HIV Escape in Cerebrospinal Fluid Despite Viral Control in Plasma
It is usually admitted that detectable CSF HIV viral load above 50 c/ml is a rare event during effective HAART. A recent study (1) shows that this situation could be more frequent. Here we discuss the implications of these results, but also raise some questions and comments about the design limitations of this study.
This cross-sectional analysis used archived CSF and plasma specimens from 2 centers (56 patients from Europe, 13 from the US). They were obtained from asymptomatic subjects treated with antiretroviral therapy for >6 months (with no change in the regimen for >3 months). These patients were treated with the 'commonly used first ART regimens', including TDF, ABC, 3TC, FTC, ZDV, ATZ, LPV, EFV.
All cases had sustained viremia <50 c/ml but in 7 cases (10%), HIV RNA level was >50 c/ml in the CSF. The median CSF HIV RNA level was 121 c/ml (range: 52-860 c/ml). CSF pleocytosis was not significantly different in patients with detectable virus. The conclusions of this study are:
- Autonomous viral replication in the brain is possible during effective ART;
- This is a 'frequent' event;
- It is related to the duration of treatment: the 7 subjects with detectable CSF HIV RNA had significantly longer treatment duration: median = 77 months, than patients with undetectable CSF virus: 33 months (p=.002).
- It is related to the number of previous blips and the history of 1 or more treatment interruptions (71% vs 15% of patients)
- It does not correlate with the CPE (Coefficient Penetration Effectiveness) score.
The importance of the brain as a sanctuary site in HIV diseases has been established for many years and is difficult to study as direct sampling is not possible. Studies on CSF are important but such a small cross-sectional study raises several questions:
- Is that really new? Letendre and colleagues have shown for at least 3 years that HIV RNA remains detectable in the CSF of 42% of patients below the 50 c/ml cutoff (2009 CROI) which is the one used in this study; in the editorial from David Clifford (2), it is stated that 10% is 'a surprisingly high proportion of patients with detectable CSF virus', but in the initial presentation of their study at CROI (Click here) the authors mentioned that this is an 'uncommon finding'...
- Were these patients really asymptomatic in terms of neurocognitive impairment? It is known that sub-clinical defects are frequent, but no formal neurocognitive testing was performed in this series;
- What was the role of resistance? This issue is discussed by authors but they state that 'CSF viral load....was too low to allow resistance analysis'. Although this is generally true -as levels are <1,000 c/ml- it is now possible to sequence low levels viremia (in particular in the patient with CSF levels at 860 c/ml)...
- What is the strength of the conclusion on the CPE rank? The number of patients with detectable CSF HIV levels was quite low, with diverse combinations. On the contrary, the CPE rank was established in a series of several hundreds of patients...
- Is the number of blips a real measure of adherence? Currently, with sensitive techniques like the Cobas Taqman, 'blips' frequently observed even in patients with strong adherence... Would the conclusions still have been valid if the authors had performed HIV RNA quantification at the 2 copies/ml level, both in plasma and CSF?
Even if several questions remain open, slow disease evolution in the brain is of major concern. Studies like this one tend to prove that the brain is a time bomb during HAART. These results are in agreement with previous reports, but the precise role of the different antiretroviral drugs is not clear in this paper. More studies are urgently needed, including prospective analysis in large cohorts of patients, as they have the potential to influence prescription choices...
1-HIV-1 Viral Escape in Cerebrospinal Fluid of Subjects on Suppressive Antiretroviral Treatment. Edén A, Fuchs D, Hagberg L, et al. J Infect Dis. 2010 Nov 4. [Epub ahead of print]
2-Viral Escape in Cerebrospinal Fluid-An Achilles Heel of HIV Therapy? Clifford DB. J Infect Dis. 2010 Nov 4. [Epub ahead of print]
You can comment this article below.
Key words: CPE rank, CSF, HIV, brain, replication, reservoir