Acute infection with HIV-1 is associated with high levels of plasma viraemia [1,2] and a subsequent vigorous response to the pathogen by the immune system [3–6] In over 50% of all cases of acute HIV-1 infection, clinical manifestations occur, most often flu-like symptoms such as fever, headache or rash [7–10]. During the first weeks, the virus replicates at very high levels, often reflected in high (≥ 106/ml) HIV-1 RNA copy numbers in peripheral blood. Subsequently, virus-specific immune responses are triggered that reduce the viral load to a lower steady-state level, balancing viral production and clearance, although lack of susceptible cells can also explain the observed pattern of HIV-1 RNA decline . This viral set point occurs after, on average, 19 weeks, but with large interpatient variability . The amount of HIV-1 RNA in plasma at that time is highly correlated with subsequent clinical course [13–17].
Combinations of drugs that inhibit HIV-1 replication at different moments in the replication cycle have greatly improved clinical outcome [18,19]. The decree of treatment of HIV-1 is: ‘hit early and hard’. The question remains whether antiretroviral therapy does indeed have an additional effect during the very early stage of infection when viral load at or just after the initial peak is still declining owing to the immune response. To assess the effect of highly active antiretroviral therapy in primary infection, we compared pre- and post-treatment rates (slopes) of plasma HIV-1 RNA change in acutely infected individuals, and we compared the post-treatment slopes with those in chronically infected antiretroviral naive patients.
In total, 17 patients were selected from an ongoing study on the effect of highly suppressive anti-HIV drug combination treatment, as described previously . All patients with at least 2 months of follow-up were included in the present study. All patients started with a five-drug combination. Only chronically infected patient ERA-005 was antiretroviral therapy experienced at the start of the study; he had received zidovudine for 16.5 months and lamivudine for 10.5 months, both of which were stopped 14 days before enrolment, and didanosine for 6 months, which had been stopped 1 year prior to enrolment.
Plasma HIV-1 RNA data of the first week after the initiation of antiretroviral therapy were used for the post-therapy slope analysis. The median frequency of HIV-1 RNA measurements in this first week was four (range three to five) for the primary infection group and five (range three to six) for the chronic infection group. Changes in immunological parameters were compared for the first 2 months of therapy.
For the five acutely infected subjects, all available HIV-1 RNA data prior to the initiation of antiretroviral therapy were also considered. The median of available pre-treatment HIV-1 RNA samples was two (range one to four). All patients with primary infection presented at the clinic with symptoms associated with acute HIV-1 infection. Therapy was started within 2 to 5 weeks after the onset of clinical symptoms and when the antibody response to HIV, as measured by two commercially available enzyme immunoassays (HIVAG-1 and IMxHIV-1/HIV-2 III plus, Abbott Laboratories, Diagnostic Division, Chicago, Illinois, USA) and Western Blot (Genelabs Diagnostics, Singapore), was still incomplete in three of the five patients.
Measurement of HIV-1 RNA
HIV-1 RNA levels in serum were measured using commercially available assays (NASBA HIV-1 RNA QT and NucliSens; Organon Teknika, Boxtel, the Netherlands) according to the instructions of the manufacturer. Initially the NASBA HIV-1 RNA QT assay was used. When HIV-1 RNA levels declined to < 1000 copies/ml (the lower detection limit of the assay), levels were quantified using the NucliSens assay. Subsequently, when levels declined to < 400 copies/ml, an ultrasensitive protocol was used. Briefly, the RNA purified according to the NucliSens protocol was eluted in a two-step procedure from the silica particles. Subsequently, RNA was concentrated by precipitation with ethanol and sodium acetate in the presence of pellet paint. The complete RNA pellet was then used in the NucliSens RNA amplification and detection procedure according to the instruction of the manufacturer. When RNA levels decreased to < 50 copies/ml, an initial input volume in the assay of 2 ml plasma was used combined with the ultrasensitive protocol adaptation, resulting in a lower quantification limit of 5 copies/ml .
Genotyping of reverse transcriptase and protease
HIV-1 RNA was isolated from 200 μl ethylenediamine tetraacetic acid-treated plasma using the method described by Boom et al. . The RNA was reverse transcribed using MMLV-RT or Superscript-RT (Life Technologies, Paisley, Scotland). The protease gene and the 5′ part of the reverse transcriptase gene (containing the first 850 nucleotides of the gene) of the HIV-1 genome were amplified by polymerase chain reaction (PCR) and, in a subsequent nested PCR, protease and a 5′-reverse transcriptase PCR fragment were generated. The nested PCR fragments were sequenced on an automatic sequencer (Applied Biosystems DNA sequencer 373A stretch and 377, Foster City, California, USA) using dye-labelled primers (ET-system, Amersham, Little Chalfont, UK).
Lymphocyte subsets and activation markers
Lymphocyte immunophenotyping on fresh cells was performed by flow cytometry using dual staining. Subsets of activated CD4 and CD8 lymphocytes, defined by membrane expression of the activation antigens HLA-DR and CD38, were measured by three-colour immunofluorescence analysis .
Measurements of plasma levels of indinavir
Indinavir concentrations in plasma were measured using a high-performance liquid chromatography procedure .
Baseline parameters are presented as means and SD, as medians and interquartile ranges (IQR), or as proportions (%). Statistical comparison was based on analysis of variance (ANOVA) or Wilcoxon test, where appropriate. Correlation between baseline parameters was studied in the whole group and in the subgroups of acutely and chronically infected patients.
Linear mixed effects (LME) models (linear regression models, allowing some parameters to vary among patients, others to remain fixed over the patient population)  were applied to the logarithm of HIV-1 RNA copies/ml. Wald's log-likelihood test was used to test statistical significance of differences in intercept and slopes between the chronically and acutely infected groups. Conditional modes gave the individual fits as presented in Figs. 1 and 2.
For the subgroup of acutely infected subjects, slopes before and after the initiation of therapy were compared using non-linear mixed effect models (NLME)  with a bilinear regression function with possibly different slopes before and after therapy and common intercept. Again, a log-likelihood test was used to assess the statistical significance of difference in slopes before and after therapy.
The relationship between the rate of decline of HIV-1 RNA and baseline virological and immunological parameters was also studied. Changes in CD4 and CD8 T cell counts, the ratio of CD4 to CD8 cells (CD4/CD8) and HIV-1 RNA (log10 copies/ml) were compared with ANOVA, after testing for normality.
For the LME and NLME analysis, the statistical package S-plus (version 4.5; Mathsoft Inc., Seattle, Washington, USA) was used. For all other analyses, SAS (version 6.12; SAS Inc. Cary North Carolina, USA) was used.
Individual plasma HIV-1 RNA data were fitted to our previously described mathematical model . A set of eight parameters was common to all patients. One could vary and was group specific. Non-linear optimization techniques were used with a negative binomial maximum likelihood measure of goodness of fit (allowing for measurement error). Log-likelihood tests were used to discriminate between results from the model that allowed one parameter to vary between groups and those from the model that had all parameters fixed.
The five patients with primary infection and 12 patients with chronic infection all started antiretroviral therapy for the first time during the study (apart from one patient). The treatment regimens are shown in Table 1. All patients with primary infection presented at the clinic with symptoms associated with acute HIV-1 infection (Table 2).
Slopes of HIV-1 RNA decline during primary infection and after the start of therapy
Comparison of HIV-1 RNA slopes before and after the start of treatment for the acutely infected patients revealed a highly significant difference (P = 0.0001), with RNA decline accelerating after the start of treatment. Mean slopes per day (standard deviation) were −0.046 (0.029) and −0.182 (0.068) log10 copies/ml pre- and post-treatment, respectively. For each of the five acutely infected patients, the post-treatment slope exceeded the pre-treatment slope (Fig. 1). Slopes before and after the start of therapy were highly correlated (P = 0.0001;Fig. 1). No correlation was found between the slopes and the time from onset of clinical symptoms related to acute HIV-1 infection to start of therapy, either before or after the start of therapy (P = 0.45 and P = 0.46, respectively).
Slopes of HIV-1 RNA decline after the start of therapy in primary versus chronic infection
At the start of therapy, no statistically significant differences were found between the groups, although there was a tendency for baseline serum HIV-1 RNA values and CD8 T cell counts to be higher in the primary infection group (P = 0.07 and P = 0.11, respectively; see Table 3). A significant inverse correlation was found in the whole group between log10 HIV-1 RNA and CD4/CD8 ratio (P = 0.024). This correlation was also found in the subgroup of chronically infected patients (P = 0.011) but not in the subgroup of acutely infected patients (P = 0.50).
HIV-1 RNA levels and slopes measured in the chronically infected patients are shown in Fig. 2. The slopes over the first 7 days post-treatment were higher for the group of chronically infected subjects (P = 0.012;Fig. 3). Mean (SD) of slopes of change in log10 copies/ml HIV-1 RNA per day were −0.291 (0.051) for the chronic group versus −0.156 (0.039) for the primary group. The difference in slopes could not be attributed to differences in sensitivity to antiviral drugs or plasma drug concentrations. No drug resistance-associated mutations were found at the start of therapy in the genes for HIV reverse transcriptase and protease. Plasma indinavir levels over the first day of therapy did not differ significantly between the primary and chronic infection group: the P values for differences between primary and chronic infection groups in area under the curve (AUC) and plasma maximum and minimum concentrations were 0.41, 0.93 and 0.70, respectively.
After taking differences in slopes between primary and chronic infection into account, there was no statistically significant difference in intercept between the primary and chronic infection groups (P = 0.13). There was an inverse correlation in the whole group between the intercept in HIV-1 RNA log10 copies/ml and the subsequent linear decline per day in HIV-1 RNA log10 copies/ml (P = 0.012). This difference seems to result largely from the difference in slopes between the primary and chronic infection group, since a significant correlation was not found in the primary infection group or in the chronic infection group (P = 0.51 and P = 0.13, respectively).
A comparison of changes in HIV-1 RNA over the first 2 months of therapy showed a tendency for a larger decrease in the chronic infection group, consistent with the difference in slopes. The difference in changes in HIV-1 RNA did not reach statistical significance (P = 0.068), probably because of the lack of statistical power resulting from the small population size and the fact that two subjects in the chronic infection group reached the lower limit of quantification of the assay.
Immunological changes after the start of therapy in primary versus chronic infection
No correlation was found between the number of CD4 or CD8 T cells or the CD4/CD8 T cell ratio at baseline and the subsequent downward slope in HIV-1 RNA log10 copies/ml, in either the whole group or the subgroups of acutely and chronically infected patients. Figure 4 shows the median (IQR) of CD4 and CD8 T cell count and CD4/CD8 T cell ratio over time during the first 2 months after the start of treatment. In this period, the group of acutely infected patients experienced a significantly larger decrease in CD8 T cell count (P = 0.04), a tendency for a larger increase in CD4 T cell count (P = 0.06) and a significantly larger increase in CD4/CD8 ratio (P = 0.01;Table 4). The immunological changes were most profound in the first week after the start of therapy.
CD4 and CD8 T cell activation at baseline in primary versus chronic infection
The number and percentages of activated (HLA-DR+CD38+) CD4 and CD8 T cells at baseline were determined for the patients in the study (Table 3) as well as for a group of 1061 non-infected controls. The median (2.5–97.5% range) of activated CD8 T cells in the non-infected controls were 10 × 106 cells/l (0–70) and 2% (1–14) and of activated CD4 T cells were 20 × 106 cells/l (10–80) and 8% (2.5–20.5), respectively. Numbers and percentages of activated CD8 and CD4 T cells were higher in HIV-1 infected patients than in non-infected controls.
Although the total number of CD8 T cells did not differ significantly between the acutely and chronically infected groups, both the number and percentage of activated CD8 T cells were significantly higher (P = 0.007 and P = 0.005, respectively) in the acutely infected group. The percentage of activated CD4 T cells and the total number of CD4 T cells were similar between groups. However, the absolute number of activated CD4 T cells appeared to be higher (P = 0.02) in primary infection than in chronic infection, which is probably a result of the extremely small variability in the number of activated CD4 T cells in chronic infection.
Results of parameter fitting
Infected cell lifetime, free virus lifetime and burst size per cell per unit time were the parameters in the mathematical model that could equally well be chosen as the ‘group-specific’ parameter. While these parameters fit the data equally well, all produced the same effect: the effective number of virions produced per productively infected cell was 3.8-fold (95% confidence limit 2.1–6.5) greater for the acutely infected than for the chronically infected group. The model wherein one parameter (infected cell lifetime, free virus lifetime or burst size) was varied between groups fitted 267 χ2 (1 degree of freedom) better than the model wherein all parameters were fixed for both groups (P < 0.0001).
The difference in decline of serum HIV-1 RNA levels before and after the initiation of therapy for the five acutely infected individuals provides evidence for an additional effect of HAART in the inhibition of primary infection. Before the start of therapy, HIV-1 RNA levels were declining in all acutely infected patients, indicating that the viral set point had not yet been reached in any patient. The relatively low CD4 T cell counts found at the start of therapy further indicated that these acutely infected patients indeed started antiretroviral treatment very early , which is confirmed by the short time period between clinical signs and symptoms of acute HIV-1 infection and the first HIV-1 RNA measurement.
The rate of HIV-1 RNA decline increases significantly in each of the five patients after the start of therapy and is in accordance with other studies [28,29]. This suggests that large amounts of drug-sensitive virus have escaped the immune system early in infection, assuming the immune response plays a role at this early stage . Therapy is helping to block de novo infection of target cells and production of new infectious virus. A similar difference in HIV-1 RNA decline in treated and untreated primary infection has been reported recently by Lillo et al. , who compared non-treated patients with primary HIV-1 infection with primary HIV-1 patients taking zidovudine and combination therapy. HIV-1 RNA was more suppressed for those patients on combination therapy. Our results support their findings from a different angle, by comparing the decline of HIV-1 RNA with and without combination therapy within the same patient.
The decline in HIV-1 RNA in the first week of antiretroviral therapy was slower for the acutely than for the chronically infected patients. This slower decline was also found when the first 2 months of therapy was considered; however, in this case, the difference between the acutely and chronically infected individuals did not reach statistical significance. This may be because HIV-1 RNA levels of two of the chronically infected patients were below the level of quantification after 2 months of therapy, whereas none of the acutely infected patients reached that level at that time point.
If the post-therapy decline in HIV-1 RNA for the acutely infected patients is to be viewed as the result of the combined efforts of immune system and therapy, the slower post-therapy first-phase decline is remarkable. It seems unlikely that the observed difference in slopes between the chronic and the primary infection group can be explained completely by the amount of HIV-1 RNA present at baseline. For the whole group, there was a significant correlation between baseline HIV-1 RNA and the exponential slope in HIV-1 RNA after the start of treatment, which is inverse to that reported earlier . However, no significant correlation was found between baseline HIV-1 RNA and slope, neither within the chronically nor within the acutely infected subgroup. Although admittedly the number of patients is lower in the subgroups, and especially in the group of acutely infected patients, the P values are sufficiently large to suggest that the observed correlation between baseline HIV-1 RNA and slope should be ascribed, to a large extent, to the difference between chronic and acute infection. It is also unlikely that the slower rate of decline in the primary infected group is caused by fewer drug-sensitive HIV strains, as we did not find resistance-associated mutations in either the reverse transcriptase or the protease gene. Neither did we find a difference in plasma indinavir levels between groups, indicating that differences in the pharmacodynamics of the antiretroviral drugs were of limited importance. We therefore concluded that the difference in the slope of HIV-RNA decline reflects merely the difference in HIV dynamics between the primary and chronic infections.
Previously, using a mathematical model of HIV-1 infection , we have demonstrated that heterogeneity in proliferation rates in CD4 cells can be explained by antigen-induced stimulation of the immune system. In this model, activation of target cells (CD4 T cells) is associated with higher proliferation rates, increasing susceptibility to HIV-1 infection and a higher production rate of new infectious HIV-1. Using this mathematical model, we could explain the slower HIV-1 RNA decline after the start of therapy found in the acutely infected patients by assuming a lower level of activation of CD4 T cells. However, this assumption could not be substantiated as no difference was found between acutely and chronically infected patients with respect to the percentage of HLA-DR+CD38+ CD4T cells, which for both groups was significantly higher than those in HIV-negative controls. Based on fitting to patient HIV-1 RNA data, the model predicted the effective number of virions produced per activated cell to be greater in the acutely than in the chronically infected group. Three factors, namely infected cell lifetime, free virus lifetime and burst size per cell per unit time, determine the number of virions produced. We do not assume the free virus lifetime to be different between primary and chronic infection. However, the lifetime of productively infected cells (and thereby the burst size per cell per unit time) may be longer in primary infection. This may be the result of a still growing immune response, allowing infected cells to live longer and produce more virus per unit time and giving HIV-1 the opportunity to establish the infection, despite antiretroviral therapy.
1. Clark SJ, Saag MS, Decker WD. et al
. High titers of cytopathic virus in plasma of patients with symptomatic primary HIV-1 infection.
N Engl J Med 1991, 324: 954 –960.
2. Daar ES, Moudgil T, Meyer RD, Ho DD. Transient high levels of viremia in patients with primary human immunodeficiency virus type 1 infection.
N Engl J Med 1991, 324: 961 –964.
3. Koup RA, Safrit JT, Cao Y. et al
. Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency virus type 1 syndrome.
J Virol 1994, 68: 4650 –4655.
4. Borrow P, Lewicki H, Hahn BH, Shaw GM, Oldstone MBA. Virus-specific CD8+ cytotoxic T-lymphocyte activity associated with control of viremia in primary human immunodeficiency virus type 1 infection.
J Virol 1994, 68: 6103 –6110.
5. Rosenberg ES, Billingsley JM, Caliendo AM. et al
. Vigorous HIV-1 specific CD4 T-cell responses associated with control of viremia.
Science 1997, 278: 1447 –1450.
6. Musey L, Hughes J, Schacker T, Shea T, Corey L, McElrath MJ. Cytotoxic-T-cell responses, viral load, and disease progression in early human immunodeficiency virus type 1 infection.
N Engl J Med 1997, 337: 1267 –1274.
7. Cooper DA, Gold J, Maclean P. et al
. Acute AIDS retrovirus infection: definition of a clinical illness associated with seroconversion.
Lancet 1985, i: 537 –540.
8. de Wolf F, Lange JMA, Bakker M. et al
. Influenza-like syndrome in homosexual men: a prospective diagnostic study.
J Coll Gen Pract 1988, 38: 443 –446.
9. Lange JM, Parry JV, de Wolf F, Mortimer PP, Goudsmit J. Diagnostic value of specific IgM antibodies in primary HIV infection.
AIDS 1988, 2: 31 –35.
10. Schacker T, Collier AC, Hughes J, Shea T, Corey L. Clinical and epidemiologic features of primary HIV infection.
Ann Intern Med 1996, 125: 257 –264.
11. Phillips AN. Reduction of HIV concentration during acute infection: independence from a specific immune response.
Science 1996, 271: 497 –499.
12. Kaufmann GR, Cunningham P, Kelleher AD. et al
. Patterns of viral dynamics during primary human immunodeficiency virus type 1 infection.
J Infect.Dis. 1998, 178: 1812 –1815.
13. Henrard DR, Phillips JF, Muenz LR. et al
. Natural history of HIV-1 cell-free viremia.
JAMA 1995, 274: 554 –558.
14. Ho DD. Viral counts count in HIV infection.
Science 1996, 272: 1124 –1125.
15. 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.
16. de Wolf F, Spijkerman I, Schellekens PTh. 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, 11: 1799 –1806.
17. Schacker TW, Hughes JP, Shea T, Coombs RW, Corey L. Biological and virological characteristics of primary HIV infection.
Ann Intern Med 1998, 128: 613 –620.
18. Fischl MA, Hammer SM, Hirsch MS. et al
. Antiretroviral therapy for HIV infection in 1997.
:Updated recommendations of the International AIDS Society-USA panel.
JAMA 1997, 277: 1962 –1969.
19. Weverling GJ, Lange JMA, Jurriaans S. et al
. Alternative multidrug regimen provides improved suppression of HIV-1 replication over triple therapy.
AIDS 1998, 12: F117 –F122.
20. Ho DD. Time to hit HIV, early and hard.
N Engl J Med 1995, 333: 450 –451.
21. Prins JM, Jurriaans S, van Praag RME. et al
. Immuno-activation with anti-CD3 and recombinant human IL-2 in HIV-1-infected patients on potent antiretroviral therapy.
AIDS 1999, 13: 2405 –2410.
22. Boom R, Sol CJA, Salimans MMM, Jansen CL, Wertheim-van Dillen PME, van der Noordaa J. Rapid and simple method for purification of nucleic acids.
J Clin Microbiol 1990, 28: 495 –503.
23. 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 1994, 170: 775 –781.
24. Foisy ML, Sommadossi JP. Rapid quantification of indinavir in human plasma by high-performance liquid chromatography with ultraviolet detection.
J Chromatogr B Biomed Sci Appl 1999, 721: 239 –247.
25. Laird NM, Ware JH. Random-effects models for longitudinal data.
Biometrics 1982, 38: 963 –974.
26. Lindstrom MJ, Bates DM. Nonlinear mixed effects models for repeated measures data.
Biometrics 1990, 46: 673 –687.
27. Ferguson NM, de Wolf F, Ghani AC. et al
. Antigen-driven CD4+ T cell and HIV-1 dynamics: residual viral replication under highly active antiretroviral therapy.
Proc Natl Acad Sci USA 1999, 96: 15167 –15172.
28. Cooper DA, Perrin, L, Kinloch, S. Intervention with quadruple HAART [Combivir (COM)/Abacavir (ABC)/Amprenavir (APV)] Intervention during primary HIV-1 infection (PHI) is associated with rapid viremia clearance and decrease of immune activation. Seventh Conference on Retroviruses and Opportunistic Infections.
San Francisco, 2000 [abstract 532].
29. Markowitz M., Vesanen M., Tenner-Racz K. et al
. The effect of commencing combination antiretroviral therapy soon after human immunodeficiency virus type 1 infection on viral replication and antiviral immune responses.
J Infect Dis 1999, 179: 525 –537.
30. Lillo FB, Ciuffreda D, Veglia F. et al
. Viral load and burden modification following early antiretroviral therapy of primary HIV-1 infection.
AIDS 1999, 13: 791 –796.
31. Notermans DW, Goudsmit J, Danner SA, de Wolf F, Perelson AS, Mittler J. Rate of HIV-1 decline following antiretroviral therapy is related to viral load at baseline and drug regimen.
AIDS 1998, 12: 1483 –1490.
Keywords:© 2000 Lippincott Williams & Wilkins, Inc.
antiretroviral therapy; HIV clearance; viral load; acute infection; chronic infection