icon star paper   Articles  
Back grey_arrow_rt.gif
Super-Infection with HIV
  "Clade B HIV-1 superinfection with wild-type virus after primary infection with drug-resistant clade B virus"
AIDS 2003; 17(7):F11-F16. Kersten K. Koelsch a; Davey M. Smith a; Susan J. Little a,b; Caroline C. Ignacio a; Theresa R. Macaranas a; Andrew J. Leigh Brown c; Christos J. Petropoulos d; Douglas D. Richman a,b; Joseph K. Wong a,b. From the aUniversity of California San Diego, La Jolla, CA, USA; bSan Diego Veterans Affairs Healthcare System, La Jolla, CA, USA; cUniversity of Edinburgh, Edinburgh, Scotland, UK; and dViroLogic, South San Francisco, CA, USA.
We present one case of HIV-1 superinfection by two HIV variants from the same viral clade in the absence of antiviral treatment. These data indicate that 4 months after infection by drug-resistant HIV this patient was infected by a second, drug-sensitive virus. This observation could alternatively be explained as co-infection with two divergent viruses, followed by late outgrowth of the variant, which was initially a minor population.
Discriminating co-infection from superinfection relies on the exclusion of a mixed viral population at the earliest timepoint post-infection. Because the detection threshold for both the dye primer sequencing and sequencing of 20 clones per timepoint is only 5-10% of the minor variant, we applied a more sensitive technique to detect population mixtures. Length polymorphism detection via Genescan technology is a highly sensitive tool for the detection of minority populations. The sensitivity for the detection of length polymorphisms goes down to one single nucleotide deletion or insertion within a PCR product. Using this technology, we confirmed that the viral population in plasma at the three earliest timepoints were homogeneous. Therefore, with reasonable certainty, we excluded the possibility that this patient had been co-infected. A different viral population was dominant at timepoint 4 and thereafter. Genescan analysis in one experiment demonstrated the co-circulation of both viral populations at the earliest timepoint after the presumptive superinfection event. As expected from the results of the clonal sequencing, the shorter initial viral variant was present as a very small minor population based on relative peak heights. This provides further evidence that Genescan analysis permits the detection of minority populations of less than 5%. The presence of the wild-type population as a minority species in the first 3 months is also less likely than superinfection, because the wild-type variant appears to be much more fit, with steady-state viral loads almost 100-fold greater than those of the initial drug-resistant variant. We did not find evidence for recombination in our case. Differences between the initial and superinfecting virus were consistent with all available analyses on pol, V3 and V4-V5. However, follow-up has been limited to 12 months.
One important issue is how superinfection impacts disease progression. Data from animal models indicate that there is a finite initial period of greater susceptibility to superinfection. Animals that seem to be protected from superinfection after primary infection had higher survival rates than animals that became superinfected.
Infection by viral variants with differing replication capacities and their variable susceptibility to the host immune response might be expected to have a significant impact on disease progression. Indeed, in the case described here, an abrupt increase in plasma viremia occurred coincident with the appearance of the second variant and consistent with the hypothesis that this second variant had greater in-vivo fitness than the initial, drug-resistant virus. The theoretical acceleration of disease progression that might result from a higher post-superinfection viral setpoint appears to be reflected in the steeper trajectory of the CD4 cell decline in this patient after superinfection.
A second issue of importance is the impact of superinfection on treatment response. An obvious scenario for concern is that of a patient with drug-sensitive virus responding well to therapy, who then becomes superinfected with drug-resistant virus. The transmission of drug-resistant virus is a common event. However, the case described here highlights a more insidious danger arising as a consequence of superinfecion. In this case, standard drug susceptibility testing at late timepoints would fail to detect the occult drug-resistant virus. Nevertheless, if this patient were to initiate therapy, it seems likely that drug-resistant virus would quickly re-emerge.
Our report coincides with three recent publications and one other report of apparent superinfection. Two of the published accounts reported superinfection by virus from a different clade, whereas in this and another recently published report, superinfection appeared to be occurring with related virus from the same clade. In the case described by Altfeld et al, the second infection occurred during a period of structured treatment interruption, despite strong immune control of the initial virus strain. This observation and the demonstration of intraclade superinfection indicates that the anti-HIV immune responses are narrowly directed, with worrying implications for vaccine design. Our case differs from most of the others and is similar to only one other case, in that it demonstrates intraclade superinfection that occurred in the absence of the modulation of natural disease by therapy or vaccination. Studies in larger cohorts as well as a more detailed analysis of humoral and cellular immune responses are needed to determine the true incidence of superinfection events and the underlying mechanisms. Together, these recent reports suggest that superinfection may occur more commonly than has previously been assumed, which has broad implications for HIV treatment, epidemiology, vaccine development and pathogenesis.
The immunological response to HIV-1 infection has been postulated to impede superinfection with a second virus; however, a few recent reports have documented cases of HIV-1 superinfection in humans either from different viral clades or from the same clade.
The immunological response established during primary HIV-1 infection results in partial control of the infecting virus, and has been proposed to limit subsequent re-infection. However, genetic mosaics resulting from interclade recombination events provide evidence for superinfection events in vivo. The occurrence of such events has been recognized in sub-Saharan Africa, Asia and South America, where divergent subtypes co-circulate. Recombination is more readily discerned after co-infection with virus from different clades. The isolation of such recombinants does not, however, always identify the individual who was infected and does not discriminate co-incident double infection from superinfection.
Two recent publications directly demonstrated superinfection by viruses from a clade different from that of the initial infecting strain. Under these circumstances, clade-specific immune responses appear to have been insufficient to prevent infection by the divergent second strain. Here we report a case of superinfection with virus of the same subtype as the initial infecting viral strain, occurring in an untreated patient during the first year of HIV infection. In another recently published report, same-clade superinfection occured in a patient during structured treatment interruption. The occurrence of intra-subtype superinfection extends concerns about superinfection to those populations infected by a single predominant viral subtype, and has additional implications for vaccine development and clinical management.
This patient was one in a cohort of patients with transmitted drug resistance discovered during primary infection. Upon enrollment, he had a CD4 cell count of 711 cells/[mu]l and low plasma viremia, with 2413 copies/ml approximately 4 weeks after infection, and had received no antiretroviral treatment. Serial plasma and peripheral blood mononuclear cell (PBMC) samples were studied by dye primer sequencing of a 500 base pair (bp) fragment of pol in an attempt to characterize the reversion of drug resistance. At month 4, a second, apparently distinct, viral population was noted (data not shown). To confirm the presence of two diverse viral strains we performed sequencing of a 1302 bp fragment of the gag-pol region from nine plasma samples taken over the course of 12 months after primary infection. This showed a second viral strain that was, in contrast to the first strain, wild type at codons 181 and 184 of HIV reverse transcriptase, and showed additional polymorphisms (K20R, M36I, L63P) in protease that were not present in the first strain.
Phylogenetic analysis showed clustering of the three sequences at baseline, months 1 and 3 after infection, which segregated from the cluster of six sequences derived from plasma samples after month 4. Genetic distances (DNADIST version 3.5c, J. Felsenstein, University of Washington, Washington, USA) between pol sequences from early (patient group 1) and late timepoints (patient group 2) averaged 6.5%, and were similar in magnitude to the genetic distances between epidemiologically unlinked patients with primary HIV infection in San Diego (mean interpatient distance 5.6%). Both early and late viral variants belonged to clade B subtype, consistent with the epidemiological history from this patient.
  icon paper stack View Older Articles   Back to Top   www.natap.org