Failure to detect nelfinavir in the cerebrospinal fluid of HIV-1-infected patients with and without AIDS dementia complex

Objective: To assess the penetration of the HIV-1 protease inhibitor, nelfinavir, into cerebrospinal fluid (CSF). Design: Nelfinavir, a commonly used HIV-1 protease inhibitor (PI), is highly effective for reducing plasma viral load. It is deployed clinically in combination with other antiretroviral agents, including nucleoside and nonnucleoside reverse transcriptase inhibitors (NRTIs and NNRTIs). Despite its potency based on plasma HIV-1 RNA results, its effectiveness in reducing HIV-1 RNA levels (i.e., viral load) in the central nervous system (CNS) is less certain. We sampled the CSF as a surrogate for brain because this fluid also is separated from the blood by a barrier to free diffusion, the blood-CSF barrier (BCB), which shares properties with the blood-brain barrier (BBB). These studies of nelfinavir CSF pharmacokinetics exploited the multiple CSF samples derived from individual study subjects who were enrolled in studies the primary objective of which was to compare viral kinetics in CSF and blood in response to antiviral therapy. Methods: Six study subjects, four with and two without AIDS dementia complex, underwent multiple lumbar punctures (LP). Intervals of CSF sampling after drug dosing were varied (from 0.48 hours to 10.3 hours after nelfinavir administration) to quantitate nelfinavir concentrations throughout the steady-state dosing interval. In four study subjects, CSF sampling was accompanied by assessment of nelfinavir levels in plasma before and after LP, whereas in the other two subjects, a single plasma sample was obtained before or after the LP. In total, 25 CSF samples were analyzed. Nelfinavir concentrations in CSF and plasma were determined using an high-performance liquid chromatography (HPLC) method with a limit of quantitation of 25 and 50 ng/ml, respectively. Results: plasma concentrations before and after LP averaged 2420 +/- 1365 ng/ml and 2528 +/- 1132 ng/ml, respectively. Nelfinavir was not detected in any of the CSF samples and levels >25 ng/ml were not present in the CSF. Thus, standard therapy with nelfinavir does not result in CSF drug concentrations at or exceeding the IC95 level for most HIV-I isolates. However, study subjects with high CSF viral loads experienced a marked reduction in the context of the combination-drug regimen including nelfinavir with two subjects showing a comparable CSF response with that in plasma. Conclusions: Nelfinavir does not appreciably penetrate into the CSF. The clinical importance of this observation is not certain, in that in four study subjects who initiated nelfinavir in combination with other antiretroviral therapy, a comparable degree of viral suppression was obtained in both the CSF and the blood when sampled 4 weeks or later after initiating therapy.