Background
Despite expanding use of antiretroviral therapy (HAART) [
1], which has clearly extended lives of persons infected with human immunodeficiency virus (HIV) [
2,
3], the virus continued to spread worldwide at nearly 5 million new infections in 2005 [
4]. Therefore, there is a need to revisit proven strategies of HIV prevention with a goal to understand their limitations and maximize their effectiveness. A strategy of post exposure prophylaxis (PEP) using highly potent antiretroviral drugs is effective in preventing human immunodeficiency virus (HIV) transmission in clinical situations where treatment can be started immediately after virus exposure. For example, in preventing vertical transmission of HIV from HIV-infected mothers to their infants [
5,
6], following occupational exposure to HIV in blood and body fluids from HIV-infected persons [
7,
8] or following sexual assault or intravenous drug use [
9,
10]. Nevertheless, major barriers to the success of the program are uncertainty as to the time of virus exposure and poor compliance in completing treatment regimen, partly due to drug toxicity [
9‐
11]. Therefore, a regimen of pre-exposure prophylaxis is being evaluated for preventing HIV infection in high-risk, HIV-negative persons, such as sex workers whereby highly potent antiviral drugs are taken before high-risk behavior [
9,
12]. The rationale for pre or post exposure prophylaxis is that after HIV exposure there is a brief window of time, before the virus spreads systemically throughout the lymphoid organs, when initiating potent antiretroviral therapy might prevent or modify viral replication. In clinical settings in which compliance to treatment is poor and a potential exist for re-exposures to virus, PEP should at least reduce virus to a level sufficient to stimulate protective immune response such as antiviral CD8+ cells and thus reduce the probability of establishing persistent, productive infection. The efficacy of such regimen depends on the timing and duration of treatment, use of highly potent antiretroviral drugs and by immune responsiveness of the host [
13,
14].
We showed previously that early treatment with [(R)-9-(2-phosphonylmethoxypropyl)adenine] (PMPA) can completely prevent SIV
mne infection in cynomolgus macaques if treatment begins within 24 hours post-inoculation (p.i.) and is continued uninterrupted for 4 weeks, but is less effective if the initiation of treatment is delayed or if the duration of treatment was shortened [
15,
16]. The highest efficacy achieved required an effective regimen (i.e. 24-hour p.i., 28-day treatment) that maintained therapeutic levels of PMPA to block the spread of virus, perhaps with a contribution from antiviral immune response. The less effective regimens such as delayed initiation of PMPA treatment or shortened duration of PMPA-treatment revealed the contribution of immune response to efficacy. These regimens resulted in either delayed establishment of virus infection or induced viral control by macaques leading to transient infections [
15]. Additionally, although the PMPA-protected macaques remain free of detectable virus or SIV antibody response, they show partial resistance to challenges with homologous or heterologous SIV [
17,
18]. Even after the onset of SIV infection in macaques, the initiation of PMPA treatment during primary infection can induce immune suppression of SIV infection [
19‐
23] mediated by CD8+ lymphocytes [
18,
24]. These studies indicated that regimens of early PEP with PMPA induce antiviral immune responses in macaques to control subsequent virus infection. The CD8+ lymphocyte-mediated control of virus replication is also a major mechanism by which early antiretroviral treatment of acute HIV-1 infection induces immune control of viral replication in HIV-1 infected persons [
25]. The high potency of early PMPA treatment may be due to the rapid intracellular phosphorylation of PMPA into its active metabolites and the long half-life of these active metabolites [
26], which disrupts the replication cycle of virus.
In the present study, we revisited our post exposure chemoprophylaxis against acute SIV
mne infection in cynomolgus macaques [
15‐
17] as an animal model for studying factors involved in efficacy of early antiretroviral intervention in HIV infection. We evaluated the impact of one or more interruptions of PMPA-treatment plus re-exposures to virus on the prevention of SIV
mne infection and induction of CD8+ lymphocyte-mediated suppression of viremia in cynomolgus macaques. We also evaluated whether the efficacy of early PMPA treatment is greatest in macaques with pre-existing immune response to SIV.
Discussion
A 4-week, uninterrupted treatment using PMPA can completely block SIV from establishing infection in macaques if treatment is started within 24 hours after intravenous SIV inoculation, but is less effective if the initiation of treatment is delayed or if the duration of treatment is shortened [
15,
16]. In the present study, we show that even when treatment begins 24 h p.i. a single interruption and virus challenge was as sufficient as multiple interruptions plus viral challenges in reducing the efficacy of PMPA, instead results in persistent SIV antibody responses and long-term CD8-cell mediated suppression of virus infection. The highest efficacy showed by a 4-week, uninterrupted treatment regimen was most probably due to maintenance of effective therapeutic level of PMPA necessary to allow the infected cells initiated within 24 hours of intravenous inoculation to decay completely without spreading virus infection. However, the reduced efficacy showed by interruption of treatment plus virus challenge was most likely due to decreased PMPA levels, which allowed transient infections to occur, thereby re-setting decay of infected cells. As a result the level of SIV infection in macaques was not sufficient to establish full infection but was sufficient to induce persistent SIV antibody responses and CD8 cell-mediated suppression of virus infection. In the present study, the first interruption of treatment occurred after a 5-day treatment from the time of SIV
mne inoculation and lasted 3 days during which macaques were re-inoculated with SIV
mne. It is possible that, for the virus-negative and antibody negative macaques (V-Ab-PEP macaques in groups B and C, if treatment was interrupted later (e.g. on day 20 of treatment) during the 28-day treatment or if the interruption lasted less than three days (e.g. 1–2 days), the efficacy of PMPA would have been preserved due to the long intracellular half-life of the active metabolites of PMPA [
26].
The findings that a majority of previously seronegative macaques developed SIV antibodies 4–8 weeks after treatment in the absence of detectable viremia or after transient viremia indicated that these macaques developed control of SIV replication by the time treatment was withdrawn, consistent with previous studies [
19‐
23]. In addition, the findings that all the previously, weakly and strongly seropositive macaques (V
-Ab
± PEP macaques) developed high titers of SIV antibodies within 2 weeks of SIV inoculation, even before the end of treatment and in complete absence of detectable viremia, indicated immune memory response [
30] to SIV in these macaques. Therefore, these results indicate that the efficacy of post exposure prophylaxis with PMPA can be significantly augmented by the pre-existing immune responses to SIV. This finding is consistent with that of other investigators that showed that the efficacy of PMPA against acute or chronic SIV infection in macaques is enhanced by the presence of CD8+ T-cells [
24]. PMPA itself can stimulate lymphocytes or macrophages to secrete cytokines such as tumor necrosis factor (TNF) or chemokines such as RANTES [
31,
32], which may have also contributed to the efficacy of PMPA.
Unlike PMPA, other antiretroviral agents such as AZT or PMEA when given in post exposure regimen are incompletely effective in blocking acute SIV infection in macaques [
33‐
35]. The high potency of PMPA may be attributable to its rapid intracellular phosphorylation to form active metabolites, the long intracellular half-life of these active metabolites [
26] and perhaps a capacity of PMPA to activate immune cells such as monocytes or lymphocytes to secrete cytokines and chemokines [
31,
32]. Therefore, efficacy of post exposure prophylaxis for HIV infection may depend critically not only on the timing of initiation and duration of treatment, but also on the pharmacological properties of specific antiretroviral agents used. Good candidates include highly potent antiretroviral agents such as PMPA, which are easily activated
in vivo, have long intracellular half-life of active drug, could activate immune system and have good safety profile.
Induction of CD8 cell-suppression of viremia is one of the mechanisms by which antiretroviral treatment, including PMPA, induces host control of virus infection [
13,
18,
24,
25]. We previously demonstrated that PEP macaques exhibit considerable CD8+ lymphocyte suppression of SIV
mne in vitro even at CD8:CD4 ratios of 1:2 and mild suppression at ratios 1:10 [
17]. Our studies show that interrupted PMPA treatment resulted in CD8+ cell-suppression of viremia that persisted for more than 2 years, even in macaques that showed no evidence of viremia. (Whether macaques previously had intermittent viremia or no detectable viremia, the results were the same: plasma viral RNA increased and then decreased to undetectable levels inversely with the levels of CD8+ lymphocytes in peripheral blood). These results demonstrate the persistence of CD8+T-cell suppression of virus infection. At the same time these results demonstrate the long-term persistence of virus in the macaques, ready to replicate immediately after removal of CD8+ T cells even in macaques without any detectable virus. Such a persistence of virus in macaque may itself be a stimulant maintaining CD8 T-cell suppression of virus replication in the macaques. These results are consistent with those of other investigators showing persistence of vaccine virus as the immune correlate of protection against late onset of AIDS in macaques [
36].
A previous study showed that when a 28-day PMPA PEP regimen is used against SIV
smE660 infection in rhesus macaques, it can induce CD8+ cell-mediated control of viral replication and resistance to homologous challenge or heterologous challenge with SIV
mac39 [
18]. However, a similar regimen fails if SIV
mac239 is used as the infecting virus in PMPA PEP [
37]. Thus, the results of PMPA PEP using SIV
mne in cynomolgus macaques in the present study may not apply necessarily to other SIV isolates or other species of macaques, or to antiretroviral regimens such as pre-exposure prophylaxis that completely block virus infection without inducing CD8+ immune responses. However, the results in the present study were obtained using SIV
mne infection in cynomolgus macaques under specified conditions in which (i) PMPA treatment is started 24 hours after SIV
mne inoculation, (ii) the first treatment interruption plus SIV
mne challenge was started after a 5-day treatment, (iii) each treatment interruption lasted 3 days and macaques were challenged with SIV
mne on the second day of interruption, and (iv) treatment was continued for at least 28 days after the last SIV
mne challenge.
One question was how much the preexisting immunity contributed to the viral control in the PEP-macaques in the present study. This question was addressed by comparing the outcome of SHIV
89.6P challenge in two PEP-macaques (95020 and M94312) that had similar outcomes from first PMPA PEP, but one macaque M94312 that received second PMPA PEP regimen whereas one (macaque 95020) did not. After the initial PMPA PEP both macaques had no detectable viremia, but they had very weak SIV-specific antibody response. Four years later, macaque M94312 was given second PEP regimen and thereafter became persistently virus-negative but strongly positive for SIV antibody response. When challenged intravenously with 10 AID
50 SHIV
89.6P. macaque 95020 failed to control active viral replication and depletion of CD4+ T cells throughout the course of primary infection. In contrast, macaque M94312 completely controlled viral replication within 4 weeks of inoculation and completely blocked the depletion of CD4+ T cells. Although the number of macaques in this challenge study was small, the marked viral control by M94312 in contrast to macaque 95020 was most probably a contribution of second PMPA PEP regimen. It is conceivable that the individual PEP-macaques or macaque groups such as V
-Ab
-, V
-Ab
± or V
-Ab
+ had different levels of pre-existing SIV specific CD8+ cell responses which contributed to the protection of macaques independent of PMPA. Van Rompay et al [
24] demonstrated that CD8+ T cells enhanced the efficacy of PMPA treatment in controlling SIV infection. Therefore, in the presence of PMPA treatment, preexisting immune response might interact additively with PMPA to control infection. The outcome would depend on the level of pre-existing immunity [
14]. Although our study has a limitation in its capacity to establish the absolute contribution in each macaque group, it provides an insight into relative contribution to efficacy by comparing the results of efficacy between macaque groups. For example, our findings show that all the seronegative (V
-Ab
-) PEP macaques in groups B and C seroconverted only when PMPA treatment had stopped for at least 4 – 8 weeks, irrespective of 5-week or 10-week duration of treatment. Similarly, the onset of transient or intermittent viremia in the three macaques in those groups also developed 2 – 8 weeks after stopping PMPA treatment. In contrast the macaques in group D that received a similar treatment regimen, but which were previously weakly seroposotive, became fully seroconverted within two weeks of SIV inoculation, even before the end of PMPA treatment. These findings suggest strongly that, viral control was more dependent on PMPA treatment in the previously seronegative PEP macaques than in the previously weakly or strongly seropositive PEP macaques.
Overall, our studies confirm that early initiation of potent antiretroviral treatment and strict compliance to the duration of treatment are critical factors for the success of post exposure prophylaxis. The initiation of PMPA treatment during chronic SIV infection fails to control the rebound of viremia after treatment is withdrawn [
38]. However, the initiation of PMPA treatment between 2 days to 6 weeks after SIV inoculation (i.e. during the primary phase of acute SIV infection) results in partial immune control of SIV rebound [
19‐
23]. Yet, for a preventive treatment against HIV to be effective, the treatment must block the HIV from ever establishing systemic infection in the lymphoid organs [
39]. We have found that the most effective regimen for blocking SIV infection in macaques is a 4-week, uninterrupted treatment using PMPA beginning before or within 24 hours of SIV intravenous inoculation [
15,
16]. Our study shows that if the PMPA treatment is interrupted and macaques are re-exposed to virus, then the complete protection from SIV infection is abrogated. Despite the lack of complete protection, a majority of macaques still benefit by developing long-term immune control of virus infection. More over the use of structured treatment interruptions of highly active antiretroviral therapy in HIV-infected persons [
13,
25] or in SIV-infected macaques [
22] has been shown to induce the immune control virus infection. The effectiveness of such regimens depends on early initiation of treatment, the potency of the antiretroviral drug, and the immune responsiveness of the host [
14]. In addition, a recent clinical trial in human found that under controlled conditions of CD4-guided treatment, scheduled treatment interruptions may also be beneficial in reducing the total amount of antiretroviral drug used, and thus reduce cost as well as prevent drug toxicity without the risk of developing drug resistance and loss of efficacy of treatment [
40]
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
PE carried out the experiments, performed data analysis and wrote the initial drafts of the manuscript, YJ helped with laboratory experiments and data interpretation, MA helped with experimental design and data interpretation, BT helped with laboratory and animal experiments and data analysis, GB assisted with animal experiments. CCT was responsible for the overall experimental design, data interpretation and implementation of the project.