 1985, 1808 were found to have been infected with HIV at the time of enrollment (HIV-seropositive men).
Demographic, treatment, medication and behavioral data are collected through a combination of self- and interviewer-guided questionnaires at semiannual visits. Physical examinations are conducted at each follow-up visit by a trained health care provider. CD4 T-lymphocyte cell counts are determined from specimens collected at the time of each follow-up interview.
Data collected at the first ten semiannual follow-up visits at the MACS (i.e., 1984–1989) were used for these analyses to maximize power. Exposure to ejaculate through RAI substantially decreased during the first 5 years of the MACS as men in the cohort began practicing safer sex. The availability and diversity of treatments for AIDS-defining conditions and HIV infection, (e.g., prophylaxis for Pneumocystis carinii pneumonia and therapy with protease inhibitors) increased steadily during the 13 years of the study. We believe that the power to detect an effect of sexual exposure to ejaculate is greatest during the earlier years of the MACS, when exposure was highest and the number of treatment-related confounders was lowest.
For this study, rapid, monotonic CD4 cell loss (as defined later in this paper) over three consecutive study visits was the measure of `failure'. At the time this study was initiated, CD4 cell count losses were the best indicator of HIV disease progression. Men who showed two successive large losses of CD4 cells measured at three consecutive semiannual follow-up visits were considered to have `failed'. Specifically, men who lost 10% or more of their CD4 cells between the first and second visits (of any three consecutive visits) and then showed a further 10% or greater loss between the second and third visits were classified as failures. The time point of a failure was recorded at the midpoint visit of a three-visit period of decline.
The number of days of continuous observation without failure or censoring formed the observation period. Men initially entered the risk set at their first complete follow-up visit to the MACS. Men who failed were not allowed to reenter the risk set. Each man contributed one period of continuous observation lasting four or more semiannual visits. Men who were present and had not failed at the tenth semiannual visit of the MACS were censored at that time.
Covariates of interest
Behavioral and demographic variables of interest included: (i) the number of self-reported RAI partners who had ejaculated into the participant's rectum (RAI-E partners) during the 12 months preceding each semiannual visit (during the last 12 months of observation); (ii) the number of RAI-E partners for the period excluding the 12 months preceding each semiannual visit (prior to the last 12 months of observation); (iii) the use of antiretroviral medications; (iv) the number of CD4 cells measured at the first visit of the observation period; (v) age; (vi) smoking; and (vii) the use of recreational drugs [i.e., marijuana and hashish (tetrahydrocannabinol), poppers (amyl nitrate), methylenedioxamphetamine (MDA), powder cocaine, downers, PCP (a dissociative hallucinogen also known as angel dust) and uppers (e.g., amphetamines other than MDA)].
The number of RAI-E partners reported during the last 12 months of observation and the number of RAI-E partners reported prior to the last 12 months of observation were treated as two distinct time-dependent covariates. The number of RAI-E partners reported during the last 12 months of observation were tallied and divided into four groups: none, 1 or 2, 3–7 and 8 or more. Because serially and exclusively monogamous men could not be distinguished from one another in these data, men with one or two (RAI-E) partners during the last 12 months of observation were grouped together. Men who reported three or more similar partners were grouped together based upon the frequency distribution of this characteristic. Similarly, the number of RAI-E partners reported for the periods prior to the last 12 months of observation were added together and divided into five groups: none (comparison), 1 to 2, 3 to 4, 5–9, 10–20, and 21 or more RAI-E partners.
To be identified as having failed (showing a monotonic loss of CD4 cells over three or more consecutive visits), study participants need to have been observed for at least the duration of the failure period. In essence, men needed to have been observed long enough to have had an opportunity to fail . Consequently, our study sample was limited to the 937 men who were observed for four or more consecutive semiannual visits .
Age (in years) at the first visit of the observation period, syphilis or gonorrhea infections, medication and recreational drug use and cigarette smoking were treated as confounders in these data. Syphilis infections were rare in these data and were, consequently, considered together with gonorrhea infections at each semiannual visit of the observation period. Men who reported using acyclovir (aciclovir) were compared with men who did not and those who used dideoxyribonucleoside analogs of cytidine (zalcitabine), thymidine (zidovudine), inosine (didanosine), or adenosine (stavudine) (antiretroviral medications) were grouped together and compared with men who reported no use. Acyclovir and antiretroviral medications were assessed at each semiannual visit and were treated as time-dependent covariates in our analyses. Men who reported recreational drug use of hashish or marijuana, poppers, MDA, cocaine downers, PCP and uppers were compared, respectively, with men who reported no use of these substances at each of the semiannual visits we evaluated. Also, cigarette smokers were compared with non-smokers at each semiannual visit of the observation period. The CD4 cell count measured at the first visit of an observation period was used to group (HIV-seropositive) men together who, initially, may have been at similar stages of HIV disease: > 850 (comparison), 500–849, 350–499, 200–349 and < 200 × 106 cells/l.
Descriptive statistics and tabular analyses were used to explore the data. Partial-likelihood Cox proportional hazards regression (SAS and BMDP) was used to obtain unadjusted and adjusted estimators. Because the model coefficients are easily transformed into hazard rate ratios reflective of short-term risk for rapid CD4 cell loss, the hazard rate ratio was used to compare groups of men in this analysis. Model fits were evaluated using deviance statistics [33,34].
Although the number of RAI-E partners reported during and prior to the 12 months preceding each semiannual visit were evaluated as categorical variables, the data for a trend was evaluated in two ways. First, the likelihood ratio test statistic that was derived from entering the number of RAI-E partners reported during the last 12 months of observation into the multivariate model was evaluated as a continuous variable. In addition, the likelihood ratio test was evaluated for an ordinal variable reflecting the four categories of exposure (1, 1, 2, 3–7, or 8 or more). The number of RAI-E partners reported prior to the last 12 months of observation were assessed in the same manner.
Data were analyzed from 937 men with four or more consecutive semiannual MACS visits. Rapid CD4 cell loss was not a rare occurrence in this study group (42%; 389/937). Men who failed showed a substantial decrease in their absolute number of CD4 cells over three consecutive visits. Using the absolute number of CD4 cells, the mean and median percentage of CD4 cells lost during the three-visit failure period was 51% and 48%, respectively. Similarly, the mean and median number of cells lost during the three-visit failure period was 297 and 331 × 106 cells/l, respectively.
The frequency of RAI that was reported by men in the larger MACS cohort and by men in our study group changed during the first several years of follow-up. The frequency of RAI partners reported by all men who were found to be infected with HIV at the first MACS visit declined sharply between visits one and three and remained low through visit 10. For these men, the mean number (standard deviation) of RAI partners reported were 7.7 (18.4), 5.1 (13.0), 2.9 (7.2), 1.9 (4.3), 2.0 (4.8), 1.8 (4.3), 1.8 (5.1), 1.8 (6.3), 1.7 (4.4) and 2.2 (7.9) for visits 1 through 10, respectively. Similar temporal patterns were observed for the mean number (standard deviation) of RAI-E partners reported by the men we studied for the 12-month period preceding each visit for visits two to ten (i.e., 9.7 (22.9), 5.5 (14.6), 3.0 (7.7), 2.0 (5.7), 1.8 (6.5), 1.3 (4.0), 0.8 (3.1), 0.5 (3.9) and 0.7 (6.2), respectively). Although there is no significant differences in the average number of RAI-E partners reported by the men we studied and HIV-seropositive men overall, differences in the visit-specific averages likely contributed to intermittent losses to follow-up that disqualified some men from inclusion in our study sample.
Most men (75%; 706/937) reported having no RAI-E partners for the 12 months preceding at least one of the semiannual visits and a few (13%; 118/937) reported no RAI-E partners in the 12 months preceding each of the visits two to ten. Conversely, more than half (52%; 57/937) of the men reported having three or more RAI-E partners during at least one of the 12-month periods preceding a semiannual MACS visit; very few men (7%; 68/937) reported having three or more RAI-E partners during the 12 months preceding each of their semiannual visits.
Most men (59%; 551/937) began their observation periods with > 500 × 106 cells/l (Table 1). Approximately one-quarter of these men reported one or more syphilis or gonorrhea infections during their observation period and a few reported using acyclovir, zalcitabine, zidovudine, didanosine or stavudine. Forty-three percent of the men reported ever smoking cigarettes during the study period (Table 1).
The final Cox-proportional hazards model included parameters estimating the effect of the number of RAI-E partners reported during and prior to the last 12 months of observation; age; smoking; recreational drug use; syphilis or gonorrhea infection; use of acyclovir, zalcitabine, zidovudine, didanosine or stavudine; and the CD4 cell count measured at the first visit of the observation period. The number of RAI-E partners that were reported prior to and during the last 12 months of observation, use of acyclovir and antiretroviral medications, recreational drug use, syphilis or gonorrhea infections and cigarette smoking were evaluated as time-dependent covariates. Sparse data in some strata may have contributed to some of the imprecision in the estimates. For example, although 278 (30%) of the 937 men reported eight or more RAI-E partners at one or more visits during their observation period, only 11 these 278 (4%) reported eight or more (RAI-E) partners after 4 or more years of continuous observation. Cox-proportional hazards models that included other time-dependent covariates  and covariate product terms were evaluated using the deviance statistic and were rejected using a 0.05 level of significance. The Cox-proportional hazards models that individually examined the relations between each covariate and rapid CD4 cell loss differed little from the findings of the multivariate model (Table 2).
The risk of rapid CD4 cell loss was more than double for men who reported one or more RAI-E partners during the last 12 months of observation as assessed by the multivariate model (Table 2). The point estimates, together with the tests for trend, imply that a greater number of RAI-E partners during the last 12 months of observation may be associated with an escalating risk of CD4 cell loss (Table 2). When the data were evaluated for trend, there was evidence for the logit to increase linearly with the number of RAI-E partners that were reported during the last 12 months of observation (χ2 1df = 26.1;P < 0.0001 when categories of exposure were evaluated as an ordinal variable and χ2 1df = 3.5;P < 0.07 when the number of partners was evaluated as a continuous variable). The relations between RAI-E partnerships during the 12 months preceding each semiannual visit persisted when the effects of other covariates in the multivariate analysis were simultaneously controlled (Table 2).
When all other covariates were controlled for, men who reported one to two RAI-E partners prior to the last 12 months of observation were at no greater risk for rapid CD4 cell loss than men who reported no such partners. However, men who reported three or more RAI-E partners prior to the last 12 months of observation appeared to be at lower risk for a period of rapid CD4 cell loss than men who reported no similar partners during that period (Table 2). As observation time increased, sparse data may have made the estimates less precise; the point estimates, together with the test for trend, imply a dose–response relationship between RAI-E partners reported prior to the last 12 months of observation and failure. Again, evaluating the data for trend, there was evidence for the logit to increase linearly with the number of RAI-E partners that were reported prior to the last 12 months of observation (i.e., χ2 1df = 15.7;P = 0.0001 when categories of exposure were evaluated as an ordinal variable and χ2 1df = 10.1;P < 0.002 when the number of partners was evaluated as a continuous variable).
We found that several other factors influenced the risk of rapid CD4 cell loss. Men who reported use of acyclovir or antiretroviral drugs were at lower risk for a period of rapid CD4 cell loss than men who did not use these drugs (Table 2). Although men who began the observation period with < 850 × 106 cells/l CD4 cells appeared to be at slightly greater risk for rapid CD4 cell loss than men who began with more, the confidence intervals and the tests for trend did not support this precise relationship when the effect of other covariates were controlled (the categories and the number of CD4 cells measured at the beginning of the risk period were evaluated as ordinal and continuous variables, respectively: χ2 1df = 1.1;P = 0.3 and χ2 1df = 1.5;P = 0.2, respectively;Table 2). The findings also suggest that syphilis and gonorrhea infections may increase the risk of rapid CD4 cell loss (Table 2).
There was no statistical association between age or smoking and rapid CD4 cell loss in the study sample. For each year of age, the hazard rate ratio changed little [i.e., hazard rate ratio: 1.003 (0.98–1.02)]. For example, the risk of rapid CD4 cell loss was 3% greater for every additional 10 years of life lived for men in this study group. Also, there was no greater risk for rapid CD4 cell loss associated with the various recreational drugs that were reported by these men (Table 2).
These data suggest that exposure to ejaculate after primary HIV infection increases the risk of rapid, CD4 cell loss, a commonly used clinical marker of HIV disease progression. Men in our study who reported one or more RAI-E partners during the last 12 months of observation were more than twice as likely to show a period of rapid CD4 cell loss than men who reported no such partners. Our findings suggest a dose–response relationship between the number of RAI-E partners and rapid CD4 cell loss. These relations persisted after we controlled in our multivariate analysis for the effects of RAI-E partnerships reported prior to the last 12 months of observation, syphilis or gonorrhea infection, recreational drug use, the use of antiretroviral medications, the number of CD4 cells measured at the first visit of the observation period and smoking.
The data suggest that some men are resistant to the effects ejaculate exposure has upon the risk of rapid CD4 cell loss. Men who reported three or more RAI-E partners prior to the last 12 months of observation were at lower risk for rapid CD4 cell loss than men who reported none. This finding may result from the innate characteristics of more resistant men (e.g., major histocompatibility complex/HLA characteristics), phenotypic or genotypic differences in their infective virus population or environmental differences (e.g., different unconditional probabilities of exposure to a sexually transmitted disease within the sexual networks these men are exposed to) that remain unmeasured in these data.
For the most part, the effects of recreational drug use did not strongly influence the risk of rapid CD4 cell loss in this study group. Although in vitro and animal studies suggest that amphetamine (MDA and other uppers), marijuana, hashish and PCP use should alter, and may specifically suppress, immune function, our findings did not support this premise. We found that men who reported using poppers, marijuana, cocaine, MDA, PCP, uppers and downers showed no statistically significant greater risk for rapid CD4 cell loss than men who reported no use of these substances. It is possible that rapid CD4 cell loss may be relatively insensitive to the immunomodulating effects of these recreational drugs.
It is highly possible that ejaculate exposure is not solely responsible for the rapid CD4 cell losses we observed. There are indications from observational studies that exposure to infectious diseases may be an underlying mechanism for the rapid CD4 cell loss. Unprotected anoreceptive intercourse is positively associated with both the acquisition of sexually transmitted diseases and the risk of developing AIDS after (primary) HIV infection [1,35,36]. Unprotected sexual relations with an AIDS-diagnosed partner seems to confer greater risk for developing of AIDS than such behaviors with HIV-infected partners who are not diagnosed with AIDS . The greater risk for bacterial and viral infections that is associated with all stages of HIV infection and the more common occurrence of opportunistic infections among those with < 200 × 106 cells/l CD4 cells infers that overall susceptibility to infection is amplified and that this vulnerability increases as (HIV) disease progresses [38–40].
Our data only allowed the effects of two sexually transmitted diseases, syphilis and gonorrhea infections, upon the risk of rapid CD4 cell loss to be examined. When we simultaneously controlled for the effects of RAI-E partners and these diseases, we found that, although positive associations persisted, our data could not precisely support these conclusions.
In vitro and in vivo studies also support our hypothesis that acute infectious diseases may cause these periods of rapid CD4 cell loss. HIV more often completes infection when CD4 cells are activated than when they remain in a quiescent state, suggesting that factors that amplify immunologic activation might also enhance cell-to-cell transmission of the endogenous infection . Both mitogens and infectious agents can enhance HIV replication rates in vitro , and replication in CD4 cells appears to be dependent upon the differentiation stage of the cells that are infected, the patient's disease stage and the type of activation signal used to stimulate the cells . In vivo, CD4 cell counts are inversely related to HIV viral replication rates and viral burden [24–26]. Conversely, HIV virus replication rates are positively associated with exposure to immunogens in vitro [12,13].
HIV replication rates and the turnover rate of virus-producing cells appear both rapid and continuous [21,41]. The capability individuals have to replenish CD4 cells may be more strained when HIV replication rates are high . Possibly, our findings extend Ho's analogy that likens clinically stable HIV infection to a partially filled sink with the tap and drain both equally wide open . Our analysis suggests that if acute infections induce bursts of viral replication, the subsequent stress placed upon the immune system may temporarily increase the drain of CD4 cells beyond the immediate regenerative capabilities of the immune system. Subsequently, though the tap and drain may again return to an equilibrium, the sink's volume of circulating CD4 cells may be substantially reduced.
We believe that infectious agents stimulate CD4 cells to increase production of HIV, which is responsible for these periods of rapid CD4 cell loss, but we are unable to analyze the majority of the sexually transmitted diseases that may be responsible for the underlying biological relations from these data (e.g., cytomegalovirus exposures). Similarly, the effect HIV virus load , ejaculatory fluids and semen from a diverse number of partners, chemokine receptor genotype [43,44] and viral phenotype have upon the relationship between ejaculate exposure and rapid CD4 cell loss cannot be clarified by our data. We recognize that there may be other residual behavioral or biological characteristics that confound our analyses and that their effect upon the risk for rapid CD4 cell loss cannot be estimated from these data.
We did not control for the effects of AIDS-defining conditions or death in our analyses. Our theoretical framework has always posited AIDS-defining conditions as either intermediates in the pathway between RAI-E partnerships and rapid CD4 cell loss or as outcomes related to the absolute number of CD4 cells that result from periods of rapid (CD4 cell) loss. Were we to control for the effects of an intermediate in the biological pathway, our estimates of the effect of RAI-E partnerships upon rapid CD4 cell loss would be spurious. Additionally, if AIDS-defining conditions are a manifestation that results from the absolute number of CD4 cells that remain after periods of rapid loss, then the (AIDS-defining) conditions do not precede our outcome of interest. Similarly, death may have occurred for some men after the period of rapid CD4 cell loss and would not be an antecedent to rapid CD4 cell loss. The factors that led to our initiating this investigation in 1992 were anecdotal reports by clinicians that they would intervene, often more vigorously, when their patients displayed monotonic periods of rapid CD4 cell loss. For these reasons, we did not include either death or AIDS-defining conditions in our analyses.
Our findings strongly suggest that behaviors do influence the clinical progression of HIV disease. These relations persisted even when we controlled for the effects of other covariates, i.e., antiretroviral medications, smoking, syphilis or gonorrhea infections, recreational drug use and CD4 cell count at the first visit of the observation period. With the exception of smoking, our findings are generally consistent with those of other investigators [45–50]. Differences between our observations and those of others may be related to the infrequency of specific exposures that occurred during our study period (e.g., use of both acyclovir and antiretroviral drugs) and to differences between our outcome and those measured by other teams (e.g., death, leukocytosis, < 200 × 106 cells/l CD4 cells, AIDS-defining conditions) [45–52]. Exposure to ejaculate, and possibly to infectious diseases that may be associated with this behavior, appears to induce rapid CD4 cell loss and to contribute to HIV disease progression. Therefore, HIV-infected individuals should be cautioned against unprotected anoreceptive intercourse.
The authors would like to thank Robert Wiley for his assistance in preparing this manuscript.
1. Phair J, Jacobson L, Detels R. et al
. Acquired immune deficiency syndrome occurring within 5 years of infection with human immunodeficiency virus type-1: the Multicenter AIDS Cohort Study. J Acquir Imm Defic Syndr 1992, 5: 490–496.
2. Gallo RC, Salahuddin SZ, Popovic M. et al
. Frequent detection and isolation of cytopathic retroviruses (HTLV-III) from patients with AIDS and at risk of AIDS. Science 1984, 224: 500–503.
3. Barre-Sinoussi F, Chermann JC, Rey F. et al
. Isolation of a T-lymphotropic retrovirus from a patient at risk for acquired immune deficiency syndrome (AIDS). Science 1983, 220: 868–871.
4. Munoz A, Schrager LK, Bacellar H. et al
. Trends in the incidence of outcomes defining acquired immunodeficiency syndrome (AIDS) in the Multicenter AIDS Cohort Study: 1985–1991. Am J Epidemiol 1993, 137: 423–438.
5. Detels R, English PA, Giorgi JV. et al
. Patterns of CD4 cell changes after HIV-1 infection indicate the existence of a codeterminant of AIDS. J Acquir Imm Defic Syndr 1988, 1: 390-395.390-395.
6. Liu Z, Cumberland WG, Hultin LE. et al
. Elevated CD38 antigen expression on CD8+ T cells is a stronger marker for the risk of chronic HIV disease progression to AIDS and death in the Multicenter AIDS Cohort Study than CD4+ cell count, soluble immune activation markers, or combinations of HLA-DR and CD38 expression. J Acquir Immune Defic Syndr Hum Retrovirol 1997, 16: 83–92.
7. Kostrikis LG, Huang Y, Moore JP. et al
. A chemokine receptor CCR2 allele delays HIV-1 disease progression and is associated with a CCR5 promoter mutation. Nat Med 1998, 4: 350–353.
8. Morawetz RA, Rizzardi GP, Glauser D. et al
. Genetic polymorphism of CCR5 gene and HIV disease: the heterozygous ( CCR5/delta ccr5 ) genotype is neither essential nor sufficient for protection against disease progression. Swiss HIV Cohort.
Eur J Immunol 1997, 27: 3223–3227.
9. Misrahi M, Teglas JP, N NG. et al
. CCR5 chemokine receptor variant in HIV-1 mother-to-child transmission and disease progression in children. French Pediatric HIV Infection Study Group.
J Am Med Asssoc 1998, 279: 277–280.
10. Update on acquired immune deficiency syndrome (AIDS) – United States. MMWR
11. Immunodeficiency among female sexual partners of males with acquired immune deficiency syndrome (AIDS) – New York. MMWR
12. Zack JA, Cann AJ, Lugo JP. et al
. HIV-1 production from infected peripheral blood T cells after HTLV-I induced mitogenic stimulation. Science 1988, 240: 1026–1029.
13. Zack JA, Arrigo SJ, Weitsman SR. et al
. HIV-1 entry into quiescent primary lymphocytes: molecular analysis reveals a labile, latent viral structure. Cell 1990, 61: 213–222.
14. Blanchard A, Montagnier L, Gougeon ML. Influence of microbial infections on the progression of HIV disease. Trends Microbiol 1997, 5: 326–331.
15. Hofmann B, Lindhardt BO, Gerstoft J. et al
. Lymphocyte transformation response to pokeweed mitogen as a predictive marker for development of AIDS and AIDS related symptoms in homosexual men with HIV antibodies. Br Med J 1987, 295: 293–296.
16. Hofmann B, Jakobsen KD, Odum N. et al
. HIV-induced immunodeficiency. Relatively preserved phytohemagglutinin as opposed to decreased pokeweed mitogen responses may be due to possibly preserved responses via CD2/phytohemagglutinin pathway.
J Immunol 1989, 142: 1874–1880.
17. Linette GP, Hartzman RJ, Ledbetter JA. et al
. HIV-1-infected T cells show a selective signaling defect after perturbation of CD3/antigen receptor. Science 1988, 241: 573–576.
18. Hofmann B, Moller J, Langhoff E. et al
. Stimulation of AIDS lymphocytes with calcium ionophore (A23187) and phorbol ester (PMA): studies of cytoplasmic free Ca, IL-2 receptor expression, IL-2 production, and proliferation. Cell. Immunol 1989, 119: 14–21.
19. Fahey JL, Taylor JM, Detels R. et al
. The prognostic value of cellular and serologic markers in infection with human immunodeficiency virus type 1. New Engl J Med 1990, 322: 166–172
20. Schnittman SM, Lane HC, Greenhouse J. et al
. Preferential infection of CD4 memory T cells by human immunodeficiency virus type 1: evidence for a role in the selective T-cell functional defects observed in infected individuals. Proc Natl Acad Sci USA. 1990, 87: 6058–6062.
21. Ho DD, Neumann AU, Perelson AS. et al
. Rapid turnover of plasma virions and CD4 lymphocytes in HIV-1 infection. Nature 1995, 373: 123–126.
22. Cayota A, Vuillier F, Scott-Algara D. et al
. Preferential replication of HIV-1 in memory CD4 subpopulation
[letter]. Lancet 1990, 336: 941.941.
23. Cayota A, Vuillier F, Scott-Algara D. et al
. Differential requirements for HIV-1 replication in naive and memory CD4 T cells from asymptomatic HIV-1 seropositive carriers and AIDS patients. Clin Exp Immunol 1993, 91: 241–248.
24. Saag MS, Crain MJ, Decker WD. et al
. High-level viremia in adults and children infected with human immunodeficiency virus: relation to disease stage and CD4 lymphocyte levels. J Infect Dis 1991, 164: 72–80.
25. Saag MS, Emini EA, Laskin OL. et al
. A short-term clinical evaluation of L-697,661, a non-nucleoside inhibitor of HIV-1 reverse transcriptase. L-697,661 Working Group.
N Engl J Med 1993, 329: 1065–1072.
26. Molina JM, Ferchal F, Chevret S. et al
. Quantification of HIV-1 virus load under zidovudine therapy in patients with symptomatic HIV infection: relation to disease progression. AIDS 1994, 8: 27–33.
27. Mellors JW, Rinaldo CR, Gupta P et al. Prognosis in HIV-1 infection predicted by the quantity of virus in plasma Science
:1167–1170. [Published erratum appears in Science 1997, 275
28. Mellors JW, Munoz A, Giorgi JV. et al
. Plasma viral load and CD4 lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med 1997, 126: 946–954.
29. Leach CT, Cherry JD, English PA. et al
. The relationship between T-cell levels and CMV infection in asymptomatic HIV-1 antibody-positive homosexual men. J Acquir Imm Defici Syndr 1993, 6: 407–413.
30. Castro BA, Homsy J, Lennette E. et al
. HIV-1 expression in chimpanzees can be activated by CD8+ cell depletion or CMV infection. Clin Immunol Immunopathol 1992, 65: 227–233.
31. Kaslow RA, Ostrow DG, Detels R. et al
. The Multicenter AIDS Cohort Study: rationale, organization, and selected characteristics of the participants. Am J Epidemiol 1987, 126: 310–318.
32. Anderson JR, Cain KC, Gelber RD. et al
. Analysis and interpretation of the comparison of survival by treatment outcome variables in cancer clinical trials. Canc Treat Rep 1985, 69: 1139–1146.
33. Dixon W. BMDP Statistical Software Manual.
Berkeley: University of California Press; 1992.
34. Hosmer DW, Lemeshow S. Applied Logistic Regression.
New York: Wiley; 1989.
35. Ostrow DG, Obermaier A. Sexual Practices History.
New York: Plenum Medical; 1983.
36. Eskild A, Samdal HH, Heger B. Co-infection with HIV-1/HTLV-II and the risk of progression to AIDS and death. The Oslo HIV Cohort Study Group.
APMIS 1996, 104: 666–672.
37. Carre N, Meyer L, Boufassa F. et al
. High risk of HIV disease progression after infection through a sexual partner with AIDS. AIDS 1996, 10: 77–80.
38. Holmberg SD, Buchbinder SP, Conley LJ et al
. The spectrum of medical conditions and symptoms before acquired immunodeficiency syndrome in homosexual and bisexual men infected with the human immunodeficiency virus. Am J Epidemiol
:395–404. [Discussion 405–406.]
39. Crowe SM, Carlin JB, Stewart KI. et al
. Predictive value of CD4 lymphocyte numbers for the development of opportunistic infections and malignancies in HIV-infected persons. J Acquir Imm Defici Syndr 1991, 4: 770–776.
40. Masur H, Ognibene FP, Yarchoan R. et al
. CD4 counts as predictors of opportunistic pneumonias in human immunodeficiency virus (HIV) infection. Ann Inter Med 1989, 111: 223–231.
41. Wei X, Ghosh SK, Taylor ME. et al
. Viral dynamics in human immunodeficiency virus type 1 infection. Nature 1995, 373: 117–122.
42. Schnittman SM, Greenhouse JJ, Psallidopoulos MC. et al
. Increasing viral burden in CD4 T cells from patients with human immunodeficiency virus (HIV) infection reflects rapidly progressive immunosuppression and clinical disease. Ann Inter Med 1990, 113: 438–443.
43. Huang Y, Paxton WA, Wolinsky SM. et al
. The role of a mutant CCR5 allele in HIV-1 transmission and disease progression. Nat Med 1996, 2: 1240–1243.
44. Michael NL, Louie LG, Rohrbaugh AL. et al
. The role of CCR5 and CCR2 polymorphisms in HIV-1 transmission and disease progression. Nat Med 1997, 3: 1160–1162.
45. Conley LJ, Bush TJ, Buchbinder SP. et al
. The association between cigarette smoking and selected HIV-related medical conditions. AIDS 1996, 10: 1121–1126.
46. Galai N, Park LP, Wesch J. et al
. Effect of smoking on the clinical progression of HIV-1 infection. J Acquir Imm Defic Syndr Hum Retrovirol 1997, 14: 451–458.
47. Burns DN, Hillman D, Neaton JD. et al
. Cigarette smoking, bacterial pneumonia, and other clinical outcomes in HIV-1 infection. Terry Beirn Community Programs for Clinical Research on AIDS.
J AIDS Hum Retrovirol 1996, 13: 374–383.
48. Nieman RB, Fleming J, Coker RJ. et al
. The effect of cigarette smoking on the development of AIDS in HIV-1-seropositive individuals. AIDS 1993, 7: 705–710.
49. Park LP, Margolick JB, Giorgi JV. et al
. Influence of HIV-1 infection and cigarette smoking on leukocyte profiles in homosexual men. The Multicenter AIDS Cohort Study.
J Acquir Imm Defici Syndr 1992, 5: 1124–1130.
50. Royce RA, Winkelstein W. HIV infection, cigarette smoking and CD4 T-lymphocyte counts: preliminary results from the San Francisco Men's Health Study. AIDS
51. Stein DS, Graham NM, Park LP. et al
. Acyclovir in human immunodeficiency virus patients. J Infect Dis 1996, 173: 504–507.
52. Youle MS, Gazzard BG, Johnson MA et al. Effects of high-dose oral acyclovir on herpesvirus disease and survival in patients with advanced HIV disease: a double-blind, placebo-controlled study. European-Australian Acyclovir Study Group. AIDS 8
: 641–649. [Published erratum appears in AIDS
1994,6: following 859.]