Most HIV-1-infected patients who are treated with highly active antiretroviral treatment (HAART) achieve viraemia levels below 500 or 50 HIV-1 RNA copies/ml [1,2] However, such HIV-1 RNA levels do not imply that viral replication has stopped; even in patients who have undetectable viraemia levels, the viral genome keeps evolving, and replication-competent HIV-1 persist in resting CD4 cells [3–6] Nevertheless, low levels of viraemia during antiretroviral treatment predict sustained virological response; patients whose viraemia fell below 200 or 20 HIV-1 RNA copies/ml are less prone to virological failure than those who stayed above these thresholds [7,8]
The ruling paradigm for the treatment of HIV-1 infection is to `hit early and hard'  The clinical benefit of `hitting hard' has been documented; patients who receive triple antiretroviral therapy have fewer AIDS events and better survival than those receiving a double therapy [1,10–12] In contrast, early antiretroviral treatment remains debatable [13–15] The possible benefits of `hitting early' include the limitation of viral spread during primary HIV infection (PHI) [16,17] the preservation of immune functions [18,19] and a reduced number of viral quasispecies, some of which may carry drug-resistance mutations [20,21]
The aim of this study was to determine whether the early initiation of antiretroviral treatment in the course of HIV-1 infection leads to lower viraemia levels. Patients treated during PHI were compared with patients receiving HAART later in the course of the disease. To address this issue an ultrasensitive viraemia assay has been developed, which has a detection limit of 3 HIV-1 RNA copies/ml.
Materials and methods
All patients were drug naïve before the initiation of antiretroviral therapy. Eligible patients had received combined antiretroviral therapy for at least 72 weeks, adhered to treatment (more than 90% of drugs taken according to patient interviews), and had reached HIV-1 RNA levels of less than 500 copies/ml after 24 weeks of therapy. Three groups were compared: patients who initiated treatment at the time of PHI (10 PHI patients); during chronic infection but with CD4 cells counts of 500/mm3 or greater (10 immunocompetent patients); or those with CD4 cell counts of less than 500/mm3 (21 immunosuppressed patients). Convenience samples of patients who fulfilled the inclusion criteria were selected for this study.
HIV-1 RNA assays
The standard determination of plasma HIV-1 RNA levels was performed according to the manufacturer's instructions (Amplicor HIV Monitor, Roche, Basel, Switzerland; or bDNA assay, Chiron Corp., Emeryville, CA, USA). Samples collected at 24 weeks and later were analysed using a modified, more sensitive version of the Amplicor assay [22,23] Briefly, plasma samples (1 ml) were centrifuged at 50 000 g (4°C) for 80 min (Heraeus Biofuge 28 RS, rotor 3740, Heraeus AG, Osterode, Germany) before RNA extraction, to increase the input of viral RNA. After the removal of plasma, the following steps were performed according to the manufacturer's instructions, except that the concentration of the internal quantitative standard was reduced 15-fold and the volume of specimen diluent was reduced 7.3-fold (55 versus 400 μl). The substrate incubation time was extended to 15 min.
The main outcome was the log10 HIV-1 RNA copies/ml measured 24–120 weeks after the initiation of antiretroviral treatment in the three groups of patients. The mean optical density (± SD) of 32 negative controls included in the successive runs was 0.074 ± 0.008. All patient samples with an optical density of less than 0.150, corresponding to the mean +10 SD of negative controls were reported as undetectable. A value of 3 copies/ml (0.48 log10) was attributed to all patient samples with calculated RNA copy numbers below 3, but with optical density of 0.150 or greater. Samples with undetectable HIV-1 RNA were assigned a value of 0.1 (−1 log10) copies/ml. The Wilcoxon paired test was used for comparison between baseline and follow-up values. Groups were compared using the Mann–Whitney test for continuous variables and a chi-square test for dichotomous variables. Associations between independent variables (baseline CD4 cell counts, HIV-1 RNA and the delay between infection and the initiation of therapy) and mean HIV-1 RNA levels were explored by means of non-parametric regression (Lowess)  Next, linear regression models were used to assess associations between baseline parameters and mean HIV-1 RNA levels. For the analysis of baseline CD4 cell counts, immunocompetent and immunosuppressed chronically infected patients were combined because CD4 cell levels defined these groups. To account for the lack of independence of observations in the same patient, we analysed residual HIV-1 RNA levels using generalized estimating equation models  as implemented on Stata (Stata Corporation, College Station, TX, USA). P values of less than 0.05 were considered to be significant.
Ultrasensitive HIV-1 viraemia assay
The limit of detection of the ultrasensitive assay was evaluated in three experiments using replicates of plasma dilution containing approximately 10, 5, 2.5 and 1.25 RNA copies/ml. All 22 samples containing 10 and 5 copies/ml, 20 out of 22 (91%) of the samples containing 2.5 copies/ml, and six out of 11 (55%) of the samples containing 1.25 copies/ml were found to be positive. Reproducibility was assessed at four concentrations on 15 replicates; the mean coefficient of variation ranged from 10 to 27% (mean ± SD: 1.72 ± 0.18, 1.32 ± 0.14, 1.02 ± 0.23, 0.81 ± 0.22 log10 RNA copies/ml). The mean optical density (± SD) of 32 negative controls included in the successive runs was 0.074 ± 0.008.
Treatment and baseline patient characteristics
Five patients with PHI received a combination of zidovudine and didanosine, and five received zidovudine, lamivudine and indinavir. In all patients with PHI, antiretroviral treatment was initiated within one month of the acute retroviral syndrome. At the time of treatment initiation, two patients had a negative Western blot, three patients had one reactive band, and five patients had less than four reactive bands. All patients with chronic infection received triple therapy, including indinavir or ritonavir and two reverse transcriptase inhibitors (zidovudine or stavudine, and lamivudine). The median duration of treatment was 108 weeks (range 84–120) for PHI patients, 84 weeks (range 72–84) for immunocompetent patients, and 96 weeks (range 72–120) for immunosuppressed chronically infected patients. No significant differences in demographic characteristics were observed between PHI, immunocompetent and immunosuppressed patients; 60, 70 and 60% were men, and the median age was 40, 37 and 34 years in the three groups, respectively. An estimated date of HIV-1 infection (median time between the last negative and first positive anti-HIV antibodies test) was available for nine (90%) immunocompetent and 14 (67%) immunosuppressed chronically infected patients. The respective median intervals between infection and the initiation of therapy were 3.07 years (range 1.04–5.83) and 5.71 years (range 2.58–11.59) (P = 0.02). Before the initiation of antiretroviral treatment, immunosuppressed chronically infected patients had higher HIV-1 RNA levels than immunocompetent chronically infected patients (P < 0.001) (Table 1).
The median decrease in HIV-1 RNA levels after 52 weeks of therapy was 4.63 log10 copies/ml (range 3.78–6.73; P = 0.005) for PHI patients, 3.75 log10 copies/ml (range 0.85–5.51; P = 0.005), and 4.02 log10 copies/ml (range 2.98–5.97; P < 0.001) for immunocompetent and immunosuppressed chronically infected patients; the differences between groups were not statistically significant. There was no difference in virological response between PHI patients treated with double or triple therapy. No significant decrease in HIV-1 RNA levels occurred between week 52 and the last available samples (72–120 weeks) for the three groups.
Residual viraemia levels
Overall, 249 HIV-1 RNA samples collected 24–120 weeks after the initiation of treatment were analysed (median six samples per patient; range 4–8) (Fig. 1). The HIV-1 RNA level was lower than 50 copies/ml in 97, 99, and 83% of samples from PHI, immunocompetent, and immunosuppressed chronically infected patients, respectively.
Five of the 10 PHI patients, but none of the 31 chronically infected patients maintained their HIV-1 RNA level at less than 3 copies/ml after reaching this level during the whole study period (Fisher's exact test; P < 0.001). For individual patients, the mean proportion of samples with HIV-1 RNA levels of less than 3 copies/ml after 6 months of treatment was 75% (range 20–100) for PHI patients, compared with 32% (range 14–57) and 8% (range 0–40) for immunocompetent and immunosuppressed chronically infected patients, respectively. The mean HIV-1 RNA level measured 24–120 weeks after the initiation of treatment was −0.54 log10 copies/ml for PHI, 0.13 log10 for immunocompetent chronically infected patients and 1.06 log10 for immunosuppressed chronically infected patients; all between-group differences were significant (all pairwise P < 0.001) (Table 2).
Regression models showed that patients who initiated treatment at the time of PHI had significantly lower HIV-1 RNA levels than patients who initiated treatment during the chronic stage of HIV infection. Higher baseline CD4 cell counts, lower baseline HIV-1 RNA levels and a shorter interval between infection and the initiation of therapy were significantly associated with lower viraemia in univariate analysis (Table 3). In multivariate analysis, only the initiation of therapy at the time of PHI and baseline CD4 cell counts were independently associated with residual HIV-1 RNA levels (model R2 = 0.68) (Table 3).
The association between the baseline HIV-1 RNA level and the mean level of viraemia measured from 6 to 120 weeks after the initiation of treatment was only apparent in immunosuppressed chronically infected patients (Fig. 2 a). For the analysis of baseline CD4 cell counts, immunocompetent and immunosuppressed chronically infected patients were combined because CD4 cell counts were used as a selection criterion. There was a clear association between baseline CD4 cell counts and residual HIV-1 RNA levels in patients who initiated treatment during the chronic phase of infection, but also in patients with PHI (Fig. 2 b). There was no evidence of an association between the delay between infection and the initiation of therapy and residual HIV-1 RNA levels within each group (Fig. 2 c). Similar results were obtained when samples with HIV-1 RNA levels of less than 3 copies/ml were assigned a value of 1 (0 log10) copy/ml (data not shown).
It was found that viraemia is more strongly suppressed when antiretroviral therapy is initiated during PHI than when it is started later in the course of HIV disease. In particular, half of the patients treated at the time of PHI achieved sustained undetectable viraemia (< 3 HIV-1 RNA copies/ml), compared with none of the patients treated after the establishment of chronic infection. Furthermore, whether treatment is initiated during PHI or later, higher CD4 cell counts at baseline predict a stronger suppression of viraemia. Both findings strengthen the case for the early initiation of antiretroviral treatment in HIV-1-infected patients. Patients in the delayed treatment group were selected in relation to favourable odds for treatment response (drug-naïve, high adherence) in order to reduce the risk of an overestimation of the benefits from treatment during PHI. Interestingly, no association was found between residual HIV-1 RNA levels and the delay between HIV-1 infection and the initiation of antiretroviral therapy after adjustment for CD4 levels in the delayed treatment group. This suggests that the integrity of the immune response is more important than the passage of time per se.
Both independent predictors of residual viraemia, i.e. the initiation of therapy at the time of PHI and baseline CD4 cell counts, suggest that the host's immune response interacts with treatment in the control of HIV-1 infection. The immune response plays a potentially important role in limiting viral replication in patients treated at the time of PHI  and abnormalities in T cell activation and maturation may be reversed when patients are treated early  This evidence and our data indicate that the initiation of treatment when the patient's immune response is still intact may attenuate long-term viral replication.
Another important finding was that pre-treatment viraemia levels did not predict residual viraemia under treatment, after adjustment for CD4 cell counts. This suggests that HAART did not merely slow down viral replication, but also prevented de-novo infection of CD4 cells. An alternative explanation would be the insufficient reliability of viraemia determinations, but both our calibration experiments and the strong relationship between baseline CD4 cell counts and residual viraemia argue against this hypothesis. Furthermore, full suppression of viral replication is plausible in this selected group of patients, who responded well to therapy (< 500 HIV-1 RNA copies/ml at 6 months) and adhered to treatment, but generalization to an unselected population of HIV-1-infected persons would be imprudent.
What then is the mechanism leading to the persistent low level of viraemia on HAART? At least three possibilities exist: active viral replication because of the low level of de-novo infection; the release of viruses trapped in the follicular dendritic cell network; and the release of viruses from infected resting CD4 cells. A low degree of de-novo infection might occur in sanctuaries such as the central nervous system or in relation to suboptimal drug levels [3,28,29] The release of viruses from follicular dendritic cells decreases markedly after 6 months of HAART  In contrast, resting CD4 cells infected with replication-competent viruses persist for years in patients on HAART [3–6] If this pool is the main source of viraemia in treated patients, these results would suggest that the size of this reservoir increases during the course of HIV-1 infection in the absence of antiretroviral therapy. Furthermore, the excellent fit of our model predicting residual viraemia from PHI and baseline CD4 cells (68% of explained variance is remarkable, given that both CD4 cell counts and residual viraemia are measured with random error) suggests that the growth of the pool of infected resting CD4 cells almost parallels the loss of circulating CD4 cells.
The association between the early initiation of antiretroviral treatment and lower residual viraemia suggests, but falls short of demonstrating, that early treatment may be clinically beneficial. Two lines of evidence support this hypothesis. First, viraemia predicts disease progression and mortality in patients with chronic HIV-1 infection [31,32] However, the latter findings were obtained in unselected cohorts of patients before the advent of HAART, and at much higher levels of viraemia. Second, the nadir of viraemia in treated patients within low viraemia values (500–20 copies/ml) predicts sustained virological response [7,8] Finally, the half-life of decay of latently infected, resting memory CD4 cells is shorter in patients with persistent undetectable viraemia compared with patients with viraemia bumps  The clinical benefit derived from the virological advantage gained by early treatment must be weighed against the greater possibility of adverse effects of therapy  or the development of viral drug resistance during the long-term treatment period. This issue may remain formally unresolved, because a direct demonstration in a randomized trial of the long-term clinical benefits from the early initiation of antiretroviral therapy may not be feasible.
The authors are indebted to Kim Zollinger for excellent technical help, and to Drs C. Renold-Moynier, J. Chodakewitz and J. Mellors for providing the patients.
1. Cameron DW, Heath-Chiozzi M, Danner S et al
. Randomised placebo-controlled trial of ritonavir in advanced HIV-1 disease.
: Advanced HIV Disease Ritonavir Study Group.
Lancet 1998, 351:543–54-9.
2. 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, 337:734–739.
3. Zhang L, Ramratnam B, Tenner-Racz K et al
. Quantifying residual HIV-1 replication in patients receiving combination antiretroviral therapy.
N Engl J Med 1999, 340:1605–1613.
4. Wong JK, Hezareh M, Günthard HF et al
. Recovery of replication-competent HIV despite prolonged suppression of plasma viraemia.
Science 1997, 278:1291–1295.
5. Finzi D, Hermankova M, Pierson T et al
. Identification of a reservoir for HIV-1 in patients on highly active antiretroviral therapy.
Science 1997, 278:1295–1300.
6. Finzi D, Blankson J, Siliciano JD et al
. Latent infection of CD4+ T cells provides a mechanism for lifelong persistence of HIV-1, even in patients on effective combination therapy.
Nat Med 1999, 5:512–517.
7. Kempf DJ, Rode RA, Xu Y et al
. The duration of viral suppression during protease inhibitor therapy for HIV-1 infection is predicted by plasma HIV-1 RNA at the nadir.
AIDS 1998, 12:F9–F14.
8. Raboud JM, Montaner JS, Conway B et al
. Suppression of plasma viral load below 20 copies/ml is required to achieve a long-term response to therapy.
AIDS 1998, 12:1619–1624.
9. Ho DD Time to hit HIV, early and hard.
N Engl J Med 1995, 333:450–451.
10. Egger M, Hirschel B, Francioli P et al
. Impact of new antiretroviral combination therapies in HIV infected patients in Switzerland: prospective multicentre study.
: Swiss HIV Cohort Study.
BMJ 1997, 315:1194–1199.
11. 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.
N Engl J Med 1997, 337:725–733.
12. Detels R, Munoz A, McFarlane G et al
. Effectiveness of potent antiretroviral therapy on time to AIDS and death in men with known HIV infection duration.
: Multicenter AIDS Cohort Study Investigators.
JAMA 1998, 280:1497–1503.
13. Perrin L, Yerly S. Clinical presentation, virological features, and treatment perspectives in primary HIV-1 infection.
In: AIDS Clinical review 1997/1998.
Volberding PA, Jacobson MA (editors). New York: M. Dekker; 1998. 67–92.
14. Levy J Caution: should we be treating HIV infection early?
Lancet 1998, 352:982–983.
15. Kinloch-De Loes S, Hirschel BJ, Hoen B et al
. A controlled trial of zidovudine in primary human immunodeficiency virus infection.
N Engl J Med 1995, 333:408–413.
16. Fauci AS, Pantaleo G, Stanley S, Weissman D Immunopathogenic mechanisms of HIV infection.
Ann Intern Med 1996, 124:654–663.
17. Pantaleo G, Cohen OJ, Schacker T et al
. Evolutionary pattern of human immunodeficiency virus (HIV) replication and distribution in lymph nodes following primary infection: implications for antiviral therapy.
Nat Med 1998, 4:341–345.
18. Dolan MJ, Clerici M, Blatt SP et al
. In vitro T cell function, delayed-type hypersensitivity skin testing, and CD4+ T cell subset phenotyping independently predict survival time in patients infected with human immunodeficiency virus.
J Infect Dis 1995, 172:79–87.
19. Carmichael A, Jin X, Sissons P, Borysiewicz L Quantitative analysis of the human immunodeficiency virus type 1 (HIV-1)-specific cytotoxic T lymphocyte (CTL) response at different stages of HIV-1 infection: differential CTL responses to HIV-1 and Epstein–Barr virus in late disease.
J Exp Med 1993, 177:249–256.
20. Zhu T, Mo H, Wang N et al
. Genotypic and phenotypic characterization of HIV-1 in patients with primary HIV infection.
Science 1993, 261:1179–1181.
21. Coffin JM HIV population dynamics in vivo: implication for genetic variation, pathogenesis, and therapy.
Science 1995, 267:483–489.
22. Schockmel G, Yerly S, Perrin L Detection of low HIV-1 RNA levels in plasma.
J Acquired Immune Defic Syndr 1997, 14:179–183.
23. Yerly S, Rutschmann OT, Opravil M, Marchal F, Hirschel B, Perrin L Cell-associated HIV-1 RNA in blood as indicator of viral load in lymph nodes.
J Infect Dis 1999, 180:850–853.
24. Cleveland WS Robust locally weighted regression and smoothing scatterplots.
J Am Stat Assoc 1979, 74:829–836.
25. Liang KY, Zeger SL Longitudinal data analysis using generalized linear models.
Biometrika 1986, 73:13–22.
26. Rosenberg ES, Billingsley JM, Caliendo AM, Boswell SL, Sax PE, Kalams SA, Walker BD Vigorous HIV-1-specific CD4+ T cell responses associated with control of viremia.
Science 1997, 278:1447–1450.
27. Bisset LR, Cone RW, Huber W et al
. Highly active anti-retroviral therapy during early HIV-infection reverses T-cell activation and maturation abnormalities.
AIDS 1998, 12:2115–2123.
28. Brack-Werner R Astrocytes: HIV cellular reservoirs and important participants in neuropathogenesis.
AIDS 1999, 13:1–22.
29. Hockett RD, Kilby JM, Derdeyn CA et al
. Constant mean viral copy number per infected cell in tissues regardless of high, low, or undetectable plasma HIV RNA.
J Exp Med 1999, 189:1545–1554.
30. Cavert W, Notermans DW, Staskus K et al
. Kinetics of response in lymphoid tissues to antiretroviral therapy of HIV-1 infection.
Science 1997, 276:960–964.
31. 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, 272:1167–1170.
32. Yerly S, Perneger TV, Hirschel B et al
. A critical assessment of the prognostic value of HIV-1 RNA levels and CD4+ cell counts in HIV-infected patients.
: The Swiss HIV Cohort Study.
Arch Intern Med 1998, 158:247–252.
33. Ramratnam B, Zhang L, Bonhoeffer S et al
. Decay of replication competent HIV-1 in resting CD4 T cells during prolonged antiretroviral treatment.
Antivir Ther 1999, 4 (Suppl. 1):117.117.
34. 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, 353:2093–2099.
Members of the Swiss HIV Cohort Study are: M. Battegay (Scientific Board Co-chairman), E. Bernasconi, Ph. Bürgisser, M. Egger, P, Erb, W. Fierz, M. Flepp (Clinical Group Chairman), P. Francioli (President of the SHCS, Centre Hospitalier, Universitaire Vaudois, Lausanne, Switzerland), H.J. Furrer, P. Grob, B. Hirschel (Scientific Board Co-chairman), B. Ledergerber, R. Malinverni, L. Matter (Laboratory Group Chairman), A. Meynard, M. Opravil, F. Paccaud, G. Pantaleo, L. Perrin, W. Pichler, J-C. Piffaretti, M. Rickenbach (Data Center Manager), P. Sudre, J. Schupbach, A. Telenti, P. Vernazza, R. Weber.