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

828 Abstracts 42254-42276

12th World AIDS Conference

the 11.5 hour determination (steady state). The 12-hour plasma PK (solid line)
curve is from day 30 and overlying individual segments (unlabeled) from plasma
samples accompanying the other CSFs. The AUC for the CSF was 4,347 and
for the plasma 12,030 hr-mcg/L (ratio of 36.1%). Data from 6-10 subjects under
study will allow detailed modeling of drug distributions in these two body-fluid
compartments.
Conclusion: Sparse sampling using multiple LPs on a small number of subjects
can be used to define the pharmacokinetics and effective CSF concentrations
of antiretroviral drugs, including abacavir. Initial estimates indicate that the CSF
exposure (AUC) of abacavir is about one-third that of plasma.
S42272 Therapeutic drug monitoring in HIV-infected patients
receiving indinavir
Laurent Perello, Cecile Goujard, J.F. Delfraissy, A.M. Taburet. Pharmacie
Hopital, Bicetre 78 Rue Du Gal, Leclerc 94270, Kremlin, Bicetre, France
Indinavir is an HIV protease inhibitor used concomitantly with anti reverse transcriptase nucleosides. Hepatic elimination via CYP3A4 leads to an important
pharmacokinetic variability and to many drug interactions. Concentrations of indinavir (IDV) were evaluated at steady state in 95 HIV infected patients. All patients
received 800 mg TID in addition to one or two nucleoside analogues. Blood
samples for indinavir dosage were drawn at any time after drug intake when
sampling for biological tests was needed. Indinavir was assayed using a reverse
phase HPLC method. The limit of quantification was 40 ng/ml and the inter-assay
variability was below 10% at all concentrations tested (200 ng/ml; 1000 ng/ml;
3500 ng/ml). Variability in indinavir peak concentrations was large ranging from
low concentrations up to 15392 ng/ml. Low value of concentrations lead retrospectively to confirm bad or no compliance after interviewing the patient. Average
trough concentrations is 70 to 200 ng/ml. High concentrations were measured
in three patients with mild hepatic dysfunction (4180 ng/ml; 5698 ng/ml; 6484
ng/ml). Patients concomitantly treated with inducers such as phenobarbital or
rifabutin had decreased concentrations, whereas addition of itraconazole to the
IDV treatment lead to an increase in concentrations. No relationship could be
found between IDV concentration and the decrease of viral load or the increase
in plasma bilirubin observed in patients treated with indinavir.
When viral load does not decrease or increase despite antiviral treatment including indinavir, drug monitoring is useful to identify problems with compliance.
Therapeutic drug monitoring of indinavir could be also useful in patients with
hepatic disease or in case of metabolic drug interactions to maintain a trough
concentration well above the in vitro IC 95% i.e 70 ng/ml.
42273 Pharmacokinetics and oral bioavailability of a novel HIV
protease inhibitor JE-2147 (KNI-764) in animals
Takamasa Ueno1, Keisuke Terashima1, T. Mimoto1, H. Satol, M. Shintani2,
H. Mitsuya3, H. Hayashi'. 1Pharm. & Biotech. Lab. Japan Energy Co. 3-17-35,
Niizo-Minami, Toda, Saitama, 335; 2Japan Energy, Tokyo, Japan; 3 Medicine
Branch, National Cancer Inst. Bethesda, MD, USA
Background: To determine the pharmacokinetics of a novel potent HIV protease
inhibitor JE-2147 (formerly called KNI-764) in animals.
Methods: JE-2147 dissolved in 50% polyethyleneglycol 400 with water was
administered intravenously or intraduodenaly to Sprague-Dawley rats, and intravenously or orally to beagle dogs. Resultant blood levels of JE-2147 were
determined by the specific and quantitative high performance liquid chromatography, and pharmacokinetic parameters were determined using a non-compartment
model.
Results: When the drug was administered intravenously to rats (dose of 10
mg/kg) and dogs (dose of 25 mg/kg), JE-2147 exhibited the distribution volume
of 5.6 and 1.6 liters/kg with the elimination half life of 50 and 94 min, respecin CEM-SS in the presence of 50% human serum. Area-under-the-curve values
of the drug were proportional to the doses when orally administered to dogs,
and feeding status of the dogs prior to oral administration did not substantialy
affect the pharmacokinetic profiles. JE-2147 was well tolerated and no significant
toxicities were found in rats when JE-2147 was repeatedly administered for 14
days at doses of 100, 300, and 1000 mg/kg/day.
Conclusions: JE-2147 has favorable pharmacokinetic properties and toxicity
profiles in animals. Further evaluation of JE-2147 is warranted as a potential
therapeutic for HIV-1 infection.
42274 Steady-state pharmacokinetics of nelfinavir mesylate
(NFV; Viracept") in HIV-infected children
Rolf P.G. van Heeswijk1, H.J. Scherpbier2, M.C.A. Van Leeuwen3,

R.M.W. Hoetelmans', J.M.A. Lange4. 1Slotervaart Hospital Department of
Pharmacy and Pharmacology Louwesweg 6 1066 EC Amsterdam; 2AMC &
Emma Childrens Hospital, Amsterdam; 3Academic Medical Center; 4AMC &
NATEC, Amsterdam, The Netherlands
Background: NFV has recently been approved by the FDA for paediatric use
in a dosage of 20-30 mg/kg tid. However, steady-state pharmacokinetic (PK)

data of this regimen in HIV-infected children are sparse. This study evaluates
the PK of NFV 30 mg/kg tid (paediatric formulation or tablets) in HIV-infected
children.
Methods: In 8 HIV-infected children (age 0.5-10 yrs, weight 3.8-23 kg), who
are treated with NFV 30 mg/kg tid as part of a triple drug regimen (stavudine,
lamivudine and NFV), steady-state PK were determined after one week by drawing 12 blood samples during 8 h. NFV in plasma was quantified with a validated
HPLC assay. The following pharmacokinetic parameters were determined by using non-compartmental pharmacokinetics: area under the plasma concentration
vs. time curve (AUC), minimum and maximum plasma concentration (Cmin and
Cmax), time to Cmax (Tmax) and elimination half-life (tl/2).
Results: Mean pharmacokinetic parameters are (+ s.e.m.): AUC(o-8 h): 36.24
~ 8.24 h*mg/L, Cmin: 3.41 ~ 0.85 mg/L, Cmax: 7.02 ~ 1.21 mg/L, Tmax: 2.25 ~
0.34 h, tl/2: 3.90 ~ 0.48 h. NFV was well tolerated in this population.
Conclusions: In children a NFV dosage of 30 mg/kg tid resulted in adequate
plasma NFV concentrations. These results warrant the evaluation of a more
practical (bid) dosing regimen for NFV in children.
42275 Low plasma levels of indinavir (IDV) are highly predictive
of virological treatment failure in patients using
IDV-containing triple therapy
David M. Burger', R.M.W. Hoetelmans2, J.W. Mulder3, P.L. Meenhorst3,
P.W.H. Hugen', K. Brinkman4, P.P. Koopmans4. ' Dept. of Clinical Pharmacy,
Univ. Hospital, Nijmegen, PO. Box 9101, 6500 HB, Nijmegen; 2Dept. of
Pharmacy, Slotervaart Hospital, Amsterdam; 3Dept. of Internal Medicine,
Slotervaart Hosp., Amsterdam; 4Dept. of Internal Medicine, Univ. Hospital,
Nijmegen, Netherlands
Objectives: To determine the importance of low IDV levels as a cause of virological treatment failure.
Design: Prospective, 24-weeks study.
Methods: All HIV-infected outpatients who were using IDV (800 mg tid) as part
of their triple drug combination therapy were included in this study. Plasma levels
of IDV were determined by HPLC at regular intervals. After 24 weeks of uninterrupted treatment virological failure was assessed if VL was not undetectable
(>200 copies/ml by Roche PCR). IDV plasma levels were compared to population
values measured at the same time interval between drug ingestion and blood
sampling. The ratio between an individual's IDV level and the population value is
called the IDV level ratio.
Results: 65 patients were included. Median baseline CD4 was 100 cells//il
(range: 5-650) and median log VL 4.9 copies/ml (range: 2.3-6.3). 78% of patients
was pretreated with antiretroviral agents, 35% was pretreated with another protease inhibitor (PI). The median IDV level ratio of the 65 patients was 0.96 (range:
0.24-4.78), but if only the lowest ratio of each patient was included, the median
IDV level ratio was 0.69 (range: 0.02-2.81). Univariate analysis demonstrated
that a low IDV level ratio was highly correlated with treatment failure (p = 0.0003).
Patients with a lowest IDV level ratio less than 0.75 had a 3.5 times higher risk
on treatment failure than patients with a lowest IDV level ratio >0.75 (55.9% vs.
16.1%). Multivariate analysis by logistic regression showed that a low IDV level
ratio, a high VL at baseline, and pretreatment with a PI were the only significant
factors in predicting virological treatment failure.
Conclusions: In addition to a high VL at baseline and pretreatment with a PI,
a low plasma level of IDV (less than 75% of population values) is a major risk
factor for virological treatment failure in patients using IDV as part of their triple
drug regimen. Monitoring IDV plasma levels is necessary to optimize triple drug
combination therapy.
42276 | Indinavir pharmacokinetics and their correlation with
virologic and immunologic parameters
David Dalmau', A.O.E. Ochoa De EuchagOen', J.M.L. Martinez-Lacasa1,
C.S.R. Sanchez Rodriguez', J.V. Vidal2, Xercavins3. 1Dr. Robert, 5 Mutua De
Terrassa Hospital Infectious Disease Unit. 082221-Terrassa; 2Laboratory of
Immunology Mutua de Terrassa Terrassa; 3Microbiology Laboratory Terrassa,
Spain
Objective: To evaluate the correlation between plasmatic levels for the HIV protease inhibitor indinavir and their intersubject variability with immunologic (CD4+
cell counts) and virologic (plasma HIV-1 RNA) parameters.
Methods: We prospectively evaluated 31 compliant HIV-1 infected patients
who were using indinavir (800 mg every 8 hours) as part of their antiretroviral
regimen during at least 24 weeks prior to the study. Plasma concentrations of
indinavir were collected and analyzed with a validated HPLC assay at 8 hours
post ingestion and 120 minutes thereafter. Viral load (VL) was routinely measured
by Nuclisens (Organon). Twenty eight patients were protease inhibitor naive prior
to receiving indinavir (mean: 8.6 months: range: 6-14 months).
Results: Eleven out of 31 patients (35%) had undetectable VL (<80 copies/mi),
whereas the remaining 20 patients (65%) had variable plasma HIV-1 RNA levels
(mean: 4.3 log 10; range: 2.5-5). There were no differences in terms of CD4 cell

counts among patients with detectable and undetectable plasma VL (mean: 276
cells//il; range: 60-697; p = 0.3). Mean plasmatic levels for indinavir among both
groups were as follows:

8 h after ingestion (mg/L)
120 minutes post-ingestion (mg/L)

Undetectable VL (range)
1,542 (0.061-11,931)
16,133 (1,473-30,414)

Detectable VL (range)
1,524 (0.016-16,590)
13,712 (1,272-52,153)