Anti-HIV immune responses and partial viral suppression

Paul Blanchard, HIV i-Base

Steve Deeks from the University of California presented some interesting data on anti-HIV immune responses from patients who remain on antiretroviral therapy despite detectable viral loads.

It has been observed that such patients, who often have highly drug resistant HIV, can frequently maintain viral load set-points below pre-therapy levels. This suggests that drug resistant virus is continuing to be suppressed either through continued activity of drugs, reduced replicative capacity or the emergence and maintenance of HIV specific immune responses.

To investigate the last possibility the group measured the fraction of CD4 and CD8 T cells that responded to HIV antigens in patients with drug-resistant viraemia. These measurements were then compared to responses seen in both untreated patients and patients on antiretroviral therapy with viral loads less than 50 copies/mL. Twenty four patients with virologic failure of HAART were studied. This group had a median period of prior treatment with protease inhibitors (PI) of 58.5 months, 45.8 months of which had been with intermittent or continuous failure. Median CD4 count was 306 cells/mm3 and median plasma HIV RNA 3.3 log copies/mL. Despite the length of time experiencing virologic failure the median CD4 was still 132 cells/mm3 above the nadir and HIV RNA was 1.7 log below baseline. All 24 patients had genotypic and/or phenotypic evidence of resistance to PI’s with a median 69-fold decreased susceptibility to the currently administered PI.

Median (IQR) Positive (>0.1%)
Untreated 0.07% (0.00 – 0.18) 3/11 (27%)
Treated, VL >50 (failure) 0.27% (0.17 – 0.44) 21/25 (84%)
Treated, VL <50 (success) 0.07%
(0.00 – 0.19)
For median values:
p<0.05 failure vs untreated
p<0.001 failure vs success
For number with positive response:
p<0.01 failure vs treated
p<0.01 failure vs success

Analysing the data another way the group also looked at whether the presence of anti-HIV CD4 or CD8 T cell responses correlated with level of viraemia. A positive correlation was found (rho=0.43, p=0.03) between viral load and % CD8 T Cells producing interferon in response to p55 antigen in treated patients with VL >50 (failure) only. There was no observed correlation between CD4 response and viral load for any of the groups. It was also shown that neither the CD4 or the CD8 response correlated with absolute CD4 T cell count for any group (treated or untreated).

Deeks concluded that gag-specific CD4 T cells were observed in many treated patients with drug-resistant viraemia. The fraction of such cells was greater in patients with drug resistant viraemia than untreated patients and patients on treatment with undetectable levels of viraemia. Additionally gag-specific CD8 cells were observed in most patients with detectable viraemia (both treated and untreated) with a positive correlation being seen between viraemia and gag-specific CD8 cells in treated but viraemic subjects.

It was hypothesised that these observations may fit with the theory that virologic rebound with a less fit virus may allow the emergence and maintenance of HIV specific CD4 T cell responses. Subsequent immunological control of resistant virus – in some patients – may then contribute to durable suppression of a drug-resistant virus population below pre-therapy set-points.


These data from Deeks are intriguing but do not allow to conclude that the immune response to HIV is truly virostatic and not only an inactive response of the immune system.

Care is also needed in comparisons with untreated patients in that the reasons why these patients remain untreated may mark them out as a group with particular characteristics of their own. Such patients may not be similar in all aspects apart from receiving no treatment as is intended.


Deeks SG, Sinclair E, Harris J et al. Partial viral suppression with combination therapy is associated with enhanced HIV-specific CD4 and CD8 T cell responses. Antiviral Therapy 2001; 6 (Supplement 1):21. (Abstract 26)

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