Bridging the Gap: Conference Record [Abstract book, International Conference on AIDS (12th: 1998: Geneva, Switzerland)]

12th World AIDS Conference

Abstracts 42254-42271 827

Methods: A population pharmacokinetic analysis of nelfinavir was performed
on data from a phase III clinical trial which consisted of 509 plasma concentrations
from 174 HIV-infected patients. The analysis was performed using NONMEM.
Results: The data were best described by a one-compartment model with
first-order absorption. Estimates of the mean population parameters (and 95%
CI) of nelfinavir were as follows: total body clearance (CL/F) was 40.7 L/hr
(37.7, 43.7), volume of distribution (Vd/F) was 733 L (531,935) and the first order
absorption rate constant, ka was 1.21 hr-1 (0.69, 1, 73). Neither body weight
(range: 42-140 kg), age (range: 21-63 years), gender (11% female), race (78%
Caucasian, 12% African American and 10% other) nor dose level (500 mg or
750 mg three times a day) appeared to be significant covariates for clearance.
Patients who had either a low baseline CD4 count (-<100 cells per /i1) or who
received concomitant therapy with a macrolide antibiotic experienced a modest
though statistically significant reduction in clearance (i.e. higher AUCs) of 18%
and 23% respectively. Addition of either of these covariates to the model resulted
in a reduction in interindividual variability in clearance. Body weight, age, gender,
dose level, disease status at baseline and concomitant medications were not
significant covariates for volume of distribution.
Conclusion: Body weight, age, gender, race and dose level did not appear to
significantly affect clearance of nelfinavir. A statistically significant but clinically
unimportant decrease in clearance was observed in patients receiving concomitant therapy with a macrolide antibiotic or in patients who had a CD4 count <-100
cells per pI at baseline.
42268 A fluorescence quenching assay for determining the
binding affinity (ka) of HIV protease inhibitors to ca-1 acid
glycoprotein
Jila Bakker, D. Tazartes, C. Flexner. Johns Hopkins University Baltimore MD
Osler 524, 600 N. Wolfe St. Baltimore, MD 21287, USA
Objective:To develop a quick and inexpensive assay for determining the ka of
binding to alpha-1 acid glycoprotein (AAG)
Design: in vitro assay
Method: The endogenous tryptophan fluorescence of AAG (20 /iM) was
quenched upon titration with increasing concentrations of drugs. The relationship
between the amount of quenching and the fraction of drug bound to AAG was
determined using the following analysis method of Weber and Young:
X = fraction of drug bound to AAG = (FO-fb)/(fa*FO) where Fo is the endogenous fluorescence intensity of AAG, fb is the fluorescence intensity of AAG at a
given drug concentration, and fa is the fraction of maximum quenching achieved
at saturating drug concentrations. Scatchard analysis provided the ka's for the
protease inhibitors.
Results: We determined the binding affinities of five protease inhibitors for
AAG.
' ka= 1 X 101 M-1
7kG- ritonavir.... ka = 1 X 104 M-1
ka= 2 X 105 M-1
saquinavir
ka= 8 X105M-1

Results (see table): RIT significantly increased the AUCO-24 of EFV by 21%
(geometric mean ratio, GMR). EFV significantly increased the AUC12-24 and
AUCO-24 of RIT by 18% (GMR) and 17% (GMR), respectively.
Conclusion: RIT inhibited EFV metabolism, most likely via the P450 3A4 pathway. In addition, EFV inhibited RIT to almost the same degree. Based on these
data, a RIT dose of 500 mg bid could be used when given in combination with
EFV. If there is intolerance attributed to ritonavir, a ritonavir dosage reduction of
100 mg may be considered. EFV does not require dose adjustment when given
with RIT.
S42270 Brain delivery of anti-HIV drugs: Rapid screening using
an in vitro model of the blood-cerebrospinal fluid (CSF)
barrier
Nathalie Strazielle', G. Gosselin2, J.F. Ghersi-Egea2. INSERM U325 Institut
Pasteur de Lille, 1 Rue du PR Calnette BP245 59019 Lille; 2CNRS UMR 5625
Universite Montpellier II, Montpellier, France
Background and Objectives: (HIV-1) is a neurotropic retrovirus and can enter
the brain shortly after infection, where it may form a protected pool of viral particles. Therefore a key issue in achieving effective treatment of AIDS is the cerebral
penetration of drugs. The choroid plexuses which form the blood-CSF barrier, are
important in controlling drug efficiency as they constitute a main route of entry for
delivery of antiretroviral agents to CSF, subarachnoid spaces and perivascular
spaces in which infected macrophages are primarily found. In addition they are
the main site of drug metabolism in the brain. The purpose of our work is to
establish an in vitro model of the blood-CSF barrier and use it for the rapid
evaluation of the cerebral delivery of antiretroviral agents.
Methods: Choroidal epithelial cells isolated from newborn rat plexus are grown
on permeable filters in a bicameral chamber system adapted for transport studies in both blood-CSF and CSF-blood directions. These transport studies are
performed using radioactive compounds or analyzed by HPLC.
Results: Choroidal epithelial cells form a confluent monolayer with a characteristic cobblestone morphology. They express epithelial markers (cytokeratin)
and specific choroidal markers (transthyretin). Tight junctions are demonstrated
by the cortical distribution of actin filaments and occludin and cell polarity is evidenced by electron microscopy (apical microvilli and cilia). The barrier properties
are supported by the formation of a hydrodynamic barrier, by a low permeability
to sucrose and an elevated transepithelial electric resistance. Drug-metabolizing
enzymes remain active in the cultured cells and specific transporters (thymidine,
L-phenylalanine) or polarized efflux mechanisms are also present. Experimental
conditions have been defined that allow precise and rapid transport measurement
in both directions, for compounds of various lipid solubility. AZT has been found to
slowly but significantly cross the in vitro blood-CSF barrier as it has been shown
in vivo. Influx and efflux rates of other anti-HIV nucleosidic derivatives as well as
newly developed molecules are now under investigation.
Conclusion: We have developed an in vitro model of the blood-CSF barrier
that closely mimics the in vivo phenotype of the choroidal epithelium, and is
suitable for transport studies. This model will allow to classify the currently used
anti-HIV drugs and other potent antiretroviral molecules for their extent of CSF
penetration and should generate useful information for the design of new antiviral
drugs.
142271 Pharmacokinetics of abacavir (1592U89) in cerebrospinal
fluid (CSF) using multiple lumbar punctures of few
subjects and "sparse sampling" techniques
Richard W. Price1, F. Aweeka2, S.E. Bellibas3, S. Staprans4,
T. Novakovic-Agopian5, W. Mahoney6, W. Symonds7. 'Neurology Rm. 4 M 62,
San Francisco Gen Hosp., San Francisco, CA 94110; 2San Francisco Gen.
Hospital/UCSF San Francisco, CA; 3Glaxo-Wellcome, Research Triangle, NC,
USA
Objectives: To define the penetration of abacavir into CSF based on plasma and
CSF pharmacokinetics (PK).
Design: Subjects on abacavir (600 mg every 12 hour) underwent 5-6 lumbar
punctures (LPs) at varying times after dosing. Each measurement representing
a distinct segment of the area under the concentration time curve (AUC), 0-12
hours. One LP was performed in the context of an intensive blood PK study, the
others accompanied by blood before and after the LP.
Methods: Abacavir was measured by validated HPLC method.
Results: This strategy allowed comparison of abacavir AUC in both plasma and
CSF. Figure shows one subject's data. CSF curve (dashed line) is constructed
from samples on indicated days; the 0- and 12-hour points are extrapolated from

SC-52151
ka = 2 X 106 M-1

5M drug

Conclusions: HIV protease inhibitors quench the fluorescence of AAG in a
manner which relates to their individual binding affinities. We used this property
to devise a quick and inexpensive assay of AAG binding which has potential use
for estimating free drug concentrations in plasma.
S42269 1Pharmacokinetics of efavirenz (EFV) and ritonavir (RIT)
after multiple oral doses in healthy volunteers
William Fiske', I.H. Benedek2, J.L. Joseph2, S. Dennis3, R. O'Dea3, A. Hsu3,
D.M. Kornhauser2. 1 Drug Metabolism and Pharmacokinetics, 2DuPont Merck
Pharmaceutical Company, Wilmington, Newark, DE; 3Abbott Laboratories,
Abbott Park, IL, USA
Objective: The objective of the study was to evaluate the potential for a pharmacokinetic interaction between EFV and RIT when administered concomitantly.
Methods: Subjects were assigned to four groups. Using an open-label, randomized, cross-over design each group received the single agent alone (EFV
600 mg qd on Days 1-10 or RIT 300 mg q12h on Day 1, 400 mg q12h on Day
2 and 500 mg q12h on Days 3-10) and in combination using the same regimen
for 10 days with 28 days between periods. Plasma samples were obtained over
a 24 hour period beginning on the evening of Day 10 for each period. EFV and
RIT were measured using validated HPLC methods.

10000 -E
0 1000 -100~

Treatment (Groups) N

RIT (1 & 3)
RIT + EFV (1 & 3)
RIT + EFV (2 & 4)
EFV (2 & 4)

RIT AUCO-12
(/lg.h/mL)
103 [62]
126 [94]
136 [57]

RIT AUC12-24
(/Ig-h/mL)
111 [51]
135 [73]
144 [43]

RIT AUCo-24
(/,gh/mL)
214 [110]
261 [165]
280 [94]

EFV AUCO-24
(/t gh/mL)
265 [74]
291 [115]
239 [92]

Day6
I s.=, Day9
Day 41
Day2 Day 30

Significant difference (p - 0.05) between combination and single agent; mean [standard deviation]

10--
0 2 4 6 8 10 12
Hours After Dosing