12:00 49.
|
Dynamics of HIV Rebound in Cerebrospinal Fluid (CSF) after Interruption of Antiretroviral Therapy (ART)
R. J. Ellis*1, S. Letendre1, S. D. W. Frost2, A. J. Leigh Brown1,2, M. E. Childers1, J. A. McCutchan1, and HIV Neurobehavioral Res. Ctr.
1Univ. of California, San Diego and 2Univ. of Edinburgh, UK
|
Background: Previously presented findings suggest that HIV in CSF represents a mixed viral population derived from at least 2 sources: extraneural trafficking, and local replication in neural tissues. Among patients with CD4 counts >200, pleocytosis (elevated CSF leukocytes) typically accompanies CSF HIV RNA levels >5000 copies/mL. To explain this link between pleocytosis and high CSF viral loads, 2 hypotheses have been proposed: trafficking leukocytes introduce virus to the CSF; or high viral replication in the CNS triggers an influx of leukocytes as part of a local immune response. We sought to distinguish between these hypotheses by comparing CSF and plasma viral dynamics after structured treatment interruption (STI).
Methods: Intensive serial sampling of CSF and blood was performed in 12 subjects with established HIV infection who underwent STI. The relationship of plasma and CSF viral setpoints to the level of pleocytosis was analyzed using a general linear model. The dynamics of plasma and CSF virus, and the level of pleocytosis over time, were analyzed using nonparametric regression.
Results: In 2 patients, HIV RNA reached detectable levels in plasma prior to CSF. (Sampling was too infrequent to detect this effect in the remaining patients.) In 4 patients, a brisk pleocytosis (WBC > 10/muL) developed during CSF viral rebound. Univariate analysis showed that pleocytosis was associated with a higher CSF viral setpoint (4.4 vs 2.9 log10 copies/ml, p<0.05) but not with a higher plasma setpoint (4.7 vs 5.0 log10 copies/mL, p>0.5). Multivariate analysis showed trends towards higher CSF viral setpoints with pleocytosis (p=0.06) and with higher plasma viral loads (p=0.07). In 1 patient in whom CSF virus rebounded initially above the CSF set point, pleocytosis clearly lagged behind CSF viral dynamics. Following 6-8 weeks off ART, patients established a fixed plasma-CSF gradient at steady state.
Conclusions: During STI, CSF HIV dynamics differ both quantitatively and qualitatively from those in blood. The finding that CSF pleocytosis lags behind changes in CSF HIV RNA suggests that pleocytosis reflects a dynamic response to local increases in viral replication, rather than a major source of CSF virions. These results are consistent with local bursts of viral replication and immunologic responses in neural tissues during STI, and add to existing knowledge of within-host compartmental interactions in HIV infection.
|
