We aimed to investigate the impact of either early, or delayed ART during primary intravenous infection of macaques with pathogenic SIVmac251. Interestingly, during this period, there is a clear relationship between peripheral and tissue viral loads both in placebo and PEP-treated animals for a given date. Our findings favor the initiation of ART before the viremia peak, even one week after infection. ART initiated after the peak of viral replication showed no impact on viral spread, due to early dissemination.
Very low levels of viral replication (SIV RNA) and dissemination (DNA) were found in mucosal tissues, if ART was started four hours after infection (below the threshold in four animals out of 5; ). The immunological benefit of preventing CD4+ T cell depletion in LN and mucosal tissues is obvious, as for acute infections: the mucosa is the dominant site of infection, and the gastrointestinal tract/other mucosal tissues contain at least half of the body's T cells 
. In macaques infected with SIV, intestinal CD4+ T cells are almost entirely depleted within three weeks of infection 
. Despite a few studies on the most acute stages of HIV-1 infection in humans, it is likely that there is a similar large and rapid loss of intestinal CD4+ T cells during the early periods of infection 
ART started seven days after infection continued to have an impact on both viral replication (RNA) and dissemination (DNA) in the gastrointestinal tract. Moreover, viral dissemination was lower in mucosal tissues than in LN. Nevertheless, delayed treatment did not stop the depletion of CD4+ T lymphocytes in the rectum, as shown by results on day 21. Verhoeven D et al
., treating SIV-infected rhesus macaques with PMPA+FTC with the same schedule (from 7 days pi), reported similar acute CD4+ T cell loss two weeks after infection in the jejunum 
ART that was started 14 days pi did not result in a significant decrease of SIV DNA and RNA in the rectum. Macaques treated with PMPA eight weeks after infection have slightly lower intestinal (samples from jejunum and colon) SIV RNA levels 
. Despite a clear activation of the immune system and modification of lymph-node architecture, which is typical of this infection 
, CD4+T lymphocyte localization seemed to be more conserved and the number of CD4+ T lymphocytes remained higher in treated animals in LN and in gut associated lymphoid tissues (GALT) than in placebo-treated animals.
Observational studies in humans, along with limited and contradictory studies using animal models, suggest that there may be a window of opportunity for initiating ART to preserve mucosal CD4 T cells or to allow a complete repopulation. Mehandru et al.
found that therapy failed to significantly repopulate mucosal CD4+ T cells in eight HIV-infected subjects who started ART early during primary infection 
. Tincati et al.
evaluated the kinetics of CD4+ T-cell decrease and ART-mediated immune reconstitution in the gastrointestinal tract of nine patients during the acute phase of HIV infection, by performing rectosigmoid colonic biopsies before and after six months of ART 
. Time from symptoms to therapy ranged from 13 to 49 days (mean 26 days) and the regimen most often used was identical to the one used in this study (AZT+3TC+IDV). HIV DNA was only marginally reduced in the gastrointestinal tract; this was associated with persistent immunological impairment in GALT. By contrast, Guadalupe et al. demonstrated that one patient who started ART within six weeks of infection showed significant mucosal repopulation 
. Verhoeven D et al
., using rhesus macaques infected with SIV and treated with FTC (emtricitabine)/PMPA at 1 weeks of infection over a period of 30 weeks, showed, despite major suppression of viral RNA levels in GALT, that ART led to a restoration of CD4+T cell levels in this tissue, in comparison with placebo-treated animals 
. Thus, ART may have been initiated too late, as a severe (and partially not reversible) depletion of CD4+ T cells in GALT had already occurred; this contrasts with the rapid and complete restoration of CD4+ T-cell levels in GALT, which is observed when the same regimen is initiated one week pi 
. This suggestion is supported by a report by George M.D. et al.; they studied mucosal CD4+T-cell restoration in a model of rhesus macaques intravenously infected with SIVmac251
and given PMPA or placebo six weeks after infection, with PMPA given over the course of 20 weeks 
. They demonstrated a marked suppression of mucosal viral loads and rapid reconstitution of CD4+T cells, in GALT of animals receiving ART. Treatment initiated one week after infection, using our combination of drugs (AZT/3TC/IDV) and over a period too short to be fully effective, reduced viral dissemination, but had no effect on LN hyperplasia. Despite the impact on viral replication and dissemination in lymph nodes and mucosal tissues, we were unable to show a clear benefit toward target cell depletion in these tissues. It may be related to a lack of potency in the treatment, and/or to the short duration of the experiment. Previous analysis of B cell functionality and distribution in these animals on day 14 (early ART) showed that ART has only a minimal effect on early Ig production and B cell distribution; however, ART stops the increase in germinal center growth and other changes observed on day 28 (D14-28 ART) in mesenteric LN and spleen 
. Taken together, these data support a limited but real immunological benefit associated with ART treatment, even if our understanding of the underlying mechanisms requires further study.
Limitations in our study included difficulty in accurately defining the significance of the number of target cells in animals, as we did not know the status (resting/activated) of these target cells, and difficulty in accurately determining the level of viral production, despite our PCR evaluation. As all the animals were killed, another limit of our study was the inability to study the natural long-term outcome of the animals, such as cellular repopulation by target cells in the various tissues (peripheral lymph nodes, rectum and lung).
The individual virological (reduction of total SIV DNA in lymphoid organs, despite lack of protection) and public health (decreasing the infectiousness of HIV in patients with acute HIV-1 infection) benefits of delayed therapy, initiated before the peak of viremia, even one week after infection, should be taken into account for persons seeking care 72 hours after HIV exposure. Further studies should include time points between 48 and 72 hours, a drug regimen resembling current HIV PEP, and long-term experiments with or without the cessation of treatment, to study the potential for long-term change in viral replication patterns and clinical outcome by stopping or limiting the primary infection