Abstract

This study used a well characterized cohort of human immunodeficiency virus type 1 (HIV‐1)–infected hemophiliacs to define the relationship between the SDF1‐3′A allele, the plasma HIV‐1 coreceptor tropism, and the natural history of HIV‐1 disease. Subjects heterozygous or homozygous for the SDF1‐3′A allele experienced higher rates of decline in CD4+ T cell counts over time than did those without the allele (P=.009). Moreover, they had an increased risk of progression to acquired immunodeficiency syndrome and death, a relationship that persisted even when baseline plasma HIV‐1 RNA levels and CD4+ T cell counts or CCR5Δ32 and CCR2‐64I genotype were controlled for. This relationship was even stronger in a subgroup of subjects for whom tropism data were available. Subjects with the SDF1‐3′A allele were also more likely to have detectable X4‐tropic viruses (P=.012), and, when tropism was included in the survival analyses, the effect of the SDF1‐3′A allele on disease progression was no longer significant. Therefore, the increased frequency of X4‐tropic viruses in subjects carrying the SDF1‐3′A allele may explain the observed adverse effect that this allele has on the natural history of HIV‐1 disease.

The natural history of HIV‐1 disease is heterogeneous and influenced by numerous virologic, immunologic, and host factors. HIV‐1 infection is usually initiated by viruses that use the CCR5 coreceptor, referred to (on the basis of their biological phenotype) as non–syncytium‐inducing viruses or (on the basis of their coreceptor tropism) as R5 viruses [1, 2]. During the course of infection, in approximately one‐half of subjects, syncytium‐inducing (SI) viruses—strains that use the CXCR4 coreceptor and are known as X4 viruses—will develop [1]. Although it is difficult to prove cause and effect, the emergence of SI or X4 viruses has been associated with both a precipitous decline in CD4+ T cell counts and accelerated disease progression [3–7]. Little is known about why SI or X4 viruses develop in only a subset of individuals and what factors may contribute to the emergence of these strains.

During the last decade, there has been interest in the relationship between genetic factors and HIV‐1 disease progression, which was recently summarized by O’Brien and Nelson [8]. Large deletions in CCR5 (CCR5Δ32), a multisite allele of the CCR5 promoter (CCR5 P1), and a mutation in CCR2, (CCR2‐64I) [9–12] have been associated with variable effects on HIV‐1 disease progression. Homozygosity for CCR5Δ32 provides near complete protection from infection [13], and heterozygosity has been associated with delayed disease progression. In contrast, CCR5 P1 has been associated with accelerated disease progression in the absence of either CCR5Δ32 or CCR2‐64I [10]. Several studies have demonstrated that the CCR2‐64I allele is also associated with delayed disease progression [11, 12], possibly as a result of the interaction between CCR2 and the CXCR4 receptor, resulting in reduced risk of development of SI or X4 viruses [14].

A polymorphism in the 3′ untranslated region of the gene encoding stromal cell–derived factor–1 (SDF‐1), also known as CXCL12, is present at a variable frequency in different populations [15, 16]. SDF‐1 is the natural ligand for CXCR4, and results of studies of the relationship between the presence of the SDF1‐3′A allele and susceptibility to HIV‐1 infection have varied: the largest study showed no protection [16], and another suggested a reduced risk of HIV‐1 infection in Thai subjects [15]. Similarly, the effects on disease progression have been conflicting: some studies suggested that it is protective [16], others showed accelerated disease progression [17–20], and yet others did not detect any effect of the SDF1‐3′A allele on the natural history of HIV‐1 disease [21–23]. Finally, the effect that this polymorphism has on SDF‐1 levels is not known: one early study proposed that levels may be increased [16], and another actually demonstrated lower levels in individuals homozygous for the SDF1‐3′A allele [24]. The present study further evaluates the relationship between this polymorphism and plasma HIV‐1 RNA levels, CD4+ T cell counts, coreceptor tropism, and HIV‐1 disease progression.

Subjects and Methods

Study Population

The Hemophilia Growth and Development Study (HGDS) is a multicenter study in the United States that, between 1989 and 1990, enrolled a population‐based cohort of 6–19‐year‐olds with hemophilia. The cohort included 207 HIV‐1–infected study subjects who acquired HIV‐1 infection through exposure to blood products, in most cases between 1982 and 1983, 6–8 years before study enrollment. Participants were followed for 7 years, up to ∼15 years from the time of infection. The details of recruitment and the characteristics of this cohort have been reported elsewhere [25, 26]. The human‐subjects committees of collaborating institutions approved the HGDS, informed consent was obtained from parents or legal guardians, and informed consent or assent was obtained from all participants, in compliance with the human‐experimentation guidelines of the US Department of Health and Human Services.

The ethnic composition of the cohort—72% European American, 15% Hispanic, 11% African American, and 2% other—resembled that of the general hemophiliac population in the United States [27]. During the 7 years of follow‐up, 0.48% of the HIV‐1–infected cohort was lost to vital status follow‐up. Although the study did not enroll individuals who died before recruitment, there is little other evidence of an introduced selection bias. At baseline and 6‐month intervals, an assessment of medical history, physical examinations, and CD4+ T cell counts were performed. Blood samples were processed within 24 h, for cryopreservation of cells, plasma, and serum. Antiretroviral therapy during the course of follow‐up was prescribed at the discretion of the primary providers. Since only 9 subjects included in the analysis had been exposed to protease inhibitor–containing triple‐drug therapy, and then only 3 had been exposed for >6 months, the effect of treatment on disease progression was minimal in this cohort.

Plasma HIV‐1 RNA Level, SDF1 Genotype, and Coreceptor Tropism

Plasma HIV‐1 RNA levels were measured at baseline and annually, from stored specimens, at a central laboratory by use of the branched DNA assay (version 2.0; Versant HIV‐1 RNA; Bayer Healthcare Diagnostics). Samples with undetectable levels of virus (<500 copies/mL) were retested by use of version 3.0 (lower limit of detection, 50 copies/mL) [28]. The SDF1‐3′A allele was detected by use of a polymerase chain reaction (PCR) restriction fragment–length polymorphism assay that has been described elsewhere [16]. Presence of the SDF1 genotype was determined for 204 of the 207 HIV‐1–infected subjects in the cohort; results were not available for the remaining subjects, because of a lack of host DNA.

Plasma samples obtained closest to baseline, with HIV‐1 RNA levels >500 copies/mL, were sent to ViroLogic for analysis of coreceptor tropism. The HIV‐1 envelope coreceptor tropism assay was developed by modifying the phenotypic drug susceptibility assay initially developed for protease and reverse‐transcriptase (RT) inhibitors [29, 30]. Pseudotyped HIV‐1 viruses were generated by use of a genomic vector carrying a luciferase reporter gene together with expression vectors containing patient virus envelope genes amplified (by RT‐PCR) from plasma HIV‐1 RNA. Coreceptor tropism is defined as the ability of the recombinant viruses to infect U87 cells that have been engineered to express CD4 with either CCR5 or CXCR4. CCR5‐tropic (R5), CXCR4‐tropic (X4), and mixed or dual tropic (R5/X4) designations were verified by blockade of coreceptor‐mediated infection by use of specific antagonists, with no distinction made between dual and mixed tropism.

Coreceptor tropism data were available for 126 of the 207 subjects in the cohort, with all but 1 having SDF1 genotype data available as well (n=125). Results of tropism studies were not available for the remainder of subjects, for a variety of reasons—including plasma HIV‐1 RNA levels <500 copies/mL (n=23), below which the tropism assay cannot be performed; inadequate stored specimens (n=18); and the inability to amplify the envelope gene, for technical reasons, such as degraded RNA template leading to failed first‐strand synthesis during RT‐PCR, failure of primer annealing because of a mutation in one of the primer binding sites, and generation of an internal restriction site in the amplicon that subsequently gets cleaved during the digestion step of the assay (n=40). The tropism assay was performed on samples obtained as close to the date when the baseline specimen was obtained, as availability allowed; these samples were obtained during the first 12 months of enrollment for 79% of the subjects for whom tropism data were available.

Study Variables

CD4+ T cell counts were measured every 6 months, and plasma HIV‐1 RNA levels were measured annually. The baseline plasma HIV‐1 RNA level represents the value at the time of the first available plasma sample from time of entry into the cohort. Participants were categorized as having progressed to AIDS if they met the 1987 Centers for Disease Control definition [31]. Of the 204 subjects for whom SDF1 genotype data were available, 20 (9.8%) had an AIDS‐defining condition at entry and were not included in the AIDS survival analysis. Ten (8.0%) of the 125 subjects for whom coreceptor tropism and SDF1 genotype data were available had received a diagnosis of AIDS at the time of enrollment and were similarly not included in the analysis.

Statistical Analysis

SDF1genotype, plasma HIV‐1 RNA level, CD4+T cell count, and coreceptor tropism. Since there were few subjects homozygous for SDF1‐3′A, analyses compared subjects with wild‐type (wt) SDF1 with subjects either heterozygous or homozygous for SDF1‐3′A. The repeated measurements of plasma HIV‐1 RNA levels and CD4+ T cell counts were modeled by use of random‐coefficient regression models [32]. Each subject’s plasma HIV‐1 RNA level and CD4+ T cell count were modeled as linear functions of the time‐varying age variable, with each subject having a different intercept and slope. The relationship between SDF1 genotype and baseline plasma HIV‐1 RNA level and CD4+ T cell counts, as well as plasma HIV‐1 RNA level and CD4+ T cell count rate of change, was examined by use of approximate F tests in the regression models. In all the models, plasma HIV‐1 RNA levels were log10 transformed, and absolute CD4+ T cell counts were square‐root transformed, to better comply with the model assumptions. In the subset of subjects for whom tropism data were available, Fisher’s exact test was used to assess the relationship between SDF1 genotype and the presence of X4 viruses.

SDF1genotype and clinical progression. Cox proportional hazards models were used to examine the effects of SDF1 genotype on progression to clinical AIDS and death in the presence of AIDS [31]. Both unadjusted and adjusted effects were assessed to examine the value of predicting survival on the basis of SDF1 genotype alone and after baseline plasma HIV‐1 RNA levels and CD4+ T cell counts were controlled for, as well as for the presence of the CCR5Δ32 and CCR2‐64I mutations, which have previously been shown to be related to disease progression in the cohort, and coreceptor tropism [26]. Kaplan‐Meier curves were also plotted, to provide a graphical comparison of time to progression to AIDS and death, by SDF1 genotype.

Results

The baseline characteristics of the subset of subjects for whom coreceptor tropism data were available were similar to those of the overall cohort (table 1), except that subjects with tropism, compared with those without tropism, had lower CD4+ T cell counts (P=.28) and higher plasma HIV‐1 RNA levels (P<.001). This is consistent with the fact that tropism data were often not available because plasma HIV‐1 RNA levels were too low for successful amplification in the tropism assay. Overall, the SDF1‐3′A allele frequency was 18.4%, with 7 subjects homozygous and 61 subjects heterozygous for this variant. Compared with the overall cohort, the proportion of subjects homozygous and heterozygous and the allele frequency were essentially the same in the subgroup for whom tropism data were available, as was the use of antiretroviral therapy. In the subset of subjects tested for plasma HIV‐1 coreceptor tropism for whom SDF1 genotype results were also available, there were 50 (40%) with R5/X4 viruses, 75 (60%) with only R5 viruses, and none with only X4 viruses. There was an increased frequency (56%) of R5/X4 viruses in subjects with the SDF1‐3′A allele, compared with those without it (32%) (P=.012).

Table 1. 

Baseline characteristics of subjects from the Hemophilia Growth and Development Study (HGDS) cohort for whom SDF1 genotype data were available and those for whom both SDF1 genotype and coreceptor tropism data were available.

Table 1. 

Baseline characteristics of subjects from the Hemophilia Growth and Development Study (HGDS) cohort for whom SDF1 genotype data were available and those for whom both SDF1 genotype and coreceptor tropism data were available.

The relationship between the presence of the SDF1‐3′A allele and CD4+ T cell counts and plasma HIV‐1 RNA levels at baseline and change over time was assessed for the entire cohort, as well as for the subset for whom coreceptor tropism data were available (table 2). For the subset of subjects for whom tropism data were available, those with and without the variant allele were no different with regard to baseline CD4+ T cell counts or antiretroviral use. However, there was a trend toward higher baseline plasma HIV‐1 RNA levels in subjects carrying the SDF1‐3′A allele (P=.096), albeit with no significant difference in the change in this measure over time. In contrast, there was a significantly greater rate of decline in CD4+ T cell counts in subjects with the polymorphism (P=.009). Similar relationships for decline in CD4+ T cell counts were seen in the subjects for whom coreceptor tropism data were available (P=.006); among these subjects, baseline plasma HIV‐1 RNA levels were significantly higher in those with the SDF1‐3′A allele than in the those with wt SDF1 (P=.023) (table 2). When the model was adjusted for tropism in these subjects, there was a marked reduction in the significance of the relationship between the presence of the SDF1‐3′A allele and baseline plasma HIV‐1 RNA level (P=.22), as well as for decline in CD4+ T cell count, with the P value decreasing from .006 to .045, which was seen even when antiretroviral use was controlled for or when only heterozygous subjects were analyzed.

Table 2. 

Baseline and change in CD4+ T cell counts and plasma HIV‐1 RNA levels in the entire Hemophilia Growth and Development Study (HGDS) cohort as well as in the subset for whom coreceptor tropism data were available, by the presence or absence of the SDF1‐3′A allele.

Table 2. 

Baseline and change in CD4+ T cell counts and plasma HIV‐1 RNA levels in the entire Hemophilia Growth and Development Study (HGDS) cohort as well as in the subset for whom coreceptor tropism data were available, by the presence or absence of the SDF1‐3′A allele.

There was an increased risk of progression to AIDS and death for subjects with the SDF1‐3′A allele, as determined by Kaplan‐Meier analysis (figure 1). By use of Cox proportional hazards models, we found that the hazard ratio (HR) for progression to AIDS in subjects with the SDF1‐3′A allele was 1.75 (95% confidence interval [CI], 1.04–2.96; P=.037) and was minimally affected by the incorporation of baseline plasma HIV‐1 RNA levels and CD4+ T cell counts into the model (table 3). Since both CCR5Δ32 and CCR2‐64I polymorphisms have been shown to influence the natural history of HIV‐1 disease, we also incorporated these variables into the model. This analysis demonstrated no significant change in the relationship between SDF1 genotype and progression to AIDS and death. Once again, the results were minimally changed when the analyses were limited to subjects heterozygous for SDF1‐3′A and when antiretroviral use was included in the model with plasma HIV‐1 RNA levels and CD4+ T cell counts (data not shown). The Kaplan‐Meier analyses and Cox proportional hazard models performed in the subgroup for whom coreceptor tropism data were available also showed an increased risk of progression to AIDS and death; HRs in subjects with the SDF1‐3′A allele were 2.75 (95% CI, 1.48–5.10; P=.001) and 2.01 (95% CI, 1.14–3.52; P=.015), respectively (figures 1 and 2 and table 3). This relationship was minimally changed by the inclusion of CCR5 and CCR2 genotypes into the model for progression to AIDS and death. Similarly, the relationship with progression to AIDS persisted when baseline plasma HIV‐1 RNA levels and CD4+ T cell counts were included in the models (table 3).

Figure 1. 

Kaplan‐Meier analysis for progression to clinical AIDS by SDF1 genotype in subjects who had not received a diagnosis of AIDS at baseline in the Hemophilia Growth and Development Study cohort (n=184) (A) and the subset of study participants for whom coreceptor tropism data were available (n=115) (B). Dotted line, wild‐type SDF1; broken line, subjects homozygous or heterozygous for the SDF1‐3′A allele.

Figure 1. 

Kaplan‐Meier analysis for progression to clinical AIDS by SDF1 genotype in subjects who had not received a diagnosis of AIDS at baseline in the Hemophilia Growth and Development Study cohort (n=184) (A) and the subset of study participants for whom coreceptor tropism data were available (n=115) (B). Dotted line, wild‐type SDF1; broken line, subjects homozygous or heterozygous for the SDF1‐3′A allele.

Table 3. 

Cox proportional hazard models for SDF1 genotype, predicting progression to clinical AIDS and death in the presence of AIDS.

Table 3. 

Cox proportional hazard models for SDF1 genotype, predicting progression to clinical AIDS and death in the presence of AIDS.

Figure 2. 

Kaplan‐Meier analysis for progression to death in the presence of AIDS by SDF1 genotype in the Hemophilia Growth and Development Study cohort (n=204) (A) and the subset of study participants for whom coreceptor tropism data were available (n=125) (B). Dotted line, wild‐type SDF1; broken line, subjects homozygous or heterozygous for the SDF1‐3′A allele.

Figure 2. 

Kaplan‐Meier analysis for progression to death in the presence of AIDS by SDF1 genotype in the Hemophilia Growth and Development Study cohort (n=204) (A) and the subset of study participants for whom coreceptor tropism data were available (n=125) (B). Dotted line, wild‐type SDF1; broken line, subjects homozygous or heterozygous for the SDF1‐3′A allele.

Since subjects with the SDF1‐3′A allele appeared to be at an increased risk of disease progression and since this group has a higher frequency of R5/X4‐tropic viruses, compared with the group with wt SDF1, it was of interest to assess how differences in tropism influenced the effect of SDF1‐3′A on disease progression. In fact, similar to what was seen for decline in CD4+ T cell counts, after adjustment for tropism was made, the relationship weakened: the HR for progression to AIDS by SDF1 genotype decreased from 2.75 to 1.68, and the relationship was no longer statistically significant (P=.115) (table 3). Again, none of the results related to disease progression were changed when antiretroviral use at baseline was included in the models or when the analyses were limited to subjects heterozygous for the SDF1‐3′A allele (data not shown).

Discussion

There has been divergence in the results of studies assessing the effect that the SDF1‐3′A allele has on HIV‐1 disease [16–23, 33]. Although the present study has focused on only one of many host genetic factors that might influence the natural history of HIV‐1 disease, it has shown that, in the HGDS cohort, this allele (in heterozygous and homozygous subjects) was associated with more‐rapid decline in CD4+ T cell counts and accelerated disease progression. Moreover, this genetic variant was associated with an increased likelihood of having detectable X4 viruses, and the results have suggested that much of the effect of SDF1‐3′A on disease progression may be related to the presence of this virus population.

A limitation of all cohort studies is the numerous confounding variables, such as differences in the characteristics of the study populations, the duration of follow‐up, and the influence of antiretroviral therapy. In fact, a possible explanation for the divergent results, compared with those from other studies, may relate to the observation that the protective effect on disease progression occurs early [16]. Since the HGDS enrolled hemophiliacs in 1989 and 1990 who had been infected ∼6–8 years before, early protective effects of the SDF1‐3′A allele could have been missed, although an effect to accelerate disease progression that may predominate later during the course of disease might be more apparent. Therefore, cohorts with large seroincident groups may show protection, whereas cohorts that include a mix of subjects during the early and later years of infection may show no overall effect on disease progression. Consistent with this is the fact that, in the present study, increased risk of disease progression associated with the SDF1‐3′A allele became evident only after the first few years of follow‐up, ∼9–10 years after the time of infection (figures 1 and 2).

Potential biological explanations for the effect of the SDF1‐3′A allele on the natural history of HIV‐1 disease has been explored by several groups. One study showed that subjects with the SDF1‐3′A allele had decreased SDF‐1 levels [24]; another showed that lower plasma SDF‐1 levels were associated with an increased frequency of SI viruses [34]. Therefore, the SDF1‐3′A allele may either causally or through positive linkage disequilibrium with causal alleles in or near SDF1 result in decreased SDF‐1 expression that may either facilitate the emergence or enhance the pathogenicity of SI or X4 viruses. If true, this effect is likely to be most prominent during the later years of infection, when X4 viruses tend to emerge, and less relevant during the early years, when R5 viruses drive much of disease pathogenesis. Although most studies have not demonstrated a relationship between the SDF1 genotype and the presence of SI viruses [17, 18, 35–38], all have been relatively small, have used biological phenotypes predominantly from cell‐associated virus cultured in vitro, and had varied patient populations. In contrast, in our study, we detected tropism by use of molecular methods on plasma virus in subjects during the later years of infection, when X4 viruses may have a greater influence on the natural history of HIV‐1 disease. The importance of our observation is further supported by the data showing that the inclusion of tropism in the models resulted in a marked decrease in the HR for progression to AIDS and death (table 3). These results suggest that the effect that the SDF1 genotype has on the presence of X4 virus in plasma may well be driving the observed increased risk of disease progression. If presence of the SDF1‐3′A allele is associated with lower plasma SDF‐1 levels, delayed disease progression during early infection may relate to the in vitro observation by Marechal et al. [39] that SDF‐1 increases tat transactivation of R5 viruses, viruses that predominate during the early years of infection.

There are several limitations of the present study, including that antiretroviral assignment was at the discretion of the provider and tropism data were not available for all subjects. Although an effect of antiretroviral therapy cannot be excluded, it seems unlikely, in light of the limited use of therapy at baseline and of potent therapy over time. Furthermore, the results were not changed when antiretroviral use was included in the models of HIV‐1 disease progression (data not shown). The lack of tropism data would not have influenced the observations relating SDF1 genotype to HIV‐1 disease progression seen in the entire cohort. In contrast, the analyses suggesting that the SDF1 genotype’s effect on disease progression is mediated by its relationship with tropism could have been influenced by the lack of data on the entire cohort. Nevertheless, this too seems unlikely, since subjects for whom tropism data were available were quite similar to subjects for whom tropism data were not available, including the use of antiretroviral therapy and the SDF1‐3′A allele frequency. Moreover, multiple‐imputation analysis was applied to impute the missing tropism data by use of baseline plasma HIV‐1 RNA levels, CD4+ T cell counts, antiretroviral use, SDF1 genotype, and other variables. The imputed tropism values were then included in the analysis of decline of CD4+ T cell counts and disease progression, with little change seen in the outcome of the analyses (data not shown).

The strength of the present observation is that it comes from a single, well‐characterized cohort, and the relationship between SDF1 genotype and longitudinal decline in CD4+ T cell counts (table 2) was consistent with the survival data (figures 1 and 2 and table 3). Moreover, the tropism data provide a plausible biological explanation for the clinical observations. Although most previous studies have focused on the effect that SDF1‐3′A homozygosity has on the natural history of HIV‐1 disease, the present study has demonstrated a significant association between heterozygosity for the SDF1‐3′A allele and accelerated disease progression. A potential explanation for this is that the HGDS cohort was restricted to subjects in the later years of disease, which excludes the early effects of this allele on disease progression and enrichment for subjects who would experience enhanced disease progression associated with the SDF1‐3′A allele. Finally, these observations add to other data suggesting that genetic markers that might influence levels of the natural ligand for coreceptors might influence the natural history of HIV‐1 disease [40]. Further studies to define how cytokines and cytokine expression might influence the natural history of HIV‐1 disease could lead to novel therapeutic strategies in the future.

Hemophilia Growth and Development Study

The following individuals are the center directors, study coordinators, or committee chairs of the study: Children’s Hospital Los Angeles: E. Gomperts, W. Y. Wong, F. Kaufman, M. Nelson, and S. Pearson; The New York Hospital–Cornell Medical Center: M. Hilgartner, S. Cunningham‐Rundles, and I. Goldberg; University of Texas Medical School, Houston: W. K. Hoots, K. Loveland, and M. Cantini; National Institutes of Health, National Institute of Child Health and Human Development: A. Willoughby and R. Nugent; Rho, Inc.: S. Donfield; Baylor College of Medicine: C. Contant, Jr.; University of Iowa Hospitals and Clinics: C. T. Kisker, J. Stehbens, S. O’Conner, and J. McKillip; Tulane University: P. Sirois; Children’s Hospital of Oklahoma: C. Sexauer, H. Huszti, F. Kiplinger, and S. Hawk; Mount Sinai Medical Center: S. Arkin and A. Forster; University of Nebraska Medical Center: S. Swindells and S. Richard; University of Texas Health Science Center, San Antonio: J. Mangos and R. Davis; Children’s Hospital of Michigan: J. Lusher, I. Warrier, and K. Baird‐Cox; Milton S. Hershey Medical Center: M. E. Eyster, D. Ungar, and S. Neagley; Indiana Hemophilia and Thrombosis Center: A. Shapiro and J. Morris; University of California–San Diego Medical Center: G. Davignon and P. Mollen; and Kansas City School of Medicine, Children’s Mercy Hospital: B. Wicklund and A. Mehrhof.

Acknowledgments

We are indebted to the children, adolescents, and parents who volunteered to participate in this study and to the members of the Hemophilia Treatment Centers. We are grateful to Mary McNally (Science Applications International, Frederick National Cancer Institute), for managing and shipping all clinical samples for this study; Bayer Healthcare Diagnostics, for performing HIV‐1 RNA assays; and Christos Petropoulos, Michael Bates, and Terri Wrin (ViroLogic, Inc.), for helpful discussion and assistance with performance of the coreceptor tropism assays.

References

1. 
Schuitemaker
H
Koot
M
Kootstra
NA
, et al.  . 
Biological phenotype of human immunodeficiency virus type 1 clones at different stages of infection: progression of disease is associated with a shift from monocytotropic to T‐cell‐tropic virus populations
J Virol
 , vol. 
66
 (pg. 
1354
-
60
)
2. 
Berger
EA
Doms
RW
Fenyo
EM
, et al.  . 
A new classification for HIV‐1
Nature
 , vol. 
391
 pg. 
240
 
3. 
Daar
ES
Chernyavskiy
T
Zhao
JQ
Krogstad
P
Chen
ISY
Sequential determination of viral load and phenotype in human immunodeficiency virus type 1 infection
AIDS Res Hum Retroviruses
 , vol. 
11
 (pg. 
3
-
9
)
4. 
Connor
RI
Mohri
H
Cao
Y
Ho
DD
Increased viral burden and cytopathicity correlate temporally with CD4+ T‐lymphocyte decline and clinical progression in human immunodeficiency virus type 1–infected individuals
J Virol
 , vol. 
67
 (pg. 
1772
-
7
)
5. 
Richman
DD
Bozzette
SA
The impact of the syncytium‐inducing phenotype of human immunodeficiency virus on disease progression
J Infect Dis
 , vol. 
169
 (pg. 
968
-
74
)
6. 
Koot
M
Keet
IPM
Vos
AHV
, et al.  . 
Prognostic value of HIV‐1 syncytium‐inducing phenotype for rate of CD4+ cell depletion and progression to AIDS
Ann Intern Med
 , vol. 
118
 (pg. 
681
-
8
)
7. 
Connor
RI
Sheridan
KE
Ceradini
D
Choe
S
Landau
NR
Change in coreceptor use correlates with disease progression in HIV‐1–infected individuals
J Exp Med
 , vol. 
185
 (pg. 
621
-
8
)
8. 
O’Brien
SJ
Nelson
GW
Human genes that limit AIDS
Nat Genet
 , vol. 
36
 (pg. 
565
-
74
)
9. 
Dean
M
Carrington
M
Winkler
C
, et al.  . 
Genetic restriction of HIV‐1 infection and progression to AIDS by a deletion allele of the CKR5 structural gene
Science
 , vol. 
273
 (pg. 
1856
-
62
)
10. 
Martin
MP
Dean
M
Smith
MW
, et al.  . 
Genetic acceleration of AIDS progression by a promoter variant of CCR5
Science
 , vol. 
282
 (pg. 
1907
-
11
)
11. 
Anzala
AO
Ball
TB
Rostron
T
O’Brien SJ
Brien
SJ
Plummer
FA
CCR2‐64I allele and genotype association with delayed AIDS progression in African women. University of Nairobi Collaboration for HIV Research
Lancet
 , vol. 
351
 (pg. 
1632
-
3
)
12. 
Smith
MW
Dean
M
Carrington
M
, et al.  . 
Contrasting genetic influence of CCR2 and CCR5 variants on HIV‐1 infection and disease progression
Science
 , vol. 
277
 (pg. 
959
-
65
)
13. 
Liu
R
Paxton
WA
Choe
S
, et al.  . 
Homozygous defect in HIV‐1 coreceptor accounts for resistance of some multiply‐exposed individuals to HIV‐1 infection
Cell
 , vol. 
86
 (pg. 
367
-
77
)
14. 
van Rij
RP
de Roda Husman
AM
Brouwer
M
Goudsmit
J
Coutinho
RA
Role of CCR2 genotype in the clinical course of syncytium‐inducing (SI) or non‐SI human immunodeficiency virus type 1 infection and in the time to conversion to SI virus variants
J Infect Dis
 , vol. 
178
 (pg. 
1806
-
11
)
15. 
Tiensiwakul
P
Stromal cell‐derived factor (SDF) 1–3′A polymorphism may play a role in resistance to HIV‐1 infection in seronegative high‐risk Thais
Intervirology
 , vol. 
47
 (pg. 
87
-
92
)
16. 
Winkler
C
Modi
W
Smith
MW
, et al.  . 
Genetic restriction of AIDS pathogenesis by an SDF‐1 chemokine gene variant
Science
 , vol. 
279
 (pg. 
389
-
93
)
17. 
van Rij
RP
Broersen
S
Goudsmit
J
Coutinho
RA
Schuitemaker
H
The role of a stromal cell‐derived factor‐1 chemokine gene variant in the clinical course of HIV‐1 infection
AIDS
 , vol. 
12
 (pg. 
F85
-
90
)
18. 
Brambilla
A
Villa
C
Rizzardi
G
, et al.  . 
Shorter survival of SDF1‐3′A/3′A homozygotes linked to CD4+ T cell decrease in advanced human immunodeficiency virus type 1 infection
J Infect Dis
 , vol. 
182
 (pg. 
311
-
5
)
19. 
Mummidi
S
Ahuja
SS
Gonzalez
E
, et al.  . 
Genealogy of the CCR5 locus and chemokine system gene variants associated with altered rates of HIV‐1 disease progression
Nat Med
 , vol. 
4
 (pg. 
786
-
93
)
20. 
Singh
KK
Barroga
CR
Hughes
MD
, et al.  . 
Genetic influence of CCR5, CCR2, and SDF1 variants on human immunodeficiency virus 1 (HIV‐1)–related disease progression and neurological impairment, in children with symptomatic HIV‐1 infection
J Infect Dis
 , vol. 
188
 (pg. 
1461
-
72
)
21. 
Ioannidis
JP
Rosenberg
PS
Goedert
JJ
, et al.  . 
Effects of CCR5‐Delta32, CCR2‐64I, and SDF‐1 3′A alleles on HIV‐1 disease progression: an international meta‐analysis of individual‐patient data
Ann Intern Med
 , vol. 
135
 (pg. 
782
-
95
)
22. 
Easterbrook
PJ
Rostron
T
Ives
N
Troop
M
Gazzard
BG
Chemokine receptor polymorphisms and human immunodeficiency virus disease progression
J Infect Dis
 , vol. 
180
 (pg. 
1096
-
105
)
23. 
Meyer
L
Magierowska
M
Hubert
JB
, et al.  . 
CC‐chemokine receptor variants, SDF‐1 polymorphism, and disease progression in 720 HIV‐infected patients. SEROCO Cohort. Amsterdam Cohort Studies on AIDS
AIDS
 , vol. 
13
 (pg. 
624
-
6
)
24. 
Soriano
A
Martinez
C
Garcia
F
, et al.  . 
Plasma stromal cell–derived factor (SDF)–1 levels, SDF1‐3′A genotype, and expression of CXCR4 on T lymphocytes: their impact on resistance to human immunodeficiency virus type 1 infection and its progression
J Infect Dis
 , vol. 
186
 (pg. 
922
-
31
)
25. 
Hilgartner
MW
Donfield
SM
Willoughby
A
, et al.  . 
Hemophilia Growth and Development Study: design, methods and entry data
Am J Pediatr Hematol Oncol
 , vol. 
15
 (pg. 
208
-
18
)
26. 
Daar
ES
Lynn
H
Donfield
S
, et al.  . 
Effects of plasma HIV RNA, CD4+ T lymphocytes, and the chemokine receptors CCR5 and CCR2b on HIV disease progression in hemophiliacs
J Acquir Immune Defic Syndr
 , vol. 
21
 (pg. 
317
-
25
)
27. 
Centers for Disease Control and Prevention
Hemophilia surveillance system report
MMWR
 , vol. 
1
 (pg. 
1
-
14
)
28. 
Dewar
RL
Highbarger
HC
Sarmiento
MD
, et al.  . 
Application of branched DNA signal amplification to monitor human immunodeficiency virus type 1 burden in human plasma
J Infect Dis
 , vol. 
170
 (pg. 
1172
-
9
)
29. 
Petropoulos
CJ
Parkin
NT
Limoli
KL
, et al.  . 
A novel phenotypic drug susceptibility assay for human immunodeficiency virus type 1
Antimicrob Agents Chemother
 , vol. 
44
 (pg. 
920
-
8
)
30. 
Richman
DD
Wrin
T
Little
SJ
Petropoulos
CJ
Rapid evolution of the neutralizing antibody response to HIV type 1 infection
Proc Natl Acad Sci USA
 , vol. 
100
 (pg. 
4144
-
9
)
31. 
Centers for Disease Control
Revision of the CDC surveillance case definition for acquired immune deficiency
MMWR
 , vol. 
36
 (pg. 
1S
-
15S
)
32. 
Longford
N
Random coefficient models
Oxford: Clarendon Press,
 
33. 
Mazzucchelli
R
Corvasce
S
Violin
M
, et al.  . 
Role of CCR5, CCR2 and SDF‐1 gene polymorphisms in a population of HIV‐1 infected individuals
J Biol Regul Homeost Agents
 , vol. 
15
 (pg. 
265
-
71
)
34. 
Llano
A
Barretina
J
Blanco
J
Gutierrez
A
Clotet
B
Stromal‐cell‐derived factor 1 prevents the emergence of the syncytium‐inducing phenotype of HIV‐1 in vivo
AIDS
 , vol. 
15
 (pg. 
1890
-
2
)
35. 
Balotta
C
Bagnarelli
P
Corvasce
S
, et al.  . 
Identification of two distinct subsets of long‐term nonprogressors with divergent viral activity by stromal‐derived factor 1 chemokine gene polymorphism analysis
J Infect Dis
 , vol. 
180
 (pg. 
285
-
9
)
36. 
Watanabe
MA
de Oliveira Cavassin
GG
Orellana
MD
, et al.  . 
SDF‐1 gene polymorphisms and syncytia induction in Brazilian HIV‐1 infected individuals
Microb Pathog
 , vol. 
35
 (pg. 
31
-
4
)
37. 
Tresoldi
E
Romiti
ML
Boniotto
M
, et al.  . 
Prognostic value of the stromal cell–derived factor 1 3′A mutation in pediatric human immunodeficiency virus type 1 infection
J Infect Dis
 , vol. 
185
 (pg. 
696
-
700
)
38. 
Lathey
JL
Tierney
C
Chang
SY
, et al.  . 
Associations of CCR5, CCR2, and stromal cell–derived factor 1 genotypes with human immunodeficiency virus disease progression in patients receiving nucleoside therapy
J Infect Dis
 , vol. 
184
 (pg. 
1402
-
11
)
39. 
Marechal
V
Arenzana‐Seisdedos
F
Heard
JM
Schwartz
O
Opposite effects of SDF‐1 on human immunodeficiency virus type 1 replication
J Virol
 , vol. 
73
 (pg. 
3608
-
15
)
40. 
Gonzalez
E
Kulkarni
H
Bolivar
H
, et al.  . 
The influence of CCL3L1 gene–containing segmental duplications on HIV‐1/AIDS susceptibility
Science
 , vol. 
307
 (pg. 
1434
-
40
)
Potential conflicts of interest: W.H. is an employee of Virologic, Inc., the company that performed the coreceptor tropism assay; E.S.D. has acted as a consultant and advisor to ViroLogic, Inc.
Financial support: National Institutes of Health (NIH; grants HD41224, AI43638, and AI27660 and contract NO1‐CO‐12400); National Cancer Institute; Universitywide AIDS Research Program (grant CCTG‐ CC99 SD003). Development of the envelope coreceptor tropism assay was supported, in part, by the small business innovative research–advanced technology grant R44 AI048990 from the NIH National Institute of Allergy and Infectious Diseases. HIV‐1 RNA measurements were performed by Bayer Healthcare Diagnostics.
Participating centers and individuals are listed after the text.