November 19, 1999
Ideally, antiretroviral therapy will control viral replication in all body compartments where virus had the ability to replicate. Two body compartments have been identified, cerebrospinal fluid and seminal fluid, in which the pharmacokinetics of drugs differ from that of blood. The blood-brain barrier is semipermeable. Protein bound drugs do not cross efficiently into the cerebrospinal fluid (CSF). In addition, specialized molecules pump certain drugs out of the CSF, effectively reducing drug concentrations of certain compounds that are able to cross into the CSF. The pharmacokinetics of drugs in CSF has not been well studied. At this meeting David Haas and colleagues presented several abstracts describing the kinetics of viral replication and the pharmacokinetics of several antiretroviral agents. They employed a novel technique, the placement of an indwelling lumbar catheter, which allowed them to sample the CSF at multiple time points over a 12-hour period. The pharmacokinetics of stavudine (d4T) and indinavir (IDV) in the CSF are described below. Dr. Kurtizkes will be reporting on the pharmacokinetics of emivirine.
Four treatment naïve patients had multiple sampling of CSF and blood four days after beginning the combination of d4T, 3TC, and nelfinavir. Mean peak concentrations of d4T of 3,051 nM were achieved in plasma one hour after taking the medication. Mean peak concentrations of 321 nM were achieved in CSF 3.5 hours after taking the medication. The half-life of d4T was longer in CSF compared to plasma, 3.27 hours compared to 1.54 hours. Because of the longer half-life in CSF, d4T concentrations in CSF exceeded that in blood after four to six hours of pill taking. Because d4T is taken twice a day, concentrations of d4T in CSF exceed that in blood for most of the day.
Eight patients receiving indinavir plus two nucleoside reverse transcriptase inhibitors volunteered to have a lumbar catheter placed for multiple CSF sampling. Protein binding of indinavir was calculated to be 58.3% in plasma and 5.7% in CSF. Free concentrations of indinavir peaked at 5,452 nM in plasma and 366 nM in CSF. Similar to the story with d4T, peak concentrations were delayed in CSF compared to plasma, and the half-life of drug in CSF was longer than that in plasma. At the end of the eight hour dosing interval, the concentration of indinavir exceeded that in plasma, 130 nM compared to 93 nM. In 4 of 8 patients, the concentration of indinavir exceeded the IC95 of virus (100 nM) throughout the eight-hour dosing interval, and the concentrations exceeded 85% of the IC95 in 7 of 8 patients.
These studies clarify several confusing issues with respect to interpreting studies that describe drug concentrations in CSF. Many investigators report a CSF to plasma ratio. Because drug entry into CSF may be delayed, and because the half-life of drug in CSF may differ from that in blood, the ratio of drug in CSF to plasma will vary substantially depending upon the timing of the specimen collection. The only variable that matters is the absolute free (non-protein bound) concentration of drug relative to the concentration needed to inhibit the virus' replication. These studies demonstrate that both d4T and indinavir are present at sufficient concentrations to inhibit virus replication in the central nervous system.