HIV infection has two main effects on the immune system, (i) a loss of CD4+ T lymphocytes, and (ii) general activation of immune cells, the latter being reflected by an increase in the percentage and absolute number of activated CD8+ T cells. The CD8+ T-cell count increases early after HIV infection and high levels of activation marker antigens (CD38, CD45RO and HLA-DR) are observed [1–4]. In a previous study we found that percentages of phenotypically distinct CD8+ subsets correlated with viral load, determined as cellular viraemia and plasma HIV RNA levels .
Changes in CD8+ lymphocyte subsets have been studied as markers of disease progression, but changes during antiretroviral therapy have been poorly documented. Previous studies involving zidovudine have shown a transient fall in T-cell activation after initiation of therapy, including decreased CD38 expression on CD8+ lymphocytes [6–8]. More potent regimens induce a more sustained reduction in these activated subsets [9,10]. These modifications have not been related to changes in viral load.
We conducted a prospective study to determine the effect of antiretroviral therapy on the relationship between CD8+ lymphocyte activation status and plasma HIV RNA levels in HIV-infected patients treated with zidovudine plus didanosine.
Patients and methods
The study participants were drawn from a prospective trial conducted in 10 centres [Add On Viral Load (AVL) ANRS 052] with 116 HIV-infected antiretroviral-naive patients at clinical stage A or B of the 1993 Centers for Diseases Control (CDC) classification. Inclusion criteria were a CD4+ cell count of 200–400 × 106/l and a plasma HIV RNA level of 4.18–5.48 log10 copies/ml. Participants gave their written informed consent before entering the study. They received 250 mg twice daily or 200 mg three times daily zidovudine (Retrovir, Glaxo-Wellcome, Marlyle-Roi, France) plus 200 mg twice daily didanosine (Videx, Bristol-Myers Squibb, La Défense, France). The patients studied were the first 30 to be enrolled in two of the participating centres.
Blood was collected for real-time measurements just prior to starting treatment, and then on day 15 and monthly from months 1 to 6 of treatment.
Immunophenotyping by flow cytometry
Flow cytometric analysis was performed on peripheral whole blood prepared using a standard lysis method, using a FACScan flow cytometer and LYSIS software (Becton Dickinson, San Jose, California, USA) as previously described . The CDC-recommended two-colour panel of monoclonal antibodies for HIV monitoring  was used to measure key lymphocyte subsets. CD8+ lymphocyte subsets were examined by means of three-colour immunophenotyping with an extensive panel of specially manufactured monoclonal antibodies (Becton Dickinson Immunocytometry Systems). A peridinin chlorophyll protein-conjugated anti-CD8 antibody was used in combination with fluorescein isothiocyanate-and phycoerythrin-conjugated antibodies. The fluorescein-conjugated antibodies were anti-HLA-DR, anti-CD45RO and anti-CD28. The phycoerythrin-conjugated antibodies were anti-HLA-DR, anti-CD38 and anti-CD28. Three-colour samples were gated on CD8bright lymphocytes, which were identified on the basis of staining intensity with an anti-CD8 monoclonal antibody. Because 99% of CD8bright cells were CD3+, CD8+ T-cell subsets were estimated from the CD8bright lymphocyte values to exclude natural killer cells.
Plasma HIV RNA was quantified using a commercial branched DNA method (detection limit, 500 copies/ml; Quantiplex, Chiron Corporation, Lyon, France) on duplicate samples collected on EDTA, according to the manufacturer's instructions.
Results are expressed as means ± SD. CD8+ subsets are given as percentages of total lymphocytes. Data were analysed using the StatView software package (version 4.5; Abacus Concepts, Berkeley, California, USA). Mean values of each variable were compared amongst the study groups by using analysis of variance. Changes over time were summarized for each variable by the difference between mean baseline and post-treatment values and by the least-square estimate of the slope of the linear regression plots of each value against time. P values of 0.05 or less were considered significant.
A total of 30 HIV-infected subjects (26 men, four women; 26 Caucasians, three Africans, one Haitian) were studied. Mean age was 37.8 years (range, 28–60 years). According to the CDC classification, 23 patients were in clinical category A and seven were in category B. Risk factors for HIV-1 infection were homo-/ bisexual contact in 20 cases, heterosexual contact in five, injecting drug use in four, and unidentified in one. Immunophenotyping data were compared with those in 23 healthy HIV-seronegative volunteers (12 men, 11 women; mean age, 36 years; range, 24–54 years).
Baseline plasma HIV RNA levels (4.81 ± 0.41 log10 copies/ml; range, 3.83–5.67 log10 copies/ml) fell sharply in the first 2 weeks of treatment (3.09 ± 0.62 log10 copies/ml on day 15) and a maximal decline of 1.8 log10 copies/ml was reached at month 3. Because the virological response versus time followed three distinct patterns, the corresponding groups of patients were analysed (Fig. 1). In 14 patients (47%), plasma HIV RNA fell to below the detection limit at month 1 and remained so thereafter (−2 log10 copies/ml at month 6; group 1, sustained responders). In 10 patients (33%) the fall was transient (−1.4 log10 copies/ml at month 6; group 2, transient responders). Finally, in six patients the plasma HIV RNA titre failed to fall below the detection limit (−0.9 log10 copies/ml at month 6; group 3, non-responders). We compared immunological changes according to the virological response in the three groups.
CD4+ and CD8+ lymphocyte subsets
The mean baseline CD4+ cell count was 262 ± 68 × 106/l (range, 139–443 × 106/l). The baseline CD8+ cell count was 1023 ± 527 × 106/l (range, 415–2786 × 106/l). Baseline values were not significantly different amongst the three groups.
Major T-cell changes were observed during the first month of therapy (Fig. 2). The CD4+ lymphocyte count first increased rapidly in the three groups of patients, reaching a mean gain of about 100 × 106/l. It was followed by a stabilization and a further moderate increase in the two groups of responders (overall CD4+ lymphocyte gain of 120 × 106/l at month 6) and a further fall in the non-responders (non-significant increase of 55 × 106/l). CD8+ lymphocyte count remained relatively stable.
Changes in CD8+ T-cell phenotype
At baseline, activated CD8+ subsets (CD38+, HLA-DR+, CD45RO+) were markedly elevated. This expansion of the CD8+ compartment was also accounted for by an increase in CD28-cells. The CD8+ cell count remained relatively stable on treatment, whereas the percentage of activated CD8+ cells declined gradually between months 1 and 6. The decrease over time was significant only in the sustained responders with regard to the following subsets: CD8+CD38+, CD8+HLA-DR+ and CD8+CD28–(relative fall at month 6, −37.2, −36.2 and −25.6%, respectively). No significant decrease was observed in the other two groups. The percentage of CD8+CD45RO+ lymphocytes fell significantly in the sustained and transient responders, and the relative fall was more pronounced in the sustained responders (−31.0 versus −27.0%).
Three-colour immunofluorescence analysis allowed us to define more precisely the CD8 subsets involved in these changes. In the 30 patients, the mean baseline percentages of the following activated CD8 subsets were markedly increased compared with the HIV-seronegative control group: CD38+HLA-DR+, 24.2 ± 10.3 versus 0.8 ± 0.8%; CD38+CD45RO+, 27.5 ±11.1 versus 1.9 ±1.2%; HLA-DR+CD45RO+, 28.1 ± 12.0 versus 2.1 ± 1.4%; CD38+CD28−, 37.9 ±10.2 versus 3.3 ±2.9%. The largest changes during treatment also involved these CD8 lymphocyte subsets (Fig. 3).The declines were marked in sustained responders, intermediate in transient responders, and non-significant in non-responders (Table 1). Although plasma HIV RNA remained undetectable at month 6 in sustained responders, the mean percentage of CD8+ lymphocytes with an activated phenotype remained significantly higher than in the HIV-seronegative control group: CD38+HLA-DR+, 11.1 ± 6.0% (P < 0.0001 versus control); CD38+CD45RO+, 14.5 ± 7.5% (P < 0.0001); HLA-DR+CD45RO+, 17.3 ±7.3% (P < 0.0001); CD38+CD28-, 20.1 ±7.1% (P < 0.0001).
We investigated changes in phenotypic markers of CD8+ cell activation during combination antiretroviral therapy of HIV infection, and its relationship with plasma HIV RNA levels in previously untreated patients. The results confirmed the reduction in CD8+ T-cell activation markers in patients on antiretroviral therapy [6,9,10]. They also emphasized the link between the reduction in CD8+ lymphocyte activation markers and the decline in plasma HIV RNA.
The 30 patients we studied were representative of the 116 patients included in the AVL trial in terms of their virological response and CD4+ T-lymphocyte increment. The degree and duration of the virological response on combined therapy with zidovudine and didanosine were heterogeneous, as in the AIDS Clinical Trials Group 175 trial . Patients in the Delta 1 trial had slightly more advanced disease but similar virological responses .
This heterogeneity led us to consider three groups of patients based on the pattern of virological response. Very consistent results were obtained: the course of immune activation ran parallel to changes in circulating viral RNA. A gradual decline in activated CD8+ T cells was seen in the sustained responders with all the marker combinations used, whereas only minor changes in CD8+ activation status or no significant change were noted in transient responders and nonresponders, respectively. In contrast with results obtained by Kelleher et al.  after treatment with ritonavir alone, showing a stable CD8+HLA-DR+ population, all the subsets of activated cells we examined decreased with the reduction in virus load. Likewise, Autran et al.  observed a decrease in both CD38+ and HLA-DR+CD8+ cells after triple-drug antiretroviral therapy. The present results provide more details of phenotypic changes, by the use of three-colour analysis, and show that the most strongly altered CD8+ subsets are those expressing more than one activation marker (CD38+CD45RO+, CD38+HLADR+, HLA-DR+CD38+, and CD38+CD28-subsets).
Plasma HIV RNA levels appear to be a valid predictor of HIV disease progression [12,14,15] and to reflect antiretroviral activity in vivo . CD38 expression on CD8+ lymphocytes is also a validated prognostic marker of HIV disease progression [2,3,17,18]. High CD38 expression in the early phase of HIV infection has a strong predictive value in addition to that provided by CD4+ cell measurement . Moreover, increased levels of activated CD8+ cells are associated with high plasma HIV RNA levels . In this study we found a correlation between the reduction in CD8+ activation and depression of plasma HIV RNA below the detection limit. However, the fall in immune activation occurred more slowly than the fall in HIV RNA. Furthermore, undetectable plasma HIV RNA in the sustained responder group was not associated with normal CD8+ lymphocyte activation status after 6 months of treatment, suggesting persistent low-level viral replication. Given the sensitivity of the technique used (branched DNA; detection limit, 500 copies/ml) we cannot rule out the persistence of low levels of plasma HIV RNA. These results may also suggest that the fall in plasma viraemia overestimates the therapeutic effect and that viral replication persists in lymphoid tissues.
The few available results indicate that the combination of zidovudine and didanosine has only a moderate effect on lymph-node HIV RNA levels [19,20]. Furthermore, a delay in lymph-node HIV clearance has been seen during triple combination therapy [21,22]. We have also found that CD8+ cell activation did not return to normal after 8 months of triple therapy including a protease inhibitor (stavudine, didanosine and ritonavir) . The slower decrease in viral replication in lymphoid tissue and the rapid decline in plasma viraemia may be due to virions trapped in the follicular dendritic cell network . By inducing chronic immune stimulation, these trapped virions may maintain a pool of activated circulating CD8+ T cells after plasma HIV RNA has reached undetectable levels. It is thus possible that the rate of decline in blood CD8+ activation markers is a better marker of a reduction in viral load in lymphoid organs .
The role of activated CD8+ T cells in HIV infection is not clear. The CD38+HLA-DR+ and CD28-phenotypes of CD8+ cells are commonly associated with cytotoxic T-lymphocyte (CTL) activity [26,27]. CTL have been detected in the early phase of infection  and may play a fundamental role in controlling HIV replication, as stronger CTL activity has been linked to slower disease progression . However, the increase in activated CD8+ T cells runs parallel to the increase in viral load (unpublished data). Moreover, in some long-term non-progressors, low viral load is associated with low CD8+ activation status and low CTL activity, suggesting that antigenic stimulation is not strong enough to induce a measurable CTL response in these patients . The correlation between the fall in the percentage of CD8+ lymphocytes expressing activation markers and suppression of plasma HIV RNA supports the concept that activation of CD8+ cells is antigen-driven and may represent an ongoing immune response to continuous HIV production. High and persistent lymphoid activation may lead to a detrimental increase in cytokine production that overwhelms the beneficial effect of CTL activity . On the other hand, viral mutations may impair the capacity of CTL to control viral replication and disease progression .
The present results have implications for patient management during antiretroviral therapy, and may help to define a threshold of CD8+ cell activation below which disease progression ceases. Recovery of normal immune activation status could be a sign of viral eradication. The interest of this immunological monitoring for patients on combination therapy with undetectable level of HIV plasma RNA has to be determined in further long-term clinical studies.
The authors thank all the patients who agreed to participate to the study and their referring physicians.
1. Giorgi JV, Detels R: T-cell subset alterations in HIV-infected homosexual men: NIAID multicenter AIDS cohort study
. Clin Immunol Immunopathol
2. Levacher M, Hulstaert F, Tallet S, et al.
: The significance of activation markers on CD8 lymphocytes in human immunodeficiency syndrome: staging and prognostic value
. Clin Exp Immunol
3. Giorgi JV, Liu Z, Hultin LE, et al.
: Elevated level of CD38+CD8+ cells in HIV infection add to the prognostic value of low CD4+ T cell levels: results of 6 years follow-up
. J Acquir Immune Defic Syndr
4. Giorgi JV, Ho HN, Hirji K, et al.
: CD8+ lymphocyte activation at human immunodeficiency virus type 1 seroconversion: development of HLA-DR+ CD38
-CD8+ cells is associated with subsequent stable CD4+ cell levels
. J Infect Dis
5. Bouscarat F, Levacher-Clergeot M, Dazza MC, et al.
: Correlation of CD8 lymphocyte activation with cellular viremia and plasma HIV RNA levels in asymptomatic patients infected by human immunodeficiency virus type 1
. AIDS Res Hum Retroviruses
6. Levacher M, Tallet S, Dazza MC, et al.
: T activation marker evaluation in ARC patients treated with AZT. Comparison with CD4+ lymphocyte count in non-progressors and progressors towards AIDS
. Clin Exp Immunol
7. Bass HZ, Hardy WD, Mitsuyasu RT, et al.
: Eleven lymphoid phenotypic markers in HIV infection: selective changes induced by zidovudine treatment
. J Acquir Immune Defic Syndr
8. Kestens L, Vanham G, Vereecken C, et al.
: Selective increase of activation antigens HLA-DR and CD38 on CD4+CD45RO+ T-lymphocytes during HIV-1 infection
. Clin Exp Immunol
9. Kelleher AD, Carr A, Zaunders J, Cooper DA: Alterations in the immune response of human immunodeficiency virus (HIV)-infected subjects treated with an HIV-specific protease inhibitor, ritonavir
. J Infect Dis
10. Autran B, Carcelain G, Li TS, et al.
: Positive effects of combined antiretroviral therapy on CD4+ T cell homeostasis and function in advanced HIV disease
11. Centers for Disease Control Working Group: Guidelines for the performance of CD4+ T-cell determinations in persons with human immunodeficiency virus infection
12. Katzenstein DA, Hammer SM, Hughes MD, et al.
: The relation of virologic and immunologic markers to clinical outcomes after nucleoside therapy in HIV-infected adults with 200 to 500 CD4 cells per cubic millimeter
. N Engl J Med
13. Brun-Vezinet F, Boucher C, Loveday C, et al.
: HIV-1 viral load, phenotype, and resistance in a subset of drug-naive participants from the Delta trial
14. Mellors J, Rinaldo Jr CR, Gupta P, et al.
: Prognosis in HIV-1 infection predicted by the quantity of virus in plasma
15. O'Brien WA, Hartigan PM, Martin D, et al.
: Changes in plasma HIV-1 RNA and CD4+ lymphocyte counts and the risk of progression to AIDS
. N Engl J Med
16. O'Brien WA, Hartigan PM, Daar ES, Simberkoff MS, Hamilton JD: Changes in plasma HIV RNA levels and CD4+ lymphocyte counts predict both response to antiretroviral therapy and therapeutic failure
. Ann Intern Med
17. Liu Z, Hultin LE, Cumberland WG, et al.
: Elevated relative fluorescence intensity of CD38 antigen expression on CD8+ T cells is a marker of poor prognosis in HIV infection: results of 6 years of follow-up
18. Mocroft A, Bofill M, Medina E, et al.
: CD8+, CD38+ lymphocyte percent: a useful immunological marker for monitoring HIV-1-infected patients
. J Acquir Immune Defic Syndr Hum Retrovirol
19. Cohen OJ, Pantaleo G, Holodniy M, et al.
: Decreased human immunodeficiency virus type 1 plasma viremia during anti-retroviral therapy reflects downregulation of viral replication in lymphoid tissue
. Proc Natl Acad Sci USA
20. Lafeuillade A, Tamalet C, Poggi C, Pellegrino P, Tourres C, Izopet J: Antiretroviral effect of zidovudine–didanosine combination on blood and lymph nodes
21. Lafeuillade A, Poggi C, Profizi N, Tamalet C, Costes O: Human immunodeficiency virus type 1 kinetics in lymph nodes compared with plasma
. J Infect Dis
22. Cavert W, Notermans DW, Staskus K, et al.
: Kinetics of response in lymphoid tissues to antiretroviral therapy of HIV-1 infection
23. Levacher M, Bouscarat F, Chau F, et al.
: Evolution of CD4 and CD8 cytokine secreting lymphocytes in HIV infected patients treated with stavudine+didanosine+ritonavir
. Sixth European Conference on Clinical Aspects and Treatment of HIV Infection
. Hamburg, October 1997 [abstract 123].
24. Cohen OJ, Pantaleo G, Lam GK, Fauci AS: Studies on lymphoid tissue from HIV-infected individuals: implications for the design of therapeutic strategies
. Springer Semin Immunopathol
25. Cohen C: Stubborn HIV reservoirs vulnerable to new treatments
26. Ho HN, Hultin LE, Mitsuyasu RT, et al.
: Circulating HIV-specific CD8+ cytotoxic T cells express CD38 and HLA-DR antigens
. J Immunol
27. Fiorentino S, Dalod M, Olive D, Guillet JG, Gomard E: Predominant involvement of CD8+CD28−lymphocytes in human immunodeficiency virus-specific cytotoxic activity
. J Virol
28. Koup RA, Safrit JT, Cao Y, et al.
: Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency type 1 syndrome
. J Virol
29. Harrer T, Harrer H, Kalams SA, et al.
: Cytotoxic T lymphocytes in asymptomatic long-term nonprogressing HIV-1 infection. Breadth and specificity of the response and relation toin vivoviral quasispecies in a person with prolonged infection and low viral load
. J Immunol
30. Ferbas J, Daar ES, Grovit-Ferbas K, et al.
: Rapid evolution of human immunodeficiency virus strains with increased replicative capacity during the seronegative window of primary infection
. J Virol
31. Pantaleo G, Fauci AS: Tracking HIV during disease progression
. Curr Opin Immunol
32. Koup RA: Virus escape from CTL recognition
. J Exp Med
Françoise Chau and Bénédicte Desforges were involved in the Immunological Study. Juliette Gerbe supervised the Clinical Coordination.
Hôpital Bichat-Claude Bernard, Paris:
J.P. Coulaud, F. Vachon (Services des Maladies Infectieuses), C. Carbon (Service de Médecine Interne), D. Descamps, F. Brun-Vézinet (Service de Virologie).
Hôpital Rotschild, Paris:
W. Rozenbaum (Service des Maladies Infectieuses), N. Aïssa (Service de Virologie).