Treatment During Primary HIV Infection as a Method of Altering the Epidemic Spread of HIV?
One goal of diagnosing patients with primary HIV infection (PHI) is to treat them with highly active antiretroviral therapy (HAART) potentially to alter their disease course. The hypothesis is that such treatment will preserve immune function, limit the development of viral diversity, and perhaps alter the viral "set point." The extent to which this intervention accomplishes these goals is still to be determined by clinical and scientific study. A second proposed and perhaps less well-known goal of treating PHI is to alter the epidemic spread of HIV infection by treating patients during a time when, because of biological factors and behavioral patterns, they may be highly infectious and likely to spread the virus to others.
While the goal of altering the disease course of patients has received a great deal of public attention, the goal of curtailing further infection and epidemic spread has received less attention. Yet, this broader population-based goal remains the focus of some researchers who see treatment during PHI as "a unique public health opportunity to abort rapid epidemic spread [of HIV] in sexual networks."1 The question of whether routine treatment during PHI could interrupt or limit the epidemic spread of HIV remains unanswered and it seems that as researchers learn more about the transmission and pathogenesis of HIV, the answer to that question becomes more complex and in some ways more elusive.
An early epidemiological model proposed per-contact infectivity to be 100 to 1,000 times higher during PHI than during chronic infection.2 A more recent model predicts a 20-fold increase in the probability of transmission between the peak viral shedding associated with PHI and its set point.3 These models assume high infectiousness of patients as a result of high viral load and peak genital viral shedding. They also assume a continuation of high-risk sexual behavior in patients. If these models hold true, it is reasonable to hypothesize that treating patients during PHI with the goal of decreasing infectiousness might be a worthwhile endeavor impacting the overall spread of HIV.
Identifying PHI in Patients
PHI is a stage of disease occurring after infection but before the development of detectable HIV-specific antibodies. Most patients develop influenza- or mononucleosis-like symptoms 2 to 6 weeks after infection. Symptoms tend to be nonspecific; thus, identifying patients solely on the basis of symptoms is difficult. A clinician therefore should be ready to question a patient about risk behaviors as well. Patients may be reluctant to disclose certain behaviors and clinicians may be forced to make clinical judgments based on symptoms and epidemiological considerations.
The most sensitive ELISA or Western blot assay, which are relatively inexpensive and readily available, cannot determine antibodies to HIV during PHI. Assays to detect p24 antigen are time sensitive and may not detect the antigen in many patients who have been symptomatic for 3 weeks or longer.3 (See "Backgrounder: Recognizing and Diagnosing Primary HIV Infection" in this issue.)
The only certain way to diagnose persons with PHI is to measure plasma HIV RNA or proviral DNA. Given the expense associated with these tests, the overall cost-effectiveness and utility of these assays in routine testing has been called into question. Patients experiencing PHI are a time-limited minority. By far, the vast majority of HIV infections are thought to occur as a consequence of the risk behaviors of chronically infected individuals. Still, epidemiologists believe that there is a substantial indirect benefit to the early diagnosis and treatment of the PHI population, such that the ripple effect of treating PHI could result in a substantial impact on the epidemic spread of HIV.3
Factors Governing Sexual Transmission
To understand better the risk of sexual transmission in the context of PHI, it is necessary to consider the general factors that govern such transmission. The mode of exposure, the infectiousness of the source patient, and the susceptibility of the infected person appear to be the 3 controlling factors that affect the risk of transmission.
The mode of exposure determines in large part the risk of becoming infected. The behaviors listed below have been estimated to carry the corresponding levels of risk:
Immunologic and genetic factors of the person exposed can influence susceptibility to infection with HIV. Deletions in CCR5 coreceptors, as well as SDF variants and HLA types, have been shown to protect against infection. Also, the presence of systemic and mucosal antibodies may decrease susceptibility such that there is a decrease in the per-sex-act risk in steady partnerships.
Factors such as the local inflammatory response caused by bacterial vaginosis, urethritis, and sexually transmitted diseases (STDs) have been associated with upregulation of CCR5 coreceptor expression and increased activation of T lymphocytes in the mucosa of women, thus leading to increased susceptibility. Some studies point to circumcision as having a protective effect. However, the data have been challenged by other findings.5
There are conflicting data about whether viral subtype or tropism (macrophage or CD4 T cell) may influence the infectiousness of the transmitter. At least one study in macaques suggests that replication capacity in vivo predicts intravaginal infection.6 Above all other factors, plasma viral load appears to be most predictive of infectiousness. This is supported by studies of vertical transmission, as well as by the large STD-intervention study in Rakai, Uganda, in which plasma HIV RNA viral load of less than 1,500 copies/mL was associated with no transmissions. Also, the rate of transmission increased progressively with increasing viral load to a maximum of approximately 23% in persons with viral loads exceeding 50,000 copies/mL.7 It should be noted that this study was done among a cohort of untreated serodiscordant heterosexual couples.
A correlation has been found between HIV RNA levels in blood and genital secretions. The concentration of HIV in semen appears to predict infectiousness. Female genital secretions may also predict infectiousness, however the data for this assertion are less certain because of assay variability and the biological fluxes caused by the menstrual cycle.5
There appear to be some exceptions to the rule of viral load in blood predicting viral load in genital secretions. Hypersecretors comprise a minority population of HIV-infected individuals in whom there is a discordancy between plasma viral load and genital viral load. Some studies have identified hypersecretors as men who consistently show higher seminal viral loads than plasma viral loads. It appears, however, that hypersecretors can be men or women. At least one study reported detectable HIV RNA in the genital tract of women despite the suppression of plasma HIV RNA to below 500 copies/mL in the plasma as a result of HAART.8
Clearly the presence of a symptomatic STD increases infectiousness by increasing the level of HIV RNA in genital secretions. Ping et al. reported a 10-fold increased level of HIV RNA in the semen of patients with gonorrhea as compared to control patients and an increase in viral shedding in the semen of patients with genital ulcerative disease.9
HIV Transmission in the Context of PHI
Before undertaking any large-scale intervention to identify and treat persons experiencing PHI, it is necessary to establish that such an endeavor will have a reasonable chance of affecting the spread of HIV. The first question to be answered then is whether transmission occurs readily during PHI and, perhaps, whether it occurs more readily than at other stages of infection as has been proposed by some models.
Pilcher et al. recently documented 5 cases of sexual HIV transmission from patients experiencing PHI to their partners, who were then subsequently confirmed to be experiencing PHI themselves.1 For purposes of this study, patients were classified as experiencing PHI if they had positive p24 antigen results, positive HIV RNA or DNA, and a negative ELISA or 2 or fewer bands on Western blot within 30 days. Two of the cases involved single sexual encounters occurring prior to the transmitter's onset of symptoms. Another case involved multiple exposures, but all occurring before day 2 of the transmitter's onset of symptoms, and one case involved a single exposure on day 7 of the transmitter's onset of symptoms. (See Figure 1.)
The issues brought forth by these cases are: when is the peak period of infectiousness in a person experiencing PHI? Does it precede the onset of PHI symptoms? And, assuming that a person with full-blown symptoms of PHI will have less interest in sex, will an intervention targeting PHI patients likely be too late to significantly affect infectiousness and transmission?
The answers to these questions are not entirely clear but more data have recently emerged that may provide some hints. A study published this year in AIDS examines the HIV RNA in various body compartments of 17 individuals with PHI.10 The time from onset of symptoms to sampling ranged from 8 to 70 days. HIV RNA in plasma was significantly higher in the group with PHI when compared to a group of historical control patients with chronic infection. HIV RNA levels in seminal plasma and cervicolavage fluids, however, were not significantly different between PHI and control patients. (It should be noted that these results differ from observations by Vernazza et al. in which they reported a significantly higher seminal viral load in PHI patients than in chronically infected patients.11) The authors of the AIDS article conclude that their results do not preclude the possible existence of a period of "uniformly high HIV shedding" very early in PHI. Still, this study leaves undetermined the exact time frame of such a period of high viral shedding.
If infectiousness is primarily a function of the level of viral shedding, this AIDS study fails to support the rationale of treating patients during PHI to minimize transmission. Those patients presumably would be no more infectious than chronically infected patients. This is a small study and its results must be confirmed by other studies before any definitive judgment can be reached on the relative levels of viral shedding in patients with PHI or chronic infection. In addition, other factors might affect the infectiousness of a patient during PHI.
Viral Fitness and Diversity
One question that has not been explored is whether the genetic diversity of the transmitter's virus has any effect on its infectiousness. Early transmission studies done mostly in men showed that a selection of variants occurs in the process of transmission. That is to say, while the transmitter may have a heterogenous population of variants, only a few variants are transmitted to the recipient. Further, the resulting homogenous viral population in the person being infected may persist for months.12
As mentioned earlier, studies in rhesus macaques demonstrate that viral replicative capacity predicts infectiousness for mucosal transmission.6 Taken together, these studies suggest that someone experiencing PHI may have a largely homogenous swarm of highly "fit" virus, which not only has increased replication capacity, but also increased infectiousness.
A study reported recently in Nature Medicine13 seems to have thrown another twist into this potential scenario. In that study researchers compared the diversity of envelope sequences in the provirus found in the peripheral blood mononuclear cells of 5 men who had recently seroconverted (within 35 to 148 days of infection) to the viral diversity found in women who had also recently seroconverted. The men were found to have a homogenous proviral population with less than 1% diversity, whereas 5 of the 6 women were found to have 5% diversity in their viral population. To confirm this difference, the researchers tested the viral diversity of an additional 26 women with recent infection (17 to 168 days) and found a heterogenous virus population in 15 of the 26 women tested. It should be noted that this study was done in a population of Kenyan patients and that genetic diversity was being tested in blood rather than in genital secretions.
The diversity of populations of provirus may predict the diversity of genital secretion populations since provirus represents the earliest archived virus. If diversity is determined by gender, one question that should be addressed is whether men may be more infectious than women during PHI. The homogeneity of the viral swarm in men could account for the difference in viral load at seroconversion between men and women. An editorial accompanying the Nature Medicine article speculates that "higher viral levels in men at seroconversion, combined with lower diversity, may indicate relatively uncontrolled replication, either because of gender differences or more frequent antigen priming among the women studied."14
Treatment with HAART during PHI has been proposed as a method to decrease infectiousness and the spread of HIV. Studies have shown that such treatment reduces viral replication not just in the blood but also in the genital tract. The timing of such treatment may be critical in determining outcome. If peak infectiousness occurs uniformly in patients before the onset of PHI symptoms, it may be that identifying PHI patients after the onset of symptoms and treating them may not have the desired effect of reducing infectiousness during the most critical period of viral spread. Even if peak infectiousness extends some time after the onset of PHI symptoms, treatment would have to begin very early to achieve the desired result.
It may be more important to identify and trace the sexual contacts of PHI patients and treat those contacts before the onset of their symptoms in order to arrest the spread of HIV. Identifying and tracing these contacts may be laborious, but it may ultimately reduce transmission rates.
The choice of antiviral treatment may also have some bearing on the transmission rate of PHI patients and their sexual contacts. It might be important to choose antiviral regimens that not only reduce plasma viral load, but also penetrate genital tract compartments. Protein binding, drug transporter affinity, and drug ionization have been theorized to influence antiviral distribution, at least into the male genital tract. These and other factors affecting antiviral distribution into the genital tract are reviewed in an article by Vernazza et al.5
Certainly, treating STDs and other genital tract infections in PHI patients and their sexual contacts will further reduce their infectiousness. Also, identifying hypersecretors may be useful in this context.
PHI patients and their sexual contacts must be counseled to avoid unsafe sexual behaviors. As the parameters of infectiousness become more defined, however, it may become necessary to re-evaluate the risk associated with certain sexual behaviors in the context of PHI.
One glaring example of a behavior in need of re-examination in the context of PHI is oral sex. The perception about the risk associated with oral sex, at least in the gay community, is that it is a relatively safe behavior especially in light of undetectable viral loads and the general success of HAART. Study after study has shown a low risk of transmission despite the many anecdotal accounts of persons attesting to have acquired HIV through oral sex. It is possible that PHI may represent a time frame during which oral sex may be riskier than at other times.
In fact, one study suggests that oral sex in general may have a higher transmission rate than what has been previously reported. In a study reported at the 7th Conference on Retroviruses and Opportunistic Infections, Dillon et al.15 reported an incidence of 7.8% of infection associated with oral sex in a cohort of 102 men who had sex with men. Stringent criteria (including detuned antibody assays to determine the recent infection of the source patients as well as partner corroboration) were used to determine oral sex as the only possible route of infection.
While this study does not address the disease stage of the source patient, it is reasonable to conclude that if the rate of transmission through oral sex and semen is greater than once thought, that rate might be greater still when the patient is experiencing PHI -- a time when genital secretions are likely to have a higher concentration of inoculum.
While it might be useful to test everyone who shows up at a doctor's office or emergency room with symptoms consistent with PHI, the expense associated with viral load tests makes this type of intervention impractical. Given that the symptoms associated with PHI can be nonspecific, a more targeted approach may be necessary to identify those patients who have been infected with HIV.
One possible approach is to identify the venues and times of likely transmission. A study recently published in the Journal of Acquired Immune Deficiency Syndromes compared the sexual risk behavior of gay and bisexual men who attended circuit parties both in the San Francisco area where they lived and other cities.16 In this study of 295 men living in San Francisco, unprotected anal sex (insertive or receptive) with partners of opposite or unknown serostatus was most prevalent during circuit parties outside the geographic area and was predicted by HIV-positive status and the use of crystal methamphetamines, sildenafil (Viagra), and amyl nitrates ("poppers"). These types of events may predict HIV transmissions and PHI. Public education efforts might be timed to coincide with these types of events and questionnaires used to identify potential PHI patients might ask not only about specific sexual behaviors but also about attendance at this type of event.
Given the spread of HIV to populations other than the gay community, other possible venues and events associated with risk-taking must be identified and taken into account.
While it is clear that transmission of HIV occurs mostly during chronic infection, PHI presents an opportunity to affect transmission dynamics that are unique to this early period of infection. Given that prevention efforts have only been partially successful and the rate of new infections in the United States continues at a constant rate, perhaps it is time to view the problem from the perspective of reducing infectiousness as well as reducing susceptibility. Identifying and treating persons with PHI may present the perfect opportunity to rethink our whole prevention effort while learning more about the natural history of this disease.
This article was provided by The Center for AIDS. It is a part of the publication Research Initiative/Treatment Action!. Visit CFA's website to find out more about their activities and publications.