Abstract

BackgroundThe objective of this work was to assess the role of human immunodeficiency virus (HIV) reservoirs in the risk of disease progression, by studying the relationship between HIV DNA level in peripheral blood mononuclear cells (PBMCs) and progression toward acquired immunodeficiency syndrome (AIDS)

MethodsHIV-1 DNA levels in PBMCs were determined by quantitative polymerase chain reaction for 383 patients enrolled in the SEROCO Cohort Study who had experienced seroconversion and had been followed up for >8 years. We compared the predictive values of HIV DNA level, HIV RNA level, and CD4+ cell count

ResultsBetween 6 and 24 months after seroconversion, HIV DNA level was a major predictor of progression to AIDS independently of HIV RNA level and CD4+ T cell count (adjusted relative risk [RR] for a 1-log10 increase, 3.20 [95% confidence interval {CI}, 1.70–6.00]). HIV DNA level was also a major predictor of disease progression during the first 6 months after seroconversion (adjusted RR, 4.16 [95% CI, 1.70–10.21]), when HIV RNA level and CD4+ T cell count were less predictive. Thus, a combination of these 3 markers provides the best estimate of the risk of disease progression for each patient

ConclusionsOur results suggest that HIV DNA level could be a useful additional marker in clinical practice and could aid in helping to define the best time to initiate treatment for each patient

HIV-1 infection results in the long-term and stable persistence of the virus in host cells [1–3 ]. The persistence of a large number of HIV-infected cells at all stages of infection represents a major obstacle to the treatment of HIV infection [4]. Some cells even persist after prolonged highly active antiretroviral therapy (HAART), constituting the so-called “viral reservoirs” [5–8 ]. HIV preferentially infects HIV-specific CD4+ T cells, which are long-lived memory cells [9]. The reservoir is established soon after infection, and the degree to which it is diminished by HAART initiated at the time of acute infection or during chronic infection varies between patients [10, 11]. The clinical implications of lifelong viral persistence are currently debated. However, all groups agree that HIV reservoirs are the main obstacle to virus eradication and that more-potent treatment will not be sufficient to clear the reservoirs [11–13 ]. Some groups have proposed that it may be possible to purge the reservoirs by inducing replication in latently infected T cells [14, 15]

Although it is generally recognized that quantification of HIV DNA level in peripheral blood mononuclear cells (PBMCs) is important for pathophysiological studies, it remains unclear how this marker could be used to monitor disease progression and treatment efficacy. HIV DNA levels have been measured in HIV-infected patients treated with interferon-α or interleukin (IL)–2 [16, 17], and the prognostic value of baseline HIV DNA level for disease progression has been reported in treated patients [17, 18]. The objectives of the present study were to look back at the natural history of HIV infection and to assess the relationship between the level of HIV DNA in blood cells and disease progression. The SEROCO Cohort Study provided the opportunity to study 383 individuals who had experienced seroconversion and who had been followed up for 8 years, from 1988 to 1996—that is, before the HAART era. We assessed the predictive value of HIV DNA level in PBMCs for assessing the risk of progression to AIDS, to a CD4+ T cell count of <200 cells/μL, or to death, taking into account the well-established markers plasma HIV RNA level and CD4+ T cell count [19–21 ]. Numerous polymerase chain reaction (PCR)–based technologies for quantifying cell-associated HIV DNA level have been described recently [22, 23]. We quantified cell-associated HIV DNA level by use of a PCR assay targeting the gag gene. This method detects all forms of intracellular HIV DNA—that is, unintegrated and integrated linear DNA, as well as 1–long terminal repeat (LTR) and 2-LTR circles [24, 25]. Finally, we used this simple assay to estimate the additive value of measuring HIV DNA level in clinical practice, in the context of the current treatment guidelines in North America and Europe, which are mainly based on CD4+ T cell count [26–28 ]

Patients and Methods

PatientsAmong the 1508 HIV-1–infected patients enrolled in the prospective multicenter French SEROCO Cohort Study from 1988 to 1996, 383 met the following criteria: known date of HIV-1 infection, AIDS-free clinical status at enrollment, no prior antiretroviral therapy (ART), and cells and serum stored at time of enrollment [21]. All 383 of these patients who had experienced seroconversion provided blood samples within 2 years of becoming infected. They were divided into 2 groups, to account for known differences in the predictive value of clinical, virological, and immunological markers between those who had samples collected early after seroconversion and those who had samples collected later after seroconversion [29, 30]. The first blood sample was collected within 6 months of infection in 112 cases and was collected between 6 and 24 months after infection in 271 cases. Aliquots of serum and PBMCs (3–5 million) from all patients had been frozen in liquid nitrogen. Patients were seen every 3 or 6 months, depending on their clinical and biological status. None of the patients received ART at enrollment, at the time when HIV-1 DNA level was evaluated. During the follow-up period, 157 (41%) patients received a mono- or bitherapy nucleoside reverse-transcriptase inhibitor (NRTI)–based regimen, and none of the patients received HAART, including protease inhibitors, before the cutoff date of the analysis (31 March 1996). This multicenter prospective cohort study was approved by an ethics committee, and all patients gave written informed consent

HIV-1 DNA level in PBMCsHIV-1 DNA level in PBMCs was quantified by use of a prototypic assay based on the Amplicor HIV Monitor test (Roche Diagnostics) and by use of an HIV-1 DNA internal standard of quantification provided by Roche, as reported elsewhere [21, 25, 31]. Briefly, PBMCs were separated by Ficoll-Hypaque gradient, and aliquots were frozen until use. The total DNA extracted from each aliquot of PBMCs was first determined with Hoechst dye (Pharmacia): 1 μg of DNA was considered to be equivalent to ∼150,000 cells. The amplification was performed using primers and a probe located in the gag region. Final results were expressed as the number or the log10 number of HIV-1 DNA copies per 106 PBMCs (threshold, 10 copies/PCR, or 60 copies/106 PBMCs)

HIV-1 RNA level in blood samplesHIV-1 RNA level in serum was quantified by use of the Amplicor HIV Monitor test. We used the same primers and probe to amplify HIV-1 DNA and RNA. Serum samples containing <400 copies/mL (the lower detection limit) were tested using the ultrasensitive specimen preparation protocol, which has a detection limit of 50 copies/mL

Statistical analysisThe Kaplan-Meier method was used to construct event-free survival curves, which were compared using the log-rank test. Crude and adjusted relative risks (RRs) and their 95% confidence intervals (CIs) were calculated using Cox proportional-hazard models to assess the predictive value of HIV-1 DNA level, independent of CD4+ T cell count and HIV-1 RNA level in blood. The following 3 outcome measures were used: clinical AIDS (1993 European definition), a CD4+ T cell count of <200 cells/μL, and death. Student’s t test and Pearson’s correlation coefficient were used to assess the association between virological markers and other variables

Results

PatientsThe main characteristics of the 383 HIV-infected patients are presented in table 1. Half of the subjects were infected before June 1988. The median follow-up time after infection was 84 months. The first study group comprised 271 patients who first gave samples 6–24 months (median, 11.0 months) after primary infection. These patients were considered to be at the so-called “plateau phase,” also known as the “steady state.” The second group comprised 112 patients who first gave blood samples within 6 months (median, 3.8 months) of primary infection. The percentage of men, baseline CD4+ T cell counts, and baseline HIV RNA levels differed significantly between the 2 groups (P = .01, P=.04, and P=.05, respectively)

Table 1

Baseline and follow-up characteristics of the 383 patients who had experienced seroconversion who were enrolled in the French SEROCO Cohort Study

Table 1

Baseline and follow-up characteristics of the 383 patients who had experienced seroconversion who were enrolled in the French SEROCO Cohort Study

Predictive value of HIV DNA level during the plateau phaseHIV DNA levels between 6 and 24 months after infection ranged from 0.70 to 4.02 log10 copies/106 PBMCs (median, 2.86 log10 copies/106 PBMCs [interquartile range {IQR}, 2.45–3.21 log10 copies/106 PBMCs). HIV DNA level was highly correlated with HIV RNA level (r=0.69; P<.0001) and with CD4+ T cell count (r=-0.40; P<.0001) (figure 1). HIV DNA levels were significantly lower in patients with a CD4+ T cell count ⩾350 cells/μL than in those with a CD4+ T cell count <350 cells/μL. HIV DNA levels were also lower in women than in men (mean, 2.47 vs. 2.86 log10 copies/106 PBMCs; P<.0001) and were lower in patients with the Δ32-CCR5 deletion than in wild-type homozygous patients (2.63 vs. 2.83 log10 copies/106 PBMCs; P<.04)

Figure 1

Relationship between HIV-1 RNA and DNA levels in 271 patients from the SEROCO Cohort with baseline measurements taken between 6 and 24 months after infection (during the plateau phase). Each circle represents a patient. PBMCs, peripheral blood mononuclear cells

Figure 1

Relationship between HIV-1 RNA and DNA levels in 271 patients from the SEROCO Cohort with baseline measurements taken between 6 and 24 months after infection (during the plateau phase). Each circle represents a patient. PBMCs, peripheral blood mononuclear cells

HIV DNA level was strongly associated with disease progression. We performed Kaplan-Meier analyses of progression to clinical AIDS separately for each of the 3 markers (HIV DNA level, CD4+ T cell count, and HIV RNA level) by tertile (figure 2). χ2 values for the log-rank test were 48.5, 35.9, and 24.0 respectively (2 df; P<.0001 for each comparison). The probabilities of developing clinical AIDS at 5 years after infection were 7.6% (SD, 3%) for the lowest tertile of HIV DNA level, 15.1% (SD, 4.0%) for the middle tertile, and 40.3% (SD, 5.7%) for the highest tertile. These probabilities were 8.7% (SD, 3.1%), 17.3% (SD, 4.2%), and 38.1% (SD, 5.7%) for the 3 tertiles of HIV RNA level and 5.6% (SD, 3.7%), 25.5% (SD, 4.8%), and 30.6% (SD, 5.1%) for the 3 tertiles of CD4+ T cell count. The crude RR of progression to clinical AIDS was 5.80 (95% CI, 3.53–9.54) for each 1-log10 increase in HIV DNA level (table 2), 2.97 (95% CI, 2.13–4.15) for each 1-log10 increase in HIV RNA level, and 1.25 (95% CI, 1.13–1.38) for each 100-cell/μL decrease in CD4+ T cell count

Figure 2

Kaplan-Meier estimates of AIDS-free survival according to baseline tertiles of HIV-1 DNA level (A) baseline tertiles of CD4+ T cell count (B) and baseline tertiles of HIV-1 RNA level (C) in 271 patients with baseline measurements taken between 6 and 24 months after infection, during the plateau phase. P<.0001 for each comparison (log-rank test). PBMCs, peripheral blood mononuclear cells

Figure 2

Kaplan-Meier estimates of AIDS-free survival according to baseline tertiles of HIV-1 DNA level (A) baseline tertiles of CD4+ T cell count (B) and baseline tertiles of HIV-1 RNA level (C) in 271 patients with baseline measurements taken between 6 and 24 months after infection, during the plateau phase. P<.0001 for each comparison (log-rank test). PBMCs, peripheral blood mononuclear cells

Table 2

Relative risks (RRs) and 95% confidence intervals (CIs) of disease progression according to HIV DNA level, HIV RNA level, and CD4+ T cell count at baseline in the 383 patients with a known date of infection who were followed in the SEROCO Cohort Study between 1988 and 1996

Table 2

Relative risks (RRs) and 95% confidence intervals (CIs) of disease progression according to HIV DNA level, HIV RNA level, and CD4+ T cell count at baseline in the 383 patients with a known date of infection who were followed in the SEROCO Cohort Study between 1988 and 1996

We assessed the predictive value of HIV DNA level in the different strata determined by tertiles of CD4+ T cell count and HIV RNA level. This showed that the predictive value of HIV DNA level is independent of HIV RNA level and CD4+ T cell count at baseline (figure 3). Patients with HIV DNA levels ⩾3 log10 copies/106 PBMCs (i.e., 1000 copies/106 PBMCs) progressed to clinical AIDS more rapidly than did those with HIV DNA levels <3 log10 copies/106 PBMCs, in each category of CD4+ T cell count (figure 3A–3C) and HIV RNA level (figure 3D–3F)

Figure 3

Kaplan-Meier estimates of AIDS-free survival according to baseline HIV-1 DNA level (<3 log10 copies/106 peripheral blood mononuclear cells (PBMCs) [A] and ⩾3 log10 copies/106 PBMCs [B]) in each stratum of the study population determined by tertiles of baseline CD4+ T cell count (A, B and C) and tertiles of baseline HIV-1 RNA level (D, E and F), in 271 patients with baseline measurements taken between 6 and 24 months after infection

Figure 3

Kaplan-Meier estimates of AIDS-free survival according to baseline HIV-1 DNA level (<3 log10 copies/106 peripheral blood mononuclear cells (PBMCs) [A] and ⩾3 log10 copies/106 PBMCs [B]) in each stratum of the study population determined by tertiles of baseline CD4+ T cell count (A, B and C) and tertiles of baseline HIV-1 RNA level (D, E and F), in 271 patients with baseline measurements taken between 6 and 24 months after infection

We chose a cutoff at 3 log10, since this value was close to the median (2.86). In a Cox model containing the 3 markers, each remained significantly associated with time to clinical AIDS. The adjusted RR was higher for each 1-log10 increase in HIV DNA level than for each 1-log10 increase in HIV RNA level (3.20 [95% CI, 1.70–6.00] vs. 1.67 [95% CI, 1.09–2.55], respectively) (table 2). The adjusted RR for a 100-cell/μL decrease in CD4+ T cell count was 1.12 (95% CI, 1.01–1.23). Further adjustment for age at infection, CCR5 genotype, history of symptomatic primary infection, and sex only marginally modified these results. HIV DNA level, HIV RNA level, and CD4+ T cell count were also found to be independently predictive of progression to a CD4+ T cell count of <200 cells/μL and to death (table 2)

Predictive value of HIV DNA level during the first 6 months after infectionIn the group of 112 patients who gave samples within 6 months of infection, HIV DNA levels ranged from 1.65 to 4.00 log10 copies/106 PBMCs (median, 2.85 [IQR, 2.52–3.23]). Similar levels were observed in samples collected in the 6–24-month period. In this group of patients studied early after infection, only HIV DNA level was a significant predictor of progression to clinical AIDS in a Cox model including these 3 markers (adjusted RR, 4.16 [95% CI, 1.70–10.21]) (table 2). When the end point was a CD4+ T cell count of <200 cells/μL or death, early HIV DNA level was again found to be the only independent predictive marker (table 2)

Clinical relevance of HIV DNA levelTo study the value of HIV DNA level as an additional marker for use in clinical practice (i.e., for deciding when to initiate treatment), we calculated Kaplan-Meier estimates of the probability of developing clinical AIDS at 5 years after infection. In accordance with the guidelines currently used in North America and Europe, which are mainly based on CD4+ T cell counts, we chose the strategic CD4+ T cell count of 350 cells/μL and a serum HIV RNA level of 15,000 copies/mL (equivalent to 30,000 copies/mL in plasma). The 271 patients at the plateau phase of HIV infection were divided into 3 groups, as follows: group 1, patients with CD4+ T cell counts <350 cells/μL (treatment or close follow-up recommended); group 2, patients with CD4+ T cell counts ⩾350 cells/μL and serum HIV RNA levels ⩾15,000 copies/mL (treatment postponed); and group 3, patients with CD4+ T cell counts ⩾350 cells/μL and serum HIV RNA levels <15,000 copies/mL (treatment not recommended). The risk of progression to clinical AIDS based on these 2 markers was estimated to be 39% for patients in group 1, 26% for patients in group 2, and 11% for patients in group 3 (figure 4A)

Figure 4

A Kaplan-Meier estimates and 95% confidence intervals of the probability of progressing to clinical AIDS at 5 years after infection according to current guidelines for initiation of highly active antiretroviral therapy (group 1: CD4+ T cell count <350 cells/μL; group 2: CD4+ T cell count ⩾350 cells/μL and serum HIV-1 RNA level ⩾15,000 copies/mL; group 3: CD4+ T cell count ⩾350 cells/μL and serum HIV-1 RNA level <15,000 copies/mL), in 271 patients with baseline measurement between 6 and 24 months after infection. P<.0001 for comparison of the 3 groups (log-rank test). B Kaplan-Meier estimates of AIDS-free survival according to baseline HIV-1 DNA level (<3 log10 copies/106 peripheral blood mononuclear cells (PBMCs) [A] and ⩾3 log10 copies/106 PBMCs [B]) within each of the 3 groups defined above

Figure 4

A Kaplan-Meier estimates and 95% confidence intervals of the probability of progressing to clinical AIDS at 5 years after infection according to current guidelines for initiation of highly active antiretroviral therapy (group 1: CD4+ T cell count <350 cells/μL; group 2: CD4+ T cell count ⩾350 cells/μL and serum HIV-1 RNA level ⩾15,000 copies/mL; group 3: CD4+ T cell count ⩾350 cells/μL and serum HIV-1 RNA level <15,000 copies/mL), in 271 patients with baseline measurement between 6 and 24 months after infection. P<.0001 for comparison of the 3 groups (log-rank test). B Kaplan-Meier estimates of AIDS-free survival according to baseline HIV-1 DNA level (<3 log10 copies/106 peripheral blood mononuclear cells (PBMCs) [A] and ⩾3 log10 copies/106 PBMCs [B]) within each of the 3 groups defined above

To estimate the additive predictive value of HIV DNA level, the 3 groups were stratified according to HIV DNA level (stratum A, <3 log10 copies/106 PBMCs; stratum B, ⩾3 log10 copies/106 PBMCs). The risk of progression to clinical AIDS differed considerably when HIV DNA level was taken into account (P < .0001 for log-rank test of comparison of the 3 groups) (figure 4B). This demonstrates that the estimate of the risk of disease progression is more accurate when all 3 markers are considered together. Moreover, our results suggest that, in group 1 (in which all patients were eligible for treatment on the basis of current guidelines), patients in stratum A, despite having a low CD4+ T cell count, have a lower risk of progressing to clinical AIDS (estimated risk at 5 years after infection, 22%) than do the patients in stratum B of group 2 and stratum B of group 3, with HIV DNA levels of ⩾3 log10 copies/106 PBMCs (estimated risk at 5 years after infection, 32% and 28%, respectively). Finally, the risk of progression to clinical AIDS at 5 years after infection was high (>25%) for all patients with HIV DNA levels of ⩾3 log10 copies/106 PBMCs, including patients from stratum B of group 3, who would not be treated on the basis of current guidelines

Discussion

The value of the HIV DNA level in PBMCs was first suggested to be of interest in 2 small studies published in 1994 [32, 33]. In 1996, after standardized HIV RNA test kits became available, a large study reported that HIV RNA level is predictive of disease progression independently of CD4+ T cell count [20]. The essential role of HIV RNA level measurements in routine clinical practice was subsequently confirmed in the context of both treatment initiation and therapeutic follow-up [34, 35]. A recent study described an HIV DNA quantification assay that detects all HIV DNA forms in infected cells, including unintegrated and integrated linear genomes and 1- and 2-LTR circles; in parallel, a specific assay for the detection of 2-LTR circles showed that the first assay was the most sensitive [24]. It has also been reported that the predictive value of cell-associated HIV DNA level is independent of HIV RNA level in plasma [24]. Our study confirms that HIV DNA level in PBMCs is a major predictor of progression to clinical AIDS, to a CD4+ T cell count of <200/μL, and to death that is independent of both HIV RNA level and CD4+ T cell count. Moreover, our results show that HIV DNA level has a higher predictive value than do the other 2 markers during the 6–24 months after infection, when the steady-state level is established in untreated patients. We also showed that HIV DNA level in PBMCs early after seroconversion (0–6 months after infection) is predictive of disease progression: in the multivariate analysis, HIV DNA level was the only factor found to be predictive of disease progression. This could be explained by the large variability in HIV RNA levels and CD4+ T cell counts during the first weeks of primary infection and by the fact that HIV DNA levels are more homogeneous and seem to be established at a very early stage of infection [2]. It is also noteworthy that the median HIV DNA level during the first 6 months after infection was very similar to that found in the PRIMO Cohort (median, 2.91 log10 copies/106 PBMCs) during the first 2 months after infection [10]

HIV DNA level reflects the number of circulating HIV-infected cells, including CD4+ T cells, that are latently infected and represents the cellular stock—that is, the capacity of infected cells for viral production. HIV RNA level reflects the rate of virus replication, which is more related to the virus and its fitness. Our data suggest that HIV DNA and RNA levels have different merits in the assessment of the risk of disease progression and are explained by partially independent mechanisms. In other terms, the disease progresses at different speeds, depending on the CD4+ T cell count and HIV RNA level; thus, HIV DNA levels can be considered to be the driving force. Although there is a strong correlation between HIV DNA level in PBMCs and HIV RNA level in plasma, it is conceivable that the stock of HIV-infected cells plays a specific role in determining the risk of progression, even though the rate of virus replication has a direct influence on the maintenance of this stock [36]

We also explored the potential role of HIV DNA level in treatment decisions. The effectiveness of HAART in delaying progression to AIDS or death has been convincingly demonstrated [34, 35]. However, the accumulation of adverse events, such as metabolic disorders and mitochondrial toxicity [37], even during treatment of early infection [38], has raised questions about the long-term use of these drugs. As a result, one of the most debated issues in HIV therapy [39, 40] concerns the optimum time to initiate ART. This has led to progressive changes in the therapeutic guidelines. Since HIV infection leads to AIDS through a continuous and progressive process, it is difficult to design treatment guidelines only on the basis of clear-cut values, such as 350 or 200 CD4+ T cells/μL or 100,000, 50,000, or 30,000 copies of HIV RNA/mL in plasma. Our study suggests that it would be useful to combine all 3 markers when trying to predict progression, to obtain a better picture of the underlying dynamics of HIV disease for each patient

One of the main questions raised by our results is that of how this marker can be used in clinical practice. Since our data clearly show that HIV DNA level provides useful additional information for each patient, this marker may complement HIV RNA level. However, it should not be considered a substitute for HIV RNA level, which remains essential for evaluation of the efficacy of ART. HIV DNA level could provide additional information at the time of acute infection or of the first HIV-positive test result. HIV DNA level may be used to tailor treatment decisions more closely for each patient, especially decisions regarding treatment initiation. Given the uncertainties regarding the long-term toxicity of HAART, this additional information could facilitate the management of treatment interruptions for patients, including children as well as adults, who have low HIV DNA levels and are at risk of developing side effects [41, 42]. Finally, our results suggest that HIV DNA level could be a helpful guide in future clinical research, with the objective of limiting treatment exposure and minimizing drug toxicity

SEROCO Cohort Study Group Members

The members of the SEROCO Cohort Study Group are as follows: M. Bary (ACCTES, Paris); C. Rouzioux, M. Burgard, and J.-P. Viard (Centre Hospitalier Universitaire Necker-Enfants Malades, Paris); P. Dellamonica and J. Durant (Hôpital L’Archet, Nice); H. Gallais and A. M. Quinson (Hôpital de la Conception, Marseille); J.-F. Delfraissy, P. Lebras, C. Goujard, Y. Quertainmont, and M. T. Ranou (Hôpital du Kremlin-Bicêtre, Le Kremlin-Bicêtre); J. J. Lefrère, J. Lerable, J. Salpetrier, and M. C. Meyohas (Hôpital Saint-Antoine, Paris); J. P. Cassuto and M. Quaranta (Hôpital Cimiez, Nice); B. Dupont and M. Eliaszewicz (Hôpital de l’Institut Pasteur, Paris); D. Vittecoq and L. Escaut (Hôpital Paul Brousse, Villejuif); S. Herson, A. Coutellier, and M. Bonmarchand (Hôpital Pitié-Salpétrière, Paris); J. A. Gastaut, G. Fabre-Costeseque, and M. P. Drogoul (Hôpital Ste-Marguerite, Marseille); J. Dormont, P. Galanaud, F. Boué, and A. Lévy (Hôpital Antoine Béclère, Clamart); D. Séréni and J. Krulick (Centre Hospitalier Universitaire Saint-Louis, Paris); D. Sicard (Hôpital Cochin, Paris); J. L. Vildé, C. Leport, U. Colassante, and W. Nouiouia (Hôpital Bichat, Paris); C. Katlama, M. Richard, and C. Rivière (Hôpital Pitié-Salpétrière, Paris); A. Sobel and M. Lechevalier (Hôpital Henri Mondor, Créteil); M. Kazatchkine, M. Buisson, and J. Vrtousnik (Hôpital Broussais and Hôpital Européen Georges Pompidou, Paris); L. Guillevin, B. Jarousse, P. Cohen, and P. Dény (Hôpital Avicenne, Bobigny); and J.-B. Hubert, C. Deveau, and L. Meyer (INSERM U 578, Le Kremlin-Bicêtre)

Acknowledgments

We acknowledge all of the patients and physicians participating in the long-term prospective cohort study, SEROCO. We thank Peter Hale for helpful discussions. This work is dedicated to Jean-Florian Mettetal and Denis Bucquet, both of whom died of AIDS and actively participated in the initiation of the SEROCO project in 1986

References

1.
Pierson
T
McArthur
J
Siliciano
RF
Reservoirs for HIV‐1: mechanisms for viral persistence in the presence of antiviral immune responses and antiretroviral therapy
Annu Rev Immunol
 , 
2000
, vol. 
18
 (pg. 
665
-
708
)
2.
Chun
TW
Engel
D
Berrey
MM
Shea
T
Corey
L
Fauci
AS
Early establishment of a pool of latently infected, resting CD4+ T cells during primary HIV‐1 infection
Proc Natl Acad Sci USA
 , 
1998
, vol. 
95
 (pg. 
8869
-
73
)
3.
Cone
RW
Gowland
P
Opravil
M
Grob
P
Ledergerber
B
Levels of HIV‐infected peripheral blood cells remain stable throughout the natural history of HIV‐1 infection. Swiss HIV Cohort Study
AIDS
 , 
1998
, vol. 
12
 (pg. 
2253
-
60
)
4.
Chun
TW
Stuyver
L
Mizell
SB
, et al.  . 
Presence of an inducible HIV‐1 latent reservoir during highly active antiretroviral therapy
Proc Natl Acad Sci USA
 , 
1997
, vol. 
94
 (pg. 
13193
-
7
)
5.
Chun
TW
Justement
JS
Pandya
P
, et al.  . 
Relationship between the size of the human immunodeficiency virus type 1 (HIV‐1) reservoir in peripheral blood CD4+ T cells and CD4+&rcolon;CD8+ T cell ratios in aviremic HIV‐1–infected individuals receiving long‐term highly active antiretroviral therapy
J Infect Dis
 , 
2002
, vol. 
185
 (pg. 
1672
-
6
)
6.
Finzi
D
Hermankova
M
Pierson
T
, et al.  . 
Identification of a reservoir for HIV‐1 in patients on highly active antiretroviral therapy
Science
 , 
1997
, vol. 
278
 (pg. 
1295
-
300
)
7.
Havlir
DV
Strain
MC
Clerici
M
, et al.  . 
Productive infection maintains a dynamic steady state of residual viremia in human immunodeficiency virus type 1‐infected persons treated with suppressive antiretroviral therapy for five years
J Virol
 , 
2003
, vol. 
77
 (pg. 
11212
-
9
)
8.
Viard
JP
Burgard
M
Hubert
JB
, et al.  . 
Impact of 5 years of maximally successful highly active antiretroviral therapy on CD4 cell count and HIV‐1 DNA level
AIDS
 , 
2004
, vol. 
18
 (pg. 
45
-
9
)
9.
Douek
DC
Brenchley
JM
Betts
MR
, et al.  . 
HIV preferentially infects HIV‐specific CD4+ T cells
Nature
 , 
2002
, vol. 
417
 (pg. 
95
-
8
)
10.
Ngo‐Giang‐Huong
N
Deveau
C
Da Silva
I
, et al.  . 
Proviral HIV‐1 DNA in subjects followed since primary HIV‐1 infection who suppress plasma viral load after one year of highly active antiretroviral therapy
AIDS
 , 
2001
, vol. 
15
 (pg. 
665
-
3
)
11.
Strain
MC
Gunthard
HF
Havlir
DV
, et al.  . 
Heterogeneous clearance rates of long‐lived lymphocytes infected with HIV: intrinsic stability predicts lifelong persistence
Proc Natl Acad Sci USA
 , 
2003
, vol. 
100
 (pg. 
4819
-
24
)
12.
Persaud
D
Siberry
GK
Ahonkhai
A
, et al.  . 
Continued production of drug‐sensitive human immunodeficiency virus type 1 in children on combination antiretroviral therapy who have undetectable viral loads
J Virol
 , 
2004
, vol. 
78
 (pg. 
968
-
79
)
13.
Pomerantz
RJ
Reservoirs of human immunodeficiency virus type 1: the main obstacles to viral eradication
Clin Infect Dis
 , 
2002
, vol. 
34
 (pg. 
91
-
7
)
14.
Scripture‐Adams
DD
Brooks
DG
Korin
YD
Zack
JA
Interleukin‐7 induces expression of latent human immunodeficiency virus type 1 with minimal effects on T‐cell phenotype
J Virol
 , 
2002
, vol. 
76
 (pg. 
13077
-
82
)
15.
Korin
YD
Brooks
DG
Brown
S
Korotzer
A
Zack
JA
Effects of prostratin on T‐cell activation and human immunodeficiency virus latency
J Virol
 , 
2002
, vol. 
76
 (pg. 
8118
-
23
)
16.
Emilie
D
Burgard
M
Lascoux‐Combe
C
, et al.  . 
Early control of HIV replication in primary HIV‐1 infection treated with antiretroviral drugs and pegylated IFNα: results from the Primoferon A (ANRS 086) Study
AIDS
 , 
2001
, vol. 
15
 (pg. 
1435
-
7
)
17.
Gougeon
ML
Rouzioux
C
Liberman
I
, et al.  . 
Immunological and virological effects of long term IL‐2 therapy in HIV‐1‐infected patients
AIDS
 , 
2001
, vol. 
15
 (pg. 
1729
-
31
)
18.
Tierney
C
Lathey
JL
Christopherson
C
, et al.  . 
Prognostic value of baseline human immunodeficiency virus type 1 DNA measurement for disease progression in patients receiving nucleoside therapy
J Infect Dis
 , 
2003
, vol. 
187
 (pg. 
144
-
8
)
19.
Mellors
JW
Kingsley
LA
Rinaldo
CR
Jr
, et al.  . 
Quantitation of HIV‐1 RNA in plasma predicts outcome after seroconversion
Ann Intern Med
 , 
1995
, vol. 
122
 (pg. 
573
-
9
)
20.
Mellors
JW
Rinaldo
CR
Jr
Gupta
P
White
RM
Todd
JA
Kingsley
LA
Prognosis in HIV‐1 infection predicted by the quantity of virus in plasma
Science
 , 
1996
, vol. 
272
 (pg. 
1167
-
70
)
21.
Hubert
JB
Burgard
M
Dussaix
E
, et al.  . 
Natural history of serum HIV‐1 RNA levels in 330 patients with a known date of infection. The SEROCO Study Group
AIDS
 , 
2000
, vol. 
14
 (pg. 
123
-
31
)
22.
Desire
N
Dehee
A
Schneider
V
, et al.  . 
Quantification of human immunodeficiency virus type 1 proviral load by a TaqMan real‐time PCR assay
J Clin Microbiol
 , 
2001
, vol. 
39
 (pg. 
1303
-
10
)
23.
Gupta
P
Ding
M
Cottrill
M
, et al.  . 
Quantitation of human immunodeficiency virus type 1 DNA and RNA by a novel internally controlled PCR assay
J Clin Microbiol
 , 
1995
, vol. 
33
 (pg. 
1670
-
3
)
24.
Kostrikis
LG
Touloumi
G
Karanicolas
R
, et al.  . 
Quantitation of human immunodeficiency virus type 1 DNA forms with the second template switch in peripheral blood cells predicts disease progression independently of plasma RNA load
J Virol
 , 
2002
, vol. 
76
 (pg. 
10099
-
108
)
25.
Christopherson
C
Kidane
Y
Conway
B
Krowka
J
Sheppard
H
Kwok
S
PCR‐based assay to quantify human immunodeficiency virus type 1 DNA in peripheral blood mononuclear cells
J Clin Microbiol
 , 
2000
, vol. 
38
 (pg. 
630
-
4
)
26.
Carpenter
CC
Cooper
DA
Fischl
MA
, et al.  . 
Antiretroviral therapy in adults: updated recommendations of the International AIDS Society‐USA Panel
JAMA
 , 
2000
, vol. 
283
 (pg. 
381
-
90
)
27.
Panel on Clinical Practices for Treatment of HIV Infection
Guidelines for the use of antiretroviral agents in HIV‐1‐infected adults and adolescents
Updated 7 April 2005
  
Available at: http://aidsinfo.nih.gov/guidelines/adult/AA_040705.pdf. Accessed 27 May 2005
28.
Delfraissy
JF
Prise en charge thérapeutique des personnes infectées par le VIH. Rapport 2002 sous la direction de JF Delfraissy
 , 
2002
Paris
Flammarion
29.
de Wolf
F
Spijkerman
I
Schellekens
PT
, et al.  . 
AIDS prognosis based on HIV‐1 RNA, CD4+ T‐cell count and function: markers with reciprocal predictive value over time after seroconversion
AIDS
 , 
1997
, vol. 
11
 (pg. 
1799
-
806
)
30.
Schacker
TW
Hughes
JP
Shea
T
Coombs
RW
Corey
L
Biological and virologic characteristics of primary HIV infection
Ann Intern Med
 , 
1998
, vol. 
128
 (pg. 
613
-
20
)
31.
Burgard
M
Izopet
J
Dumon
B
, et al.  . 
HIV RNA and HIV DNA in peripheral blood mononuclear cells are consistent markers for estimating viral load in patients undergoing long‐term potent treatment
AIDS Res Hum Retroviruses
 , 
2000
, vol. 
16
 (pg. 
1939
-
47
)
32.
Verhofstede
C
Reniers
S
Van Wanzeele
F
Plum
J
Evaluation of proviral copy number and plasma RNA level as early indicators of progression in HIV‐1 infection: correlation with virological and immunological markers of disease
AIDS
 , 
1994
, vol. 
8
 (pg. 
1421
-
7
)
33.
Chevret
S
Kirstetter
M
Mariotti
M
Lefrere
F
Frottier
J
Lefrere
JJ
Provirus copy number to predict disease progression in asymptomatic human immunodeficiency virus type 1 infection
J Infect Dis
 , 
1994
, vol. 
169
 (pg. 
882
-
5
)
34.
Hammer
SM
Squires
KE
Hughes
MD
, et al.  . 
A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. AIDS Clinical Trials Group 320 Study Team
N Engl J Med
 , 
1997
, vol. 
337
 (pg. 
725
-
33
)
35.
Gulick
RM
Mellors
JW
Havlir
D
, et al.  . 
Treatment with indinavir, zidovudine, and lamivudine in adults with human immunodeficiency virus infection and prior antiretroviral therapy
N Engl J Med
 , 
1997
, vol. 
337
 (pg. 
734
-
9
)
36.
Perelson
AS
Neumann
AU
Markowitz
M
Leonard
JM
Ho
DD
HIV‐1 dynamics in vivo: virion clearance rate, infected cell life‐span, and viral generation time
Science
 , 
1996
, vol. 
271
 (pg. 
1582
-
6
)
37.
Carr
A
Samaras
K
Thorisdottir
A
Kaufmann
GR
Chisholm
DJ
Cooper
DA
Diagnosis, prediction, and natural course of HIV‐1 protease‐inhibitor‐associated lipodystrophy, hyperlipidaemia, and diabetes mellitus: a cohort study
Lancet
 , 
1999
, vol. 
353
 (pg. 
2093
-
99
)
38.
Goujard
C
Boufassa
F
Deveau
C
Laskri
D
Meyer
L
Incidence of clinical lipodystrophy in HIV‐infected patients treated during primary infection
AIDS
 , 
2001
, vol. 
15
 (pg. 
282
-
4
)
39.
Hogg
RS
Yip
B
Chan
KJ
, et al.  . 
Rates of disease progression by baseline CD4 cell count and viral load after initiating triple‐drug therapy
JAMA
 , 
2001
, vol. 
286
 (pg. 
2568
-
77
)
40.
Opravil
M
Ledergerber
B
Furrer
H
, et al.  . 
Clinical efficacy of early initiation of HAART in patients with asymptomatic HIV infection and CD4 cell count >350 × 106/l
AIDS
 , 
2002
, vol. 
16
 (pg. 
1371
-
81
)
41.
Saitoh
A
Hsia
K
Fenton
T
, et al.  . 
Persistence of human immunodeficiency virus (HIV) type 1 DNA in peripheral blood despite prolonged suppression of plasma HIV‐1 RNA in children
J Infect Dis
 , 
2002
, vol. 
185
 (pg. 
1409
-
16
)
42.
Pellegrin
I
Caumont
A
Garrigue
I
, et al.  . 
Predictive value of provirus load and DNA human immunodeficiency virus genotype for successful abacavir‐based simplified therapy
J Infect Dis
 , 
2003
, vol. 
187
 (pg. 
38
-
46
)
Presented in part: 1st Workshop on HIV Persistence, Saint Martin, France, 7–9 December 2003
Financial support: Agence Nationale de Recherche sur le Sida (grants to the SEROCO Cohort Study)
SEROCO Cohort Study Group members are listed after the text