icon star paper   Articles  
Back grey_arrow_rt.gif
A Patient with HIV-1 Superinfection: safe sex, vaccine implications
  "...superinfection may precipitate more rapid progression of disease. Infected and noninfected persons should therefore exercise the same degree of vigilance to prevent HIV-1 exposure. With sexual activity seemingly increasing among persons with HIV-1 infection, this is a public health message that needs to be broadcast loud and clear..." editorial below
This study and others strongly suggest it is possible to get reinfected with a 2nd HIV virus after having HIV. This case and others appear to show the second infection resulting from sex, but I suspect that reinfection can also occur from IVDU exposure. Interestingly, this patient also acquired HCV after being HCV negative for years. HCV appeared to be acquired around the time of the 2nd HIV infection or recently prior to the HIV infection. This study report does not address how or when HCV was contracted but the insufficient information available leads me to speculate that perhaps this patient may have contracted HCV and HIV sexually from the same person.
This study raises questions about developing a vaccine for HIV. Where does this leave HIV-1 vaccine initiatives, if it can be demonstrated that anti-HIV-1 immunity cannot provide protection against infection with different clades of the virus? The NEJM editorial (Bruce Walker, Phillip Goulder) below discusses this question, but the authors speculate that superinfection is an infrequent occurrence and therefore would not greatly imperil vaccination initiatives.
New England Journal of Medicine, Sept 5, 2002, Vol 347: 731-36, number 10
Authors- Stephanie Jost, B.Sc., Marie-Charlotte Bernard, M.D., Laurent Kaiser, M.D., Sabine Yerly, M.S., Bernard Hirschel, M.D., Assia Samri, Ph.D., Brigitte Autran, M.D., Li-Ean Goh, Pharm.D., and Luc Perrin, M.D.
The detection of an increasing number of circulating recombinant strains of human immunodeficiency virus type 1 (HIV-1) indicates that genetic recombination can occur in cells infected with two strains of HIV-1. Coinfection with two circulating strains of HIV-1 has been detected in a few subjects in communities where HIV-1 infection is endemic. Coinfection may result from exposure to a second virus either shortly after the initial infection or during the course of established HIV-1 infection; the latter circumstance is called superinfection.
Most viral infections induce lifelong immunity, but reinfection with respiratory viruses such as respiratory syncytial virus is common, most likely because immunity becomes nonprotective or fades. It is thought that HIV-1 superinfection is a rare event and that it is prevented by previous viral exposure through a phenomenon called superinfection immunity.
However, HIV-1 superinfection has been induced experimentally in chimpanzees. In this animal model and in superinfection induced with the simian immunodeficiency virus in macaques, the second infection produces a slower deterioration in immunity than does the initial infection, and there is more efficient control of viremia. In this article, we report a case of HIV-1 superinfection.
In November 1998, a 38-year-old man presented with an acute retroviral syndrome. Anti-HIV-1 antibodies were undetectable, the level of p24 antigen was greater than 100 pg per millimeter, the plasma level of HIV-1 RNA was 805,000 copies per milliliter, and the CD4 cell count was 684 per cubic millimeter. Sequences of the HIV-1 reverse transcriptase and protease genes revealed no mutations associated with drug resistance and identified the HIV-1 as subtype AE. For years, the patient had had sexual contacts with multiple unknown male partners. He enrolled in the QUEST trial in November 1998 and received highly active antiretroviral treatment (HAART) with zidovudine, lamivudine, abacavir, and amprenavir for 27 months. From month 21 to month 27 he participated in a vaccination trial and was randomly assigned to receive ALVAC vector vCP1452.11
Six weeks after the initiation of HAART, the plasma level of HIV-1 RNA declined to 1000 copies per milliliter. Treatment was then interrupted for six weeks because of toxic effects on the liver. After HAART was resumed, the HIV-1 RNA level decreased rapidly, to less than 50 copies per milliliter. After vaccination, HAART was again interrupted (on January 21, 2001) as part of the vaccine research protocol. In February 2001, the patient's plasma HIV-1 RNA level rose to 80,000 copies per milliliter (the first rebound) and then decreased to 21,000 copies per milliliter. A rapid increase in the HIV-1 RNA level was next observed on April 10 (the second rebound), and for the next four months the level fluctuated between 200,000 and 400,000 copies per milliliter. The patient's symptoms (transient fatigue and fever) were mild, and he declined to resume HAART during this period. His history revealed that he had had several unprotected sexual contacts in Brazil three weeks before the second rebound of viremia. Four months after the second rebound, however, HAART was resumed, and the plasma level of HIV-1 RNA rapidly decreased. Treatment was again interrupted after an increase in the alanine aminotransferase level to 800 U per liter, as compared with a level of 200 U per liter before HAART. At this time, serologic data and quantification of hepatitis C virus (HCV) RNA documented an acute HCV infection. The increase in alanine aminotransferase was attributed both to HCV infection and to drug-induced toxic effects. Treatment with pegylated interferon and ribavirin was followed by clearance of HCV RNA (to less than 500 copies per milliliter) within two months.
(edit comments by Jules Levin: The patient had been previously tested for HCV antibody & HCV RNA but was negative until testing positive now. This NEJM report does not discuss how or when this patient contracted HCV. Perhaps the contracted HCV through IVDU recently and prior to the sexual contact which is proposed to have superinfected the patient or perhaps the patient contracted HCV & HIV together sexually from the same person).
The patient we describe had HIV-1 seroconversion in November 1998 as a result of infection with HIV-1 subtype AE; the infection was successfully treated with HAART. After the interruption of HAART at the end of January 2001, subtype AE viremia rebounded (in February); the viremia then declined during the following two weeks before a second rebound, at which time the HIV-1 RNA level plateaued at 400,000 copies per milliliter. This second rebound involved HIV-1 subtype B, which rapidly replaced subtype AE. This subtype B virus segregated with isolates from Brazil, a finding consistent with the patient's recent travel and sexual exposures in that country.
This observation provides strong evidence of HIV-1 superinfection. To rule out coinfection, we attempted to amplify subtype B by means of subtype-specific PCR. We failed to detect subtype B in plasma RNA or in proviral DNA samples obtained up until the second viremia rebound. In addition, the subtype B isolate had a higher replicative capacity in vitro than the subtype AE isolate and, once detected, rapidly replaced subtype AE in vivo.
Strong, HIV-1-specific CD8-cell responses were directed against a single AE-derived epitope during the period of infection with subtype AE. None of the subtype B epitopes derived from the patient's sequences were recognized at any time during either subtype AE or subtype B infection. The drop in subtype AE CD8 cells at the time of the switch in subtypes could be attributed to the sequence variation in the B epitope that affected HLA-binding capacity. That the inhibition of subtype AE replication was due to an immune response to ALVAC epitopes after immunization cannot be ruled out, since the p17 peptide SLYNTVATL in the patient's AE was identical to the "subtype B" gag sequence inserted into ALVAC.11
The biologic course after HIV-1 subtype B superinfection was characterized by the persistence of high plasma levels of HIV-1 RNA and a loss of 300 CD4 cells per cubic millimeter within the four months after the emergence of subtype B. The subtype B superinfection led to rapid progression of disease,16 in contrast to the pattern of disease observed after experimental superinfection in monkeys.7,8,9 The early initiation of HAART in our patient may have limited his exposure to HIV-1 antigens and his HIV-1-specific immune responses. In North America and western Europe, subtype B predominates and diversifies rapidly within individual patients,17 rendering the detection of superinfection especially difficult; the frequency of superinfection might thus be underestimated.
Our data indicate that natural infection does not necessarily induce cross-clade protection. There should be close monitoring of circulating HIV-1 strains in the context of vaccine development. Our observations also support the use of safe-sex precautions even among HIV-1-infected persons.
NEJM Editorial
HIV-1 Superinfection - A Word of Caution

Philip J.R. Goulder, M.D., Ph.D.
Bruce D. Walker, M.D.
Massachusetts General Hospital
Boston, MA 02114
Acute human immunodeficiency virus type 1 (HIV-1) infection is typically associated with an influenza-like syndrome, with levels of virus in the bloodstream during the initial symptomatic period typically in excess of 10 million particles per milliliter. These acute symptoms usually resolve within one to two weeks, and there is a gradual reduction in viremia to a mean of about 30,000 particles per milliliter one year after infection. This dramatic drop in viremia is thought to be due to the induction of partially effective immune responses against the virus, although without treatment, viremia slowly rises over time. The conclusion that the initial containment of the virus represents a partially protective immune response is supported by studies in animal models of AIDS, in which initially high levels of virus persist when the CD8 subgroup of T lymphocytes is experimentally removed. Together, these observations suggest that effective immune responses against HIV-1 are generated, even if they are ultimately incapable of preventing the progression of disease.
Another way to address the effectiveness of immunity induced by HIV-1 infection is to examine whether infection with one HIV-1 strain can provide protection against superinfection with a second strain after partial containment of the first virus. More than a dozen different subtypes of HIV-1, called clades, have been identified around the world, and they can differ from one another by 30 percent in the env gene sequence and 15 percent in the gag gene sequence. Although numerous cases of simultaneous new infection with two distinct HIV-1 strains have been reported, there have been few, if any, clear cases of superinfection, in which a second virus infects the host well after a quasi-steady state with the first virus has been achieved. Although experimental models of AIDS in animals indicate that superinfection is possible, documentation of such a phenomenon in humans has been lacking. The assumption that superinfection does not occur in humans infected with HIV-1 has bolstered hopes that it might be possible to develop a broadly cross-protective AIDS vaccine and probably has also influenced decisions among consenting HIV-1-infected partners about whether to follow safe-sex recommendations.
In this issue of the Journal, Jost et al.6 challenge these assumptions and provide convincing evidence that HIV-1 superinfection can occur long after an initial infection is established. In the case they describe, primary infection with HIV-1 clade AE, which is prevalent in Southeast Asia, was documented in 1998. For the ensuing 28 months, the only virus detected in the patient was of the same AE clade. During this period, the patient was successfully treated with a four-drug regimen against HIV-1. Then, three months after he stopped treatment because of drug toxicity and three weeks after he had additional sexual contacts in Brazil, B clade HIV-1, which is endemic in Brazil, was detected when he had a sudden rebound in viremia. Although the B clade-infected person in Brazil who transmitted the virus could not be identified and thus unequivocal evidence of superinfection could not be documented, these data strongly support the conclusion that this patient was first infected with an AE clade virus and was superinfected more than two years later with a B clade virus that accelerated the progression of disease. Similar HIV-1 superinfection with two different clades has also recently been described in two injection-drug users in Thailand.
Where does this leave HIV-1 vaccine initiatives, if it can be demonstrated that anti-HIV-1 immunity cannot provide protection against infection with different clades of the virus? There are aspects of this case that need not leave us in despair. First, the HIV-1-specific immune responses that were detectable were narrowly directed and insubstantial, with cytotoxic T lymphocytes targeting only a single region of the virus identified. The typical gradual broadening of immune responses may have been blunted by early and effective treatment and may have further waned during the period of effective antiviral therapy. Second, cases of superinfection occurring as a consequence of exposure to a genetically different clade of virus do not define the likelihood of such events. Exposure to a homologous clade of virus is probably far more frequent than exposure to genetically different clades and therefore more relevant to vaccine programs. Thus, if a vaccine can induce, in fully immunocompetent persons, broadly based HIV-1-specific immune responses that are associated with successful control of viremia, the vaccine may be successful even in the face of a rare challenge by a very different clade of virus. Finally, most programs of HIV-1 vaccine development already assume that protection against multiple clades will be difficult to achieve. Indeed, the viral variation in HIV-1 infection far exceeds that observed in influenza, yet the influenzavirus diverges sufficiently over time to require a new vaccine annually. Studies of animal models have already shown that achieving protection against divergent AIDS-causing strains of virus is likely to be a formidable task5 and may require the development of polyvalent vaccines.
The less optimistic view of this case is that it may be the tip of the iceberg -that superinfection may be frequent and may explain the presence of recombinant viruses worldwide. The fact that the first descriptions of superinfection in humans involve heterologous clades of virus may simply reflect the fact that it is easier to demonstrate superinfection when two distinct clades are involved than when only one clade is involved. It is worth remembering that only a tiny fraction of CD4 T lymphocytes are infected at any one time, presumably leaving an ample number of uninfected cells available as targets for the superinfecting viral strain. Caution should be exercised in concluding that the immune response was necessarily weak in the patient described by Jost et al., since only previously defined epitopes of cytotoxic T lymphocytes were tested for recognition and since virus-specific helper T-lymphocyte activity and neutralizing antibody levels were not determined. It is also difficult to ignore the fact that this patient expressed two of the HLA class I molecules most consistently associated with effective control of HIV-1 -namely, HLA-B*57 and HLA-B*51. In the worst-case scenario, superinfection with a virus differing by only one amino acid in a critical epitope might be associated with lack of immune control and progression to AIDS.10
The public health implications of this and other cases of superinfection are clear. The data show that infection with HIV-1 does not necessarily provide protection against superinfection with another clade of HIV-1, even though they do not reveal how likely it is that such superinfection will occur. Almost by definition, a superinfecting strain of virus will be one that is not well contained by the immune response to the primary strain. Thus, superinfection may precipitate more rapid progression of disease. Infected and noninfected persons should therefore exercise the same degree of vigilance to prevent HIV-1 exposure. With sexual activity seemingly increasing among persons with HIV-1 infection,11 this is a public health message that needs to be broadcast loud and clear.
The current case report leaves a number of questions open, but these questions will clearly help focus future research efforts. For example, to what extent is the increasing number of recombinant viruses worldwide the result of superinfection? Does superinfection occur within clades? Can such cases be used to begin to dissect the components of protective immunity in HIV-1 infection? We urgently need detailed evaluations of the immune responses and of the sequences of circulating virus before and after loss of control of viremia in cases of superinfection. The magnitude of the task of creating an effective AIDS vaccine will then become more apparent.
  icon paper stack View Older Articles   Back to Top   www.natap.org