A number of individuals infected with more than one strain of HIV have been identified over the past few years. Should people already diagnosed with HIV be concerned? Given the limited number of cases seen so far, the risk of multiple infections (also called dual infection) appears to be quite low. And there are many more pressing health concerns facing people living with HIV/AIDS. Nevertheless, two trends are worth noting: dual infection seems more likely to happen under certain conditions, and it is associated with faster progression to AIDS-related events. This article describes current evidence and theories behind this emerging phenomenon.

Gathering Evidence

Experts once hoped that a single HIV infection would prevent further infections, much like a vaccination. In the mid-1990s, however, studies using analogous viruses in primates showed that sequential infections were possible. Some people believed it was only a matter of time before something similar would be seen in humans.

Compelling evidence of dual HIV infection in humans appeared in 2002. A report in the Journal of Virology in August of that year strongly suggested reinfection in two injection drug users (IDUs) from Thailand (one female, one male). The woman was initially diagnosed with HIV subtype AE only, followed by detection of subtype B approximately two months later. The man was apparently reinfected with subtype AE virus approximately six to ten months after his primary diagnosis with subtype B virus. Neither individual was being treated for HIV during the study period.

In the September 5, 2002 issue of the New England Journal of Medicine, researchers from the University of Geneva reported on a man initially diagnosed with subtype AE virus in November 1998 whose viral load became undetectable (below 50 copies/mL) with antiretroviral therapy. He stopped treatment in January 2001 and shortly thereafter traveled to Brazil, where he had multiple unprotected sexual contacts. In April 2001, three weeks after his return from South America, his viral load spiked to 400,000 copies/mL and he reported symptoms of acute retroviral syndrome, which can signal a new HIV infection. Lab tests subsequently detected a second strain of HIV -- subtype B, which is common in Brazil. The researchers concluded that reinfection had occurred.

More recently, Davey Smith, MD, of the University of California at San Diego and colleagues reported in the August 12, 2005 issue of AIDS that a man with wild-type (drug-sensitive) subtype B virus was apparently reinfected about a year after his first infection with a different subtype B virus resistant to protease inhibitors, which he had never taken, and 3TC (lamivudine, Epivir), which he started only after the second infection. Another case of dual infection with two subtype B viruses with discordant drug sensitivity was reported by the same research group in 2003. In that case, however, the subject was first diagnosed with drug-resistant subtype B virus and then found to have wild-type HIV of the same subtype four months later. Like the man in the 2005 report, this individual had not taken antiretroviral therapy before the apparent reinfection event.

Other cases of multiple HIV infection have been identified in the past four years, although the total number remains small -- only 16 apparent reinfections by one measure (a 2005 Medscape survey of the scientific literature done by a group from the Gladstone Institute of Virology and Immunology in San Francisco). The Gladstone researchers, however, did not consider cases of coinfection. In addition, dual infection rates may be higher than reported, since few people with HIV have been tested for multiple strains. Only larger future studies using more sophisticated technologies and better tracking of source partners can provide a clearer picture of the incidence (rate of new cases) and prevalence (total number of existing cases) of coinfection and reinfection in a given population.

Impact on Disease Progression

Dual infection in humans has been linked to disturbances in immune control and poorer prognosis. In the case of the man who traveled to Brazil, the emergence of his subtype B virus while off therapy coincided with a loss of 300 CD4 cells/mm³ and a dramatic rebound in viral load before he resumed highly active antiretroviral therapy (HAART) four months later.

In a report from 2004, Smith and colleagues analyzed the two dual infection cases mentioned above plus a third man with apparent secondary infection (wild-type followed by drug-resistant virus). Among the three men, CD4 cell counts dropped an average of 132 cells/mm³ within six months of acquiring the second strain, while viral load levels increased an average of 1.6 logs -- a 40-fold increase.

Geoffrey Gottlieb, MD, of the University of Washington in Seattle and colleagues retrospectively located five individuals with dual infection (four U.S. gay men, one female sex worker from South Africa). Four were coinfected near the time of seroconversion, while the other was reinfected 1.3 years after initial infection. All five had rapid disease progression: from seroconversion to below 200 CD4 cells/mm³ within 3.1 years on average, and to an AIDS diagnosis or death within 3.4 years. Time from seroconversion to AIDS typically takes 8-10 years in untreated individuals.

In a letter to The Lancet in June 2005, Gottlieb proposed that the case of unusually rapid HIV disease progression in a New York City man described by local health officials in February 2005 might also be due to dual infection rather than the emergence of a so-called "supervirus" (see "News Briefs" in this issue).

Several factors might explain an association between dual infection and a surge in HIV disease progression. For now, these are hypothetical and could be related to viral dynamics and the way the second virus attacks the immune system or evades immune responses.

Acquiring a drug-resistant viral strain, for instance, would increase the likelihood of losing a response to antiretroviral therapy. This was seen in Smith's 2005 report as well as others. For those not on treatment, overwhelming a drug-resistant virus (considered less able to replicate) with a new wild-type virus (considered more virulent) could result in a higher viral load and speed progression of disease.

Viral recombination might play a significant role in accelerating HIV disease (see table below). Recombination increases viral diversity more rapidly than mutations that evolve slowly through replication errors. Recombinant viruses may be less sensitive to anti-HIV drugs and are potentially more virulent than nonrecombinant viruses. This might result from altered tropism -- specifically, the virus' ability to use the CXCR4 coreceptor to enter cells, as was the case in the New York man; CXCR4-using viruses are associated with worse disease outcomes than viruses that use the CCR5 coreceptor.

Genetically mixed viruses might also be more adept at evading immune responses in a type of evolutionary strategy. At the 3rd International AIDS Society (IAS) conference this past July, Carolyn Williamson, PhD, from the University of Cape Town and colleagues reported finding recombinant virus in six of six dually infected subjects, along with evidence of viral evasion of cellular immune defenses and neutralizing antibodies. The South African team proposed that dual infection "enables recombination to contribute significantly to viral adaptation to immune responses ... and may help explain rapid disease progression."

Alternatively, the link between dual infection and disease progression might be a product of individual characteristics. Gottlieb has speculated about whether certain people who are inherently predisposed to faster disease progression may also be more susceptible to reinfection. His team noted in their 2004 report, for example, that the one subject believed to be reinfected "had rapid CD4 decline immediately after initial infection, suggesting a host susceptibility to infection with a second virus."

Susceptibility and Protection

As to when reinfection might occur, data collected so far show an interesting trend. Researchers at the Gladstone Institute pointed out in their survey of the literature that multiple infections have not been reported in anyone beyond three years after his or her first infection. (Only a female sex worker from Kenya with recombinant AC virus might have been reinfected after three years, but the exact date is unknown due to a nine-year gap in blood sampling.) This observation has been borne out in recent studies in which dual infection was not observed in chronically infected individuals, even among IDUs who consistently shared needles and HIV positive individuals who had partners with different strains and high risk of re-exposure.

While this trend may be an inaccurate observation based on coincidence or testing errors, it has also been seen in primates. A study done in the late 1990s by Ron Otten, PhD, and colleagues from the Centers for Disease Control and Prevention (CDC) showed that macaque monkeys could be infected with two strains of HIV-2 up to four weeks after a first infection, but not between eight and 72 weeks afterwards. Humans might have a similar window of susceptibility to reinfection of approximately three years.

The lack of evidence for dual infection during chronic (long-term) HIV disease suggests a protective mechanism at work, such as immune responses that evolve over time or "viral interference" -- the ability of the original virus to ward off acquisition of another. Any protective role played by anti-HIV therapy in chronic infection would appear to be negligible, since multiple infections have not been reported in untreated chronically infected people after three years.

Dual infection therefore seems to occur only during acute or early infection -- and in these cases, anti-HIV therapy might well make a difference. Evidence suggests that multiple infections happen only in people with acute or early infection who are not being treated or only intermittently treated with anti-HIV drugs. This implies that antiretroviral therapy has a protective effect, at least during early HIV disease, either in blocking secondary infections or in preventing certain coinfecting strains from asserting themselves. Antiretroviral agents used as pre-exposure prophylaxis (PREP), taken before a high-risk incident, might work in a similar way to block a first infection. However, using anti-HIV agents as PREP remains experimental and unproven.

Although continuous antiretroviral therapy (during early disease) and chronic infection (regardless of treatment) each appear to provide protection against dual infection, more research is needed to understand and confirm these observations. Studies are likewise needed to identify any individual characteristics that might make some people more prone to acquiring a second virus. These factors are currently unknown, although cases such as the one described by Gottlieb in 2004 point to the possibility.

Managing Dual Infection

The appearance of genetically distinct viruses within an individual complicates the management of HIV disease. Because multiple infections often lead to signs of accelerated disease progression, the typical clinical response has been to begin or resume anti-HIV treatment. Some people among the recently documented cases have controlled their secondary infection with standard antiretroviral therapy. Others, even if responding well to a first regimen before reinfection, have required salvage or rescue regimens containing four or more drugs. Resistance tests may help guide clinicians in their choice of therapy. Newer drugs and drug classes might likewise improve the chances of treatment success, especially in cases of secondary infection with a drug-resistant virus.

The current understanding of dual infection raises complex questions for people with HIV, clinicians, and prevention workers alike. What approach, if any, should be taken given the small number of cases? Should people with early HIV disease be counseled to start treatment to avoid reinfection, even if their virus is under control? What impact will reinfection have on HIV positive individuals who "serosort," or choose to have sex only with other positive people? Will those with chronic infection feel freer to have unprotected sex despite the risk of acquiring other sexually transmitted infections (STIs)?

As always, the best guide to risk management is reliable information. The Gladstone researchers wisely counsel that "clinicians and researchers should provide balanced and broad views of the risks of unprotected sex between HIV-1 infected persons, and avoid exaggerated or sensational claims about superinfection that could undermine behaviors such as serosorting and serodisclosure that can help to curtail the spread of HIV." Beliefs about multiple infections can affect behavior. In interviews with 193 HIV positive men who have sex with men (33% Latino, 29% African American), researchers from the San Francisco Department of Public Health reported in 2003 that the 83% who believed reinfection was damaging to health were significantly less likely to report unprotected anal sex with an HIV positive partner or any partner compared with those who did not share this belief.

Vaccine Design

Recent dual infection news has been sobering for vaccine researchers, who study the mechanisms the immune system uses to control pathogens (disease-causing organisms) and work to develop agents that will elicit the same immune responses. The specific protective mechanisms, or "correlates of protection," necessary to subdue HIV are unknown, which has been a major obstacle in HIV vaccine research since the beginning of the epidemic. The task is now made more difficult by the knowledge that the immune system cannot reliably prevent reinfection even when responding vigorously to an initial infection.

At the 2003 Retrovirus conference, for example, Todd Allen, PhD, of Massachusetts General Hospital and colleagues reported that a robust and specific immune response to one HIV subtype (B) did not inhibit reinfection with another. The fact that virus-specific immune responses are unable to stop other invading viruses -- even those of the same subtype, as seen in the cases reported by Smith's group -- suggests that priming the immune system with a vaccine to control one viral subtype will not be sufficient, and that designing a vaccine broadly protective against a range of HIV strains might be impossible.

Still, with the added challenge comes a silver lining: the apparent protection afforded by chronic infection, antiretroviral drugs, or individual characteristics. Figuring out how these or other factors allow the immune system to prevent dual infection could be a significant breakthrough and may help guide researchers toward their elusive goal. Given the moribund state of HIV vaccine development, no time should be wasted in exploring this possibility.

Conclusion

What little is known about dual infection has been sketched from a handful of case reports. Uncertainty will prevail until scientists resolve the issue of whether reinfection occurs independently of coinfection. If all dual infections are in fact coinfections acquired at or near the same time, there would theoretically be no risk of later being reinfected with a second strain of HIV. Studies would then focus on why, when, and in whom coinfection takes place, as well as why some coinfecting HIV strains emerge virulently and only during early infection.

If, however, reinfection is a distinct phenomenon, researchers will need to determine precisely what conditions are necessary for multiple infections to occur, who might be more susceptible to them, and what are the clinical implications. Only a fuller understanding of dual infection can help people with HIV make informed decisions about risk. (For information about the Positive Partners study, which looks at whether reinfection occurs between sexual partners, see "Open Clinical Trials.") And, with luck, investigating the dynamics of multiple infections might lead to the ultimate protection: an HIV vaccine.

Nicholas Cheonis is the former editor of BETA.

Selected Sources

1. Allen, T.M. and others. HIV-1 superinfection despite broad CD8+ T-cell responses containing replication of the primary virus. 10th Conference on Retroviruses and Opportunistic Infections (CROI). Boston. February 10-14, 2003. Abstract 307.

2. Blackard, J.T. and others. Human immunodeficiency virus superinfection and recombination: current state of knowledge and potential clinical consequences. Clinical Infectious Diseases 34(8): 1108-1114. April 15, 2002.

3. Gottlieb, G.S. and others. Dual HIV-1 infection associated with rapid disease progression. The Lancet 363(9409): 619-622. February 21, 2004.

4. Jost, S. and others. A patient with HIV-1 superinfection. New England Journal of Medicine 347(10): 731-736. September 5, 2002.

5. Marcus, J.L. and others. HIV superinfection vs dual initial infection: what clinicians and patients should know. Medscape 11(1). 2005. Accessed November 4, 2005.

6. Otten, R.A. and others. Identification of a window period for susceptibility to dual infection with two distinct human immunodeficiency virus type 2 isolates in a Macaca nemestrina (pig-tailed macaque) model. Journal of Infectious Diseases 180(3): 673-684. September 1999.

7. Plantier, J.C. and others. HIV-1 group M superinfection in a HIV-1 group O-infected patient. 12th CROI. Abstract 288.

8. Ramos, A. Intersubtype human immunodeficiency virus type 1 superinfection following seroconversion to primary infection in two injection drug users. Journal of Virology 76(15): 7444-7452. August 2002.

9. Smith, D.M. and others. HIV drug resistance acquired through superinfection. AIDS 19(12): 1251-1256. August 12, 2005.

10. Weiser, B. and others. Multidrug resistant HIV-1 resulting from intrapatient viral recombination. 3rd International AIDS Society (IAS) Conference on HIV Pathogenesis and Treatment. Rio de Janeiro. July 24-27, 2005. Abstract MoFo0306.

11. Williamson, C. and others. Lower rates of adaptive evolution in HIV-1 dual infections compared to single infections. 3rd IAS Conference. Abstract MoFo0305.