HIV reinfection cases reported at CROI

Simon Collins, HIV i-Base

A poster discussion session included studies looking at different aspects of dual infection and reinfection and included two case studies that showed where this had a clinical outcome. Two other studies, one from London and one from San Francisco, reported on aspects of dual infections.

Erika Castro from University of Vaudois Hospital, Lausanne, and colleagues presented a useful case of reinfection between two men (M1 and M2) who had been sexual partners since 2006. [1]

M1 was initially diagnosed in 2000 during primary infection and had been suppressed on HAART through to 2007 with no history of drug resistance. M2 had been on HAART for five years with detectable viral load (range 3–4 log) and documented triple class resistance. In February 2008, viral load in M1 rebounded to 280 copies/mL and continued increasing. Resistance and phylogenetic tests were compared from 2000 and 2008 (86 sequences:  whole-genome (n=28), env (n=28), and gag (n=25).

All sequences were sub-type B. The genotypic analysis from 2008 showed 25 new related drug resistance mutations in M1 (11 in RT and 14 in protease), of which 23 were also present in M2. Additionally, M1-2008 sequences clustered within the M2-2008 branches and distinct from M1-2000 sequence clusters in all trees. Recombination between the original M1 and M2 strains was not observed, with M1 being replaced with M2 sequences following superinfection.

The case is important for highlighting several aspects of HIV reinfection:

  • That reinfection can occur in established infection.
  • That it can occur between regular partners (that no immune protection develops to repeated exposure to one virus).
  • That reinfection can occur after several years of exposure (as with initial infection, chance and probability are low).
  • That there is a clinical risk from reinfection when partners have different resistance profiles (most reinfection cases are only discovered because of unexplained viraemia in stable patients).
  • That ART did not offer protection against reinfection, probably because any PEP effect would be negligible if the new MDR virus was resistant to the ARVs in that combination.

Martine Braibant from University de Tours Hospital and colleagues presented another case of sub-type B being reinfected with subsequent sub-type B infection. [2]

This patient entered a long-term non-progressor (LTNP) cohort in 1995 aged 58, following a ten year history of HIV infection, with a CD4 count >600 cells/mm3 for the previous 5 years and viral load of 135 copies/mL on study entry. From 1995–1999, viral load slowly increased to around 10,000 copies/mL and CD4 count dropped steadily to <500 cells/mm3. The initial infection was found to have a 20 nucleotide deletion in nef (consistent in 28 sequences) and the loss of viral control and immunologic progression from 1995 was associated with detection of subsequent sequential reinfection with two fully competent phylogentically different strains. Both new strains were also sub-type B.

The patient responded well to HAART, achieving viral suppression and CD4 recovery >700 cells/mm3 within the first year of therapy, but potentially would have maintained the option to remain off-treatment for many years if reinfection hadn’t occurred.

This case highlighted that:

  • Progression rates may be determined by both virologic and immunologic factors.
  • Reinfection in the absence of resistance may have clinical implications on disease progression, requiring earlier treatment.
  • That long-term exposure to low level sub-type B virus did not promote an immune response that was protective of subsequent sub-type B infections.

The authors highlighted in their conclusion that this last point showed the inherent difficulties for development of a preventative vaccine.

Jane Deayton and colleagues from St Barts Hospital, London report three cases of inter-clade dual infection detected during routine genotype testing. The three cases were Caucasian UK-born MSM whose only risk factor was sexual exposure in the UK. Case 1 had been diagnosed in 2001 and received a genotype test in 2008 prior to starting therapy that indicated dual infection with sub-types B and G. Cases 2 and 3 were tested after their HIV diagnoses in 2007 and 2008 and showed dual infections with B and CRF02_AG, and B and A, respectively. Both these patients were reported to be stable off treatment.

None of the cases included significant drug resistance mutations.

The authors concluded that these were the reports of cross-clade dual infection from sexual transmission in MSM in the UK are rare (only one other was known). Additionally this indicated onward transmission of clades associated with African epidemics and an increasing cross-over of viruses between different demographic groups.

Finally, Larry Bragg and colleagues in San Francisco presented a poster looking at cases where majority viruses change in an individual though competitive expression following dual infection. Whilst difficult to distinguish from superinfection (ie reinfection after an initial infection) sequential expression of dual infections (SEDI) theoretically could come from dual initial infection or reinfection shortly after initial infection, especially prior to seroconversion. [4]

The group included 220 recently infected persons with at least two genotypes (560 person-years of follow-up). The mean age was 37 and median time from infection to first test was 107 days. Divergent viruses appeared in 7 cases, an overall incidence density of 1.24/100 person-years.

Their model estimated a risk of SEDI that was 16-fold higher in the first year post-infection compared to after one year. The estimated rate of SEDI was 4.1/100 person-years (95% CI, 1.8 to 9.2) in the first year following infection and was 0.2 per 100 person-years beyond 1 year post-infection (95% CI, 0.03 to 1.8).

This study highlighted that in the case of dual infection or early reinfection the predominant infection is determined within the first year.


1. Castro E et al. HIV-1 superinfection with a drug-resistant strain in a patient successfully controlled with ART. Poster abstract 480.

2. Braibant M et al. Disease progression after intrasubtype superinfection in an HLA-B57+ asymptomatic LTNP initially infected with a nef-defective HIV-1 strain. Poster abstract 302.

3. Deayton J et al. Inter-clade dual HIV-1 infection: an emerging phenomenon. Poster abstract 447.

4 Bragg L et al. HIV-1 Superinfection Surveillance in an Acute Infection Cohort Using pol Sequences from Resistance Genotyping: 1996 to 2008. Poster abstract 446.

Links to other websites are current at date of posting but not maintained.