In this study, we investigated the relationship between PrEP antiviral activity and protection by using a repeat-exposure macaque model that closely resembles human transmission. At dosing equivalent to that used in humans, we found that daily FTC was partially protective, and that the addition of TDF increased effectiveness. We also show that subcutaneous FTC and high-dose tenofovir completely blocked rectal transmission. These findings show that full protection against repeated exposures by daily PrEP is possible in a primate model, and that PrEP effectiveness correlates with increases in antiviral activity.
None of the protected animals showed any evidence of transient systemic infection, likely reflecting the ability of FTC and tenofovir to effectively block the earliest SHIV infections, possibly at the mucosal point of entry, before significant systemic dissemination of the virus occurs. Since inhibiting virus establishment shortly after exposure may be critical to PrEP efficacy, we also explored if PrEP given intermittently around each virus exposure could be equally protective to daily PrEP. Earlier studies in newborn macaques exposed orally to a highly virulent SIVmac251 strain have shown the promise of short PrEP interventions [33
]. Intermittent PrEP modalities are highly desirable because of their convenience, potential cost-effectiveness, and lower risks of drug toxicity. Both FTC and tenofovir have long (40 to >60 h) intracellular half-lives in humans [16
] suggesting the possibility of extended prophylactic activity when administered around virus exposures. We found FTC/tenofovir given intermittently as a two-dose PrEP around each of 14 weekly virus exposures to be as fully protective as the same regimen given daily. Therefore, intermittent PrEP with potent regimens are highly promising modalities. Evaluation of intermittent PrEP with different drug combinations, possibly of different classes, and defining minimal dose requirement and optimal timing relative to virus exposure will all be important.
While many biologic similarities exist between rectal and vaginal HIV transmission, some differences in the early events of mucosal infection and dissemination kinetics are possible [8
]. Therefore, it is important to confirm the PrEP efficacy of these regimens against vaginal transmission in appropriate macaque models. Although daily PrEP with a Truvada-equivalent dosing was highly effective, a regimen with more tenofovir was required to completely block transmission in this model. However, the dose of tenofovir in this regimen would likely be toxic in humans. More work in macaque models could possibly identify two- or three-drug combinations that are fully protective and yet carry low risks of toxicity. The increasing availability of new drugs in different classes such as those that block viral integration or entry through CCR5 will provide additional possibilities. The results of such animal studies may help guide designs of clinical trials that will ultimately measure effectiveness of various PrEP regimens against sexual HIV transmission.
Initial macaque studies with tenofovir used single and non-physiologic doses of SHIV or SIV (103
) capable of yielding high infection rates in untreated controls [10
]. We used a more physiological virus dose that fell within the upper range of viral load observed in human semen during acute HIV-1 infection [20
]. The repeated nature of the model has also the advantage of evaluating protection over multiple transmission events. The infection of control macaques after a median of two exposures suggests that treated animals that remain uninfected after 14 challenges were protected over a median of seven transmission events. This model maintains high stringency, increases statistical power, provides improved estimations of risk reductions, and reduces the number of macaques [36
Although a repeat-challenge model is more relevant to human transmission, which typically requires multiple exposures, a disadvantage of the model is that it is logistically demanding over a long period of time. This limits the ability to do multiple concurrent arms, which raises a potential for bias. However, animal studies with non-concurrent arms can be well controlled. In our study, we have staggered interventions because of logistic feasibility and to prevent unnecessary use of animals. All animal procedures were done under identical conditions by the same personnel and experimental protocol, thus minimizing the potential for bias. Likewise, it is not known if repeated exposures to the virus can ultimately alter susceptibility to infection. The similar infection rates observed among previously or newly exposed animals suggest that the impact of repeated exposures on susceptibility to infection is minimal [12
Several important observations were made from the longitudinal analysis of the breakthrough infections. The finding that wild-type SHIV initiated all six infections suggests that PrEP failure in these animals is due to residual virus replication in cells not protected by drugs, rather than a rapid selection of a drug-resistant virus. Of the four animals infected during FTC treatment, only one selected resistant viruses, while an FTC-resistant virus emerged in one of two animals that failed FTC/TDF PrEP. The absence of tenofovir resistance in both macaques is consistent with clinical observations showing resistance to FTC and not tenofovir as the most frequent pathway of resistance to Truvada [38
]. The two macaques in which FTC-resistant mutants emerged had the highest peak viremias, suggesting that selection of drug resistance may be facilitated by higher virus replication. Thus, lower acute viremias may have diminished the risk of resistance during extended treatment with FTC or FTC/TDF. Similar blunted viremias during early infection have been noted in macaques failing PrEP with an orally administered CCR5 inhibitor [39
]. These data underscore the potential differences in virus load and drug resistance dynamics during PrEP failures from those in mono- or dual drug therapy of established infections [21
The attenuated acute viremias may have additional clinical and public health implications. It is well established that massive depletion of CD4+
T cells, specifically CD4+
memory T cells, begins in the acute stage of SIV as well as HIV-1 infection in the gut and other lymphoid tissues, and is generally proportional to the degree of virus replication [42
]. Substantial reductions in acute viremia may conceivably reduce CD4+
T cell depletion, help preserve immune function, and attenuate the course of HIV infection. In humans, reduction in virus set points by 1 log10
have been estimated to double the time to progression to HIV disease [45
]. Reductions in virus loads in the animals that failed PrEP were apparent at the first time points before seroconversion and were sustained under continued drug treatment after all the animals seroconverted. Blunted viremia during acute infection in persons who fail PrEP will likely depend on the period of drug exposure and the potency of the PrEP regimen. PrEP-treated populations will likely be monitored by serologic testing for infection to minimize drug exposure and reduce the risks of drug resistance. The period of drug exposure will thus depend on the frequency of serologic testing but will inevitably be at least several weeks long, enough to affect early CD4+
T cell depletion. Individuals with acute infections who have very high virus loads may also play a key role in the epidemic spread of HIV-1 because they are more infectious than individuals with chronic infections who have lower virus loads [46
]. Therefore, reductions in acute viremia during PrEP treatment may contribute to decreases in HIV-1 transmissibility at the population level and could add to the overall effectiveness of PrEP.
The data from this study demonstrate the potential high effectiveness of daily or intermittent PrEP against sexual HIV transmission, support expanding PrEP trials in humans and identify promising PrEP modalities.