Discordant responses during antiretroviral therapy: role of immune activation and T cell redistribution rather than true CD4 T cell loss

We studied T cell dynamics in four patients who initially responded well to highly active antiretroviral therapy (HAART) but subsequently experienced virological failure. Maintenance of peripheral blood CD4T cell counts was associated with low levels of immune activation. Low reactivity to rebounding virus may preserve normal T lymphocyte distribution over blood and tissues and be associated with stable peripheral blood T cell numbers in virological failures to HAART. A proportion of HIV-1-infected individuals receiving highly active antiretroviral therapy (HAART) fails to suppress the virus completely, or goes through one or more periods of virus relapse after initial adequate control. Virus rebound is usually associated with a decline in peripheral blood CD4 T cell numbers; however, a subset of patients was found to maintain peripheral blood CD4 T cell counts or even experienced an increase in CD4 T cell numbers in the blood, despite high levels of plasma HIV-1 RNA 1–4. This may be attributable to protease inhibitors (PI), which were reported to reduce the susceptibility of CD4 T cells to apoptosis 5,6, or that may interfere with cell cycle progression and thereby enhance T cell survival 7,8. Alternatively, PI or reverse transcriptase (RT)-resistant virus strains may be less cytopathic to mature or immature (intrathymic) CD4 T cells 9,10. Alternative mechanisms remain to be explored, such as the role of immune activation and T cell redistribution. We studied the association between plasma HIV-1-RNA levels, naive, memory and effector CD4 and CD8 peripheral blood T cell numbers and T cell activation longitudinally in patients who initially adequately suppressed the virus, but subsequently experienced virological failure despite continued HAART. We selected participants of the Amsterdam Cohort Studies on HIV-1 Infection and AIDS, of whom sequential cryopreserved peripheral blood mononuclear cells were available before and during HAART, and who initially experienced sustained virus suppression (plasma HIV-1 RNA less than 400 copies/ml for more than 6 months) but subsequently developed virological failure (plasma HIV-1 RNA ≥ 1 log increase) for at least 6 months. Only four patients were identified who complied with these strict criteria, as in most virological failures treatment is adjusted within 6 months and the increase in plasma HIV-1 RNA is less pronounced. Selected individuals were men and the median age at initiation of this study was 31.5 years (range 22.8–46 years). All patients were HIV-1 seropositive when they entered the cohort; follow-up was between 40 and 70 months. Details of the HAART regimens are given in Fig. 1. T cell division was assessed by measuring the expression of Ki67 (Immunotech, Marseille, France), a protein that is expressed exclusively by cells that are in cell cycle, and naive (CD27+ CD45RO−), CD27+ memory (CD27+ CD45RO+), CD27− memory (CD27− CD45RO+) and effector (CD27− CD45 RO−) CD4 and CD8 T cells were defined as described previously 11. Plasma HIV RNA was assessed using Roche Amplicor Monitor Standard Assay or Ultra Monitor Assay (Roche Diagnostics, Branchburg, NJ, USA), NucliSens HIV-1 QT assay and nucleic acid sequence-based amplification HIV-1 RNA QT (Organon Teknika, Boxtel, the Netherlands), or Quantiplex HIV-1 RNA 3.0 branched DNA monitor assay (Chiron Corp., Emeryville, CA, USA).Fig. 1.: (a–d) Plasma HIV-1 RNA, CD4 T cell numbers and the proportion of Ki67-positive CD4 T cells of four patients during untreated HIV-1 infection, during highly active antiretroviral therapy and during subsequent virological failure. –– number of CD4 T cells (per μl blood); – – – plasma HIV-1 RNA (copies per ml); vertical grey bars: Ki67-positive CD4 T cells (%); horizontal grey bars: period during which each patient was treated with highly active antiretroviral therapy (HAART). Characters represent drugs: Z, zidovudine; L, lamivudine; SA, saquinavir; N, nelfinavir; D, didanosine; ST, stavudine; I, indinavir; R, ritonavir. (e) Rebound of immune activation during virological failure to pre-treatment levels. Depicted are the proportions of Ki67-positive CD4 T cells before the initiation of HAART (pre-HAART), during successful virus suppression (HAART) and during virus rebound (VF) of patients 6181 (•), 6156 (▪), 3558 (▴), and patient 8324 (♦). For comparison, median proportions of Ki67-positive CD4 T cells of a group of untreated HIV-1-infected individuals (grey bar, n = 16) and a group of healthy donors (white bar, n = 5) are shown 11. (f) Correlation between the activation of CD4 and CD8 T cells. For each patient the proportion of Ki67-positive CD4 T cells is plotted against the proportion of Ki67-positive CD8 T cells at all timepoints. Symbols are similar to those in (a).Individual data are depicted in Fig. 1a–d. All patients initially had high plasma HIV-1-RNA concentrations that rapidly reached less than 400 copies/ml when HAART was initiated. CD4 T cell numbers increased and Ki67 expression declined in this period. During subsequent virological failure, two patterns could be distinguished. In patients 6181 and 6156, high plasma HIV-1-RNA levels were associated with increased proportions of Ki67-positive CD4 T cells and a decline in CD4 T cell numbers (Fig. 1a–b), whereas in patients 3558 and 8324, virus rebound was associated with relatively low proportions of Ki67-positive CD4 T cells and stable CD4 T cell numbers (Fig. 1c–d). We plotted the fraction of Ki67-positive CD4 T cells measured at three timepoints (before the initiation of HAART, during virus suppression, and during virus rebound) of the four patients, and compared these with the proportion of Ki67-positive CD4 T cells of a group of untreated HIV-1-infected individuals and a group of healthy laboratory donors 11 (Fig. 1e). The proportions of dividing CD4 T cells of patients 3558 and 8324 that were relatively low before the initiation of HAART remained low during virological failure, whereas patients 6156 and 6181 rapidly rebounded to high levels of CD4 Ki67 expression. The proportion of Ki67-positive CD4 T cells was associated with the proportion of Ki67-positive CD8 T cells for each individual at all timepoints (Fig. 1f). In all patients, Ki67 expression of naive, memory and effector T cell subsets followed patterns that were parallel to each other (data not shown). HIV-1 infection characteristically leads to hyperactivation of the immune system, reflected in upregulated cytokine levels and lymphocyte homing receptors. This has been shown to increase the proportion of T cells migrating to lymphoid tissues 12. In the first weeks after the initiation of HAART, when plasma HIV-1 RNA is significantly reduced, immune hyperactivation diminishes and previously sequestered lymphocytes are rapidly released into the circulation, leading to a normalized distribution of lymphocytes between blood and tissues 13. We observed relatively rapid changes in peripheral blood CD4 T cell numbers involving both naive and memory T cells after treatment failure, which reversed quickly during the subsequent successful adjustment of treatment. The decline in peripheral blood CD4 T cell numbers in patients who responded to rebounding virus with high proportions of Ki67-positive dividing T cells may be related to increased activation-induced sequestration of T cells in lymphoid tissues. Interestingly, the maintenance of peripheral blood CD4 T cell numbers despite virological failure was associated with limited immune activation, which may result in low activation-induced T cell sequestration to lymphoid tissues. Peripheral blood CD4 T cell counts are commonly used to evaluate HIV disease progression. It has, however, been shown that in untreated HIV-1 infection the decline in peripheral blood CD4 T cell numbers may overestimate the true CD4 T cell loss, as depletion is less pronounced in lymphoid tissue 14. Our data suggest that this may also be the case during virological failure to HAART. In fact, these data and those from other laboratories 15 indicate that the transient reduction in peripheral blood CD4 T cell numbers during structured treatment interruptions in chronically HIV-1-infected patients may be caused by immune activation and increased tissue sequestration, rather than to a true loss of T cells. Mette D. Hazenberga Sigrid A. Ottoa Ferdinand W. N. M. Witb,c Joep M. A. Langeb,c Dörte Hamanna Frank Miedemaa,c