Total raltegravir concentrations in cerebrospinal fluid exceed the 50-percent inhibitory concentration for wild-type HIV-1.
ABSTRACT HIV-associated neurocognitive disorders continue to be common. Antiretrovirals that achieve higher concentrations in cerebrospinal fluid (CSF) are associated with better control of HIV and improved cognition. The objective of this study was to measure total raltegravir (RAL) concentrations in CSF and to compare them with matched concentrations in plasma and in vitro inhibitory concentrations. Eighteen subjects with HIV-1 infection were enrolled based on the use of RAL-containing regimens and the availability of CSF and matched plasma samples. RAL was measured in 21 CSF and plasma pairs by liquid chromatography-tandem mass spectrometry, and HIV RNA was detected by reverse transcription-PCR (RT-PCR). RAL concentrations were compared to the 50% inhibitory concentration (IC(50)) for wild-type HIV-1 (3.2 ng/ml). Volunteers were predominantly middle-aged white men with AIDS and without hepatitis C virus (HCV) coinfection. The median concurrent CD4(+) cell count was 276/μl, and 28% of CD4(+) cell counts were below 200/μl. HIV RNA was detectable in 38% of plasma specimens and 4% of CSF specimens. RAL was present in all CSF specimens, with a median total concentration of 14.5 ng/ml. The median concentration in plasma was 260.9 ng/ml, with a median CSF-to-plasma ratio of 0.058. Concentrations in CSF correlated with those in with plasma (r(2), 0.24; P, 0.02) but not with the postdose sampling time (P, >0.50). RAL concentrations in CSF exceeded the IC(50) for wild-type HIV in all specimens by a median of 4.5-fold. RAL is present in CSF and reaches sufficiently high concentrations to inhibit wild-type HIV in all individuals. As a component of effective antiretroviral regimens or as the main antiretroviral, RAL likely contributes to the control of HIV replication in the nervous system.
- [show abstract] [hide abstract]
ABSTRACT: To compare cerebrospinal fluid (CSF) and serum orosomucoid (alpha-1-acid glycoprotein-AAG) concentrations in various subgroups of patients with multiple sclerosis (MS). CSF and serum AAG concentrations, AAG quotient (i.e., CSF AAG/serum AAGx10(3)) and index were determined in a group of 59 patients with clinically definite or probable MS. Patients were subdivided according to the disease form, disease severity according to an expanded disability status scale (EDSS), its treatment, disease duration and sex. CSF AAG was increased in 52.5% of the patients and AAG quotient even in 64.4%. An increase in the CSF AAG concentration, as well as in AAG quotient and index, appear only after several years of disease duration, while no significant correlation with age has been found. This suggests that CSF AAG changes in MS represent a secondary, unspecific phenomenon and that this protein is not relevant for the aethiopathogenesis of the disease. Nevertheless, the finding of subnormal CSF AAG levels in some MS patients in remission (never observed in those in the attack) implies the possibility that CSF AAG may be used as a "state marker" in MS. Serum AAG levels were significantly lower in secondary progressive form and in severely disabled patients. This observation suggest that serum AAG values determination might have some prognostic significance. Further studies are, however, needed. Serum AAG should be investigated in parallel with other CSF and serum protein fractions in order to establish a pannel of examinations enabling multiple statistical analyses. This approach may lead to the finding of a "complex state marker" enabling thus to evaluate more precisely disease course in individual patients and to accept appropriate therapeutic measures.Clinica Chimica Acta 09/2003; 334(1-2):107-10. · 2.85 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Lopinavir (LPV) is highly bound to plasma proteins and is a substrate for active drugs transporters, which may greatly limit the access of LPV to the central nervous system (CNS). However, even low lopinavir concentrations may be sufficient to inhibit HIV replication. Prior anecdotal reports indicated that lopinavir concentrations were below detection in cerebrospinal fluid (CSF). LPV was measured by liquid chromatography/mass spectrometry in 31 CSF-plasma pairs from 26 HIV-infected individuals who were taking LPV-containing antiretroviral regimens. The lower limit of quantification was 3.7 microg/l. Seven of the sample pairs had very low plasma (and CSF) LPV concentrations, with a mean estimated plasma trough of 274 microg/l (range, < 3.7 to 608; typical trough values approximately 4000 microg/l), suggesting poor recent adherence. In the remaining 24 sample pairs, the median LPV concentration was 5889 microg/l [interquartile range (IQR), 4805-9620] and all CSF samples had measurable LPV concentrations: median 17.0 microg/l (IQR, 12.1-22.7). The median CSF-plasma ratio was 0.23% (range, 0.12-0.75). All CSF concentrations in these samples were more than double the 50% inhibitory concentration for wild-type HIV virus. In patients with typical plasma levels of LPV, the drug is detectable in the CSF at concentrations that exceed those needed to inhibit HIV replication. Despite being > 98% bound to plasma proteins, LPV penetrates into the CNS and may contribute to the control of HIV in this potential reservoir.AIDS 06/2005; 19(9):949-52. · 6.41 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: To rigorously evaluate the time course of cognitive change in a cohort of individuals with HIV-associated neurocognitive disorders (HAND) initiating combination antiretroviral therapy (CART), and to investigate which demographic, laboratory, and treatment factors are associated with neuropsychological (NP) outcome (or "any NP improvement"). Study participants included 37 HIV+ individuals with mild to moderate NP impairment who initiated CART and underwent NP testing at 12, 24, 36, and 48 weeks thereafter. NP change was assessed using a regression-based change score that was normed on a separate NP-stable group thereby controlling for regression toward the mean and practice effect. Mixed-effect regression models adjusting for loss to follow-up were used to evaluate the time course of cognitive change and its association with baseline and time-varying predictors. In persons with HAND initiating CART, cognitive improvement happens soon after initiation (13% at week 12), but more often 24, 36, and up to 48 weeks after initiation (up to 41%), with fewer than 5% demonstrating significant worsening. In multivariate analyses, unique predictors of NP improvement included more severe baseline NP impairment and higher CART CNS penetration index. Greater viral load decrease was associated with NP improvement only in univariate analyses. Clinically meaningful neuropsychological improvement seemed to peak around 24-36 weeks after combination antiretroviral therapy initiation and was prolonged over the 1-year study period. This study also provides new evidence that benefit may be maximized by choosing antiretroviral medications that reach therapeutic concentrations in the CNS.Neurology 06/2009; 73(5):342-8. · 8.25 Impact Factor
ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Dec. 2010, p. 5156–5160
Copyright © 2010, American Society for Microbiology. All Rights Reserved.
Vol. 54, No. 12
Total Raltegravir Concentrations in Cerebrospinal Fluid Exceed the
50-Percent Inhibitory Concentration for Wild-Type HIV-1?
David Croteau,1* Scott Letendre,2Brookie M. Best,3,5Ronald J. Ellis,1Sheila Breidinger,7
David Clifford,8Ann Collier,9Benjamin Gelman,10Christina Marra,9Gilbert Mbeo,11
Allen McCutchan,2Susan Morgello,12David Simpson,12Lauren Way,4Florin Vaida,6
Susan Ueland,2Edmund Capparelli,3and Igor Grant4for the CHARTER Group
Departments of Neurosciences,1Medicine,2Pediatrics (Rady Children’s Hospital),3and Psychiatry,4Skaggs School of
Pharmacy and Pharmaceutical Sciences,5and Department of Family and Preventive Medicine, Division of
Biostatistics and Bioinformatics,6University of California—San Diego, San Diego, California;
Merck Research Laboratories, West Point, Pennsylvania7; Washington University, St. Louis,
Missouri8; University of Washington, Seattle, Washington9; University of Texas Medical Branch,
Galveston, Texas10; Johns Hopkins University, Baltimore, Maryland11; and
Mt. Sinai School of Medicine, New York, New York12
Received 14 April 2010/Returned for modification 31 May 2010/Accepted 28 August 2010
HIV-associated neurocognitive disorders continue to be common. Antiretrovirals that achieve higher con-
centrations in cerebrospinal fluid (CSF) are associated with better control of HIV and improved cognition. The
objective of this study was to measure total raltegravir (RAL) concentrations in CSF and to compare them with
matched concentrations in plasma and in vitro inhibitory concentrations. Eighteen subjects with HIV-1
infection were enrolled based on the use of RAL-containing regimens and the availability of CSF and matched
plasma samples. RAL was measured in 21 CSF and plasma pairs by liquid chromatography–tandem mass
spectrometry, and HIV RNA was detected by reverse transcription-PCR (RT-PCR). RAL concentrations were
compared to the 50% inhibitory concentration (IC50) for wild-type HIV-1 (3.2 ng/ml). Volunteers were pre-
dominantly middle-aged white men with AIDS and without hepatitis C virus (HCV) coinfection. The median
concurrent CD4?cell count was 276/?l, and 28% of CD4?cell counts were below 200/?l. HIV RNA was
detectable in 38% of plasma specimens and 4% of CSF specimens. RAL was present in all CSF specimens, with
a median total concentration of 14.5 ng/ml. The median concentration in plasma was 260.9 ng/ml, with a
median CSF-to-plasma ratio of 0.058. Concentrations in CSF correlated with those in with plasma (r2,
0.24; P, 0.02) but not with the postdose sampling time (P, >0.50). RAL concentrations in CSF exceeded
the IC50for wild-type HIV in all specimens by a median of 4.5-fold. RAL is present in CSF and reaches
sufficiently high concentrations to inhibit wild-type HIV in all individuals. As a component of effective
antiretroviral regimens or as the main antiretroviral, RAL likely contributes to the control of HIV
replication in the nervous system.
Cells in the central nervous system (CNS) are infected early
in the course of HIV infection, and HIV RNA is often de-
tected in the cerebrospinal fluid (CSF) of individuals with
chronic disease. This infection can lead to HIV-associated
neurocognitive disorders (HAND), which continue to be com-
mon despite potent combination antiretroviral therapy (ART):
based on recent estimates, 39 to 58% of treated HIV-infected
individuals have global cognitive impairment (12). HIV drives
the pathogenesis of HAND, leading to structural and func-
tional changes to dendrites and synapses (17). Supporting the
importance of the virus in pathogenesis, HIV RNA concentra-
tions in CSF have, in general, been associated with HAND in
cross-sectional and longitudinal studies (9, 10, 18). A substan-
tial proportion of people with HAND do normalize their neu-
ropsychological performance with combination ART, but the
majority do not return to normal functioning, suggesting that
current therapeutic practices undertreat the CNS and can be
HIV in the CNS is thought to be a mixture from blood-
derived and CNS-derived sources, and this mixture can differ
based on factors including CD4?cell count and CSF leukocyte
count. Individuals with higher CD4?cell counts are more
likely to have blood-derived HIV in CSF through trafficking
lymphocytes, while those with more advanced immune sup-
pression, as well as those with moderate to severe HAND, are
more likely to have HIV in CSF that derives from cells in the
CNS, such as macrophages and microglia (8, 25). These
differences are apparent when HIV RNA decay is moni-
tored in CSF and plasma following the initiation of ART,
with dissociated slopes in predominantly CNS derived
sources (7, 8, 15).
Antiretrovirals differ in their distribution in—or penetration
into—the CNS; some drugs penetrate at concentrations similar
to those in blood, and others penetrate at less than 1% of
concentrations in blood. Only antiretrovirals that penetrate
into the CNS in therapeutic concentrations will be able to
reduce HIV replication in glial cells and macrophages. This is
reflected by the relationship between penetration estimates
* Corresponding author. Mailing address: HIV Neurobehavioral
Research Center, University of California—San Diego, 220 Dickinson,
Suite B, #8208, San Diego, CA 92103. Phone: (619) 543-4755. Fax:
(619) 543-1235. E-mail: firstname.lastname@example.org.
?Published ahead of print on 27 September 2010.
and either the level of HIV RNA in CSF or neuropsychological
performance: the higher the penetration estimates, the lower
the HIV RNA levels in CSF (15, 16) and the better the neu-
ropsychological performance (5, 14, 21, 23), although not all
studies are consistent (16).
Raltegravir (RAL) is the first drug in the new class of inte-
grase inhibitors that prevent the covalent insertion of uninte-
grated linear HIV DNA into the host cell genome, ultimately
limiting the formation of HIV provirus. RAL has shown ex-
ceptional potency in suppressing HIV replication, with a mean
plasma viral load change of ?2.2 log10copies/ml after 10 days
of monotherapy (20). The objectives of this analysis were to
measure RAL concentrations in CSF and to estimate whether
they are in the therapeutic range.
MATERIALS AND METHODS
Eighteen subjects enrolled in the project at the HIV Neurobehavioral Re-
search Center (HNRC) at the University of California—San Diego (UCSD)
between April 2007 and February 2009. Five subjects were coenrolled in clinical
trials of the antiviral effectiveness of RAL sponsored by the manufacturer, and
the remaining participants were coenrolled in observational cohort studies, in-
cluding the CNS HIV Antiretroviral Therapy Effects Research (CHARTER)
project. This study and all linking studies were approved by the UCSD Human
Research Protections Program. Eligible subjects had HIV-1 infection, were ART
naïve or experienced, were receiving RAL from either a clinical trial or their
medical provider, were able to give informed consent, and were willing to un-
dergo lumbar punctures. Subjects with contraindications to lumbar puncture,
psychiatric disease potentially interfering with study participation, opportunistic
infections within 30 days, or moderate to severe cognitive impairment were
excluded. Blood and CSF samples were obtained after informed consent was
provided. CSF was obtained by lumbar punctures performed with aseptic tech-
niques by experienced operators using 22-gauge pencil-point needles. Blood was
obtained within 1 h of CSF collection by routine phlebotomy. All specimens were
stored at ?80°C until analysis. The study was designed to distribute the sampling
times evenly after RAL dosing. Twenty-two CSF-plasma pairs were obtained.
One was obtained more than 24 h after the reported dose and was excluded from
analyses. The remaining 21 specimen pairs were obtained from 17 subjects, of
whom 14 provided a single pair each, 2 provided 2 pairs each, and 1 provided 3
pairs. The interval between multiple samplings varied from 2.5 weeks to 5
months, with a median of 4 weeks.
RAL concentrations were measured at Merck Research Laboratories based on
a previously described method (19). In brief, RAL and the internal standard
(13C6-labeled RAL) were measured in plasma and CSF by reverse-phase high-
performance liquid chromatography with tandem mass spectrometry detection
employing an atmospheric-pressure chemical-ionization interface in the positive
ionization mode. Sample preparation consisted of 96-well liquid-liquid extraction
of 200 ?l of plasma or 250 ?l of CSF. The multiple-reaction-monitoring transi-
tions were m/z 445 to 109 for the drug and m/z 451 to 367 for the internal
standard. The lower limit of quantitation (LLOQ) was 2 ng/ml for plasma, with
a linear calibration range from 2 to 1,000 ng/ml, and 0.25 ng/ml for CSF, with a
linear calibration range from 0.25 to 100 ng/ml. Plasma study samples were
analyzed over 2 days. The interday accuracy of the plasma quality control sam-
ples was 104.6 to 108.0%, and the interday precision was 1.3 to 5.3%. CSF study
samples were analyzed in one analytical run. The intraday accuracy of the
CSF quality control samples was 98.7 to 101.2%, and the intraday precision
was 1.9 to 3.0%.
HIV RNA was quantified by reverse transcription-PCR (RT-PCR) with a
Roche TaqMan real-time assay (Roche Diagnostics) with an assay LLOQ of 50
copies/ml. Blood CD4?T-cell counts were determined by flow cytometry, and
hepatitis C virus (HCV) serostatus was determined by immunoassay. The lowest
past CD4?cell count (i.e., the nadir) was obtained by a combination of self-
report and record review.
Descriptive and bivariate statistics were generated using standard methods.
RAL concentrations were log transformed to improve their distribution. Con-
centrations of the drug in CSF were compared to concentrations of the drug in
plasma and to the 50% inhibitory concentration (IC50) for wild-type HIV (3.2
ng/ml) (6). Pearson’s correlation method was used to assess the relationship
between RAL concentrations in plasma and RAL concentrations in CSF. In
addition, pharmacokinetic data were fit to a population semiphysiologic two-
compartment model (ADVAN4, TRANS1) by the NONMEM VI program
(ICON) using a model approach used previously with abacavir (3). RAL con-
centrations in plasma alone and combined plasma-plus-CSF RAL concentrations
were modeled sequentially, with CSF penetration estimated using the FOCE
(first-order conditional-estimation) subroutine, with interaction and separate
residual-error terms for concentrations in plasma and CSF. The covariance
estimation step was required to be successful for the model to be accepted.
Different initial estimates were used to ensure the avoidance of convergence at
a local minimum. The half-life from each compartment was calculated as 0.693
divided by the appropriate rate constant. Given the limited data, no covariate
analysis model building was attempted.
Subjects were predominantly middle-aged (median age, 46
years; range, 32 to 70 years) white (89%) men (94%) who had
AIDS (83%). Nearly all (16/18 [88%]) were HCV seronega-
tive. The median CD4?cell count at the time of sampling was
276/?l (range, 20 to 1,216/?l), with 28% of values falling below
200/?l. The median nadir CD4?cell count was 51/?l (range, 0
to 323/?l). Disease severity was categorized as stage C for 55%,
stage B for 17%, and stage A for 28% based on the CDC
classification system. The median plasma HIV RNA concen-
tration was 1.70 log10copies/ml (interquartile range [IQR],
1.70 to 2.25).
The median duration of RAL use was 4.2 months (IQR, 1.4
to 27.4 months). All subjects reported the use of 400 mg of
RAL twice daily. Concurrent antiretrovirals included one to
three nucleoside/nucleotide reverse transcriptase inhibitors
(NRTIs) for all but 1 patient (94%), one non-NRTI for 6
patients (33%), one ritonavir-boosted protease inhibitor for 11
patients (61%), and one fusion inhibitor for 5 patients (28%).
All subjects reported taking at least 95% of their antiretroviral
doses in the 4 days preceding sampling, except for one. The last
RAL dose was taken with food prior to sampling in 20 of 21
RAL concentrations in plasma and CSF are displayed in
Fig. 1, and aggregate data are summarized in Table 1. RAL
was present in all CSF specimens, with a median concentration
of 14.5 ng/ml (IQR, 9.3 to 26.1 ng/ml). The median concen-
tration in plasma was 260.9 ng/ml (IQR, 72.0 to 640.4 ng/ml).
The median CSF-to-plasma ratio was 0.058 (IQR, 0.021 to
FIG. 1. Raltegravir concentrations in plasma (squares) and CSF
(circles). Horizontal dashed line, IC50; vertical dashed line, end of the
typical dosing interval. Solid and dashed boldface lines, plasma and
CSF population pharmacokinetic modeling.
VOL. 54, 2010RALTEGRAVIR IN CSF EXCEEDS WILD-TYPE HIV-1 IC50
0.18). The CSF-to-plasma ratio increased across the dosing
interval, with higher ratios later in the dosing interval (r2, 0.25;
P, 0.02). Sampling was moderately well balanced across the
dosing interval, with 3 pairs obtained between 1 and 2 h after
the dose, 7 pairs between 2 and 4 h, 5 pairs between 5 and 8 h,
5 pairs between 8 and 12 h, and 1 pair at 15 h. RAL concen-
trations in CSF exceeded the IC50for wild-type HIV in all
specimens by a median of 4.5-fold (IQR, 2.7- to 7.7-fold). HIV
RNA was undetectable in 20 of 21 (95%) CSF specimens and
in 13 of 21 (62%) plasma specimens. There was a statistically
significant correlation between RAL concentrations in CSF
and those in plasma (r2, 0.24; P, 0.02) (Fig. 2) but not
between RAL concentrations and the postdose sampling
time (P, ?0.50). Higher RAL concentrations in plasma were not
associated with undetectable HIV RNA levels in plasma (t, 0.37;
P, ?0.20). Too few CSF samples had detectable levels of HIV
RNA to enable reliable statistical analyses comparing HIV
RNA levels with RAL concentrations in CSF. The results were
not substantially different when the analyses were performed
with the inclusion of a single time point for each subject, using
only the first time point for subjects with multiple CSF-plasma
pairs. Population pharmacokinetic modeling estimated the typ-
ical CSF-to-plasma ratio at 0.078 and showed that the average
RAL half-life for the population was longer in CSF than in
plasma. Possible interactions between RAL and other se-
lected antiretrovirals, such as atazanavir (n, 3), efavirenz (n,
1), etravirine (n, 6), and ritonavir (n, 9), were explored despite
the small sample size. RAL levels in plasma or CSF did not
differ between those who reported use of these drugs and those
who did not, with the exception of atazanavir. RAL concen-
trations in plasma (t, 3.0; P, 0.008) and in CSF (t, 2.3; P, 0.03)
were higher for those reporting atazanavir use than for those
who used other drugs.
RAL concentrations in plasma were highly variable, a find-
ing consistent with prior reports (24). RAL concentrations in
CSF were relatively less variable and exceeded the concentra-
tion required to inhibit wild-type HIV in vitro in all individuals
(median, 4.5-fold higher than the IC50for wild-type HIV). The
concentrations in CSF are consistent with the physicochemical
characteristics of RAL, including its low molecular weight and
its relatively low level of plasma protein binding. Although only
total drug concentrations were measured, it appears unlikely
that protein binding in CSF would result in unbound RAL
concentrations below the IC50for wild-type HIV, given the
much lower concentration of binding proteins, such as albumin
and alpha-1 acid glycoprotein, in CSF than in plasma (100 to
1,000-fold lower) and the findings of a recent pharmacokinetic
study measuring bound and unbound indinavir in plasma and
CSF (1, 11). The correlation between RAL concentrations in
CSF and those in plasma, although statistically significant,
demonstrated some variability, which may suggest some limi-
tation in using plasma RAL concentrations to estimate RAL
concentrations in the nervous system. RAL is 83% bound to
plasma proteins, and assuming that all free drug penetrates
into the CSF, one would expect the concentrations in CSF to
be approximately 15 to 20% of those in plasma. The lower
CSF-to-plasma ratio of 0.058 suggests that there may be other
barriers to CSF penetration, but these barriers are not as great
as those seen with protease inhibitors (2, 4, 13). The discor-
dance between the CSF and plasma RAL concentration curves
in the population pharmacokinetic model suggests that the
influx into and efflux from the CSF are slower than elimination
from plasma. The biphasic plasma pharmacokinetics of RAL,
and the fact that samples were not evenly distributed during
the postdosing interval corresponding to this biphasic pattern,
with more samples between 0 and 4 h, may have led to some
underestimation of the CSF RAL concentrations. However,
population pharmacokinetic modeling partly helps compen-
sate for the uneven sampling times of a formal intensive phar-
macokinetic sampling approach.
Although 38% of the pairs had detectable plasma viral loads,
HIV was undetectable in all but one CSF sample. This rela-
tively high proportion of subjects with detectable plasma viral
loads despite a median RAL use duration of 4 months is
consistent with the fact that RAL is often used in highly treat-
ment experienced patients with suspected or proven resistance
to other classes of antiretrovirals. Although no pretreatment
CSF viral loads were measured in this analysis, the high rate of
HIV suppression in CSF supports the potency of RAL in the
CNS. The single subject with persistently detectable viral loads
in CSF (222 copies/ml) was reportedly receiving 400 mg of
RAL twice daily for 33 months, was on a five-drug regimen,
FIG. 2. Correlation of CSF and plasma raltegravir concentrations
in log-transformed values (circles). Solid line, linear fit; dashed lines,
95% confidence intervals.
TABLE 1. Summary of concentration and sampling data
Value (ng/ml) for:Ratio
14.5260.9 0.058 4.5
aExperiments were performed with 21 pairs of CSF and plasma samples. CSF
samples were collected at 6.1 ? 4.2 h postdose, and plasma samples were
collected at 6.2 ? 4.3 h postdose; the difference between the times of blood and
CSF collection was 0.10 ? 0.54 h.
5158CROTEAU ET AL.ANTIMICROB. AGENTS CHEMOTHER.
had a CSF RAL concentration below the median (9.0 ng/ml),
and had the highest plasma viral load (60,100 copies/ml) of the
The 33% dissociation observed between HIV RNA suppres-
sion in plasma (62% of pairs) and in CSF (95% of pairs)
suggests that RAL may have relatively greater potency in mi-
croglia and macrophages, which primarily produce HIV in the
nervous system, than in lymphocytes, which primarily replicate
HIV in the blood. A second explanation, not necessarily ex-
cluded by the first, is that HIV may have adapted to (or “com-
susceptible strains in the CSF and resistant strains in blood. A
third explanation, which is compatible with the preceding two
explanations, is that the nervous system is relatively protected
by the limited immune activation that can occur with failing
antiretroviral therapy (compared with no antiretroviral therapy
at all) (22). Although an antiretroviral regimen with drugs with
greater effectiveness at CNS penetration and/or a higher un-
bound fraction could cause such dissociation, the 7 pairs with
plasma-CSF dissociation came from patients on various anti-
retroviral regimens, suggesting that this is a less likely expla-
nation here. Prospective, controlled clinical trials are needed in
order to better understand the contribution of these and other
mechanisms to the control of HIV in the nervous system, but
the data support the notion that RAL-containing antiretroviral
regimens continue to control HIV in the nervous system even
after they fail in blood.
The CSF RAL concentrations in this analysis (median, 14.5
ng/ml) are comparable to those recently reported in another
small study (median, 18.4 ng/ml), by Yilmaz et al. (26). Our
somewhat lower median RAL concentration in CSF may be
explained by the lower plasma concentrations in our analysis
(medians, 260.9 ng/ml versus 448 ng/ml). This difference be-
tween plasma concentrations likely accounts for the differences
in CSF-to-plasma ratios between the studies (0.058 versus
0.03). Other potential factors responsible for the differences
include specimen collection, processing, and storage; the RAL
assay system; baseline subject characteristics (e.g., age, body
mass index); and RAL adherence. The distribution of postdos-
ing sampling times appears similar to ours, making this factor
unlikely to explain the difference. The reports also differ in
their conclusions about how well RAL may treat HIV in the
nervous system: our report concludes that all CSF RAL con-
centrations fall in the therapeutic range, but Yilmaz et al.
conclude that approximately 50% do. This apparent disagree-
ment is due to the standard used to support the conclusion.
The IC50was used in this and prior reports, since it is generally
less variable than the standard used by Yilmaz et al., the 95%
inhibitory concentration (IC95), and is commonly referenced in
clinical resistance testing reports. Each approach has its merits,
and combining them in the interpretation of antiretroviral con-
centrations in CSF may enable additional differentiation be-
tween drugs in the future. In this instance, for example, 100%
of our concentrations exceed the IC50of 3.2 ng/ml, and 48%
(10 of 21) exceed the upper limit of the IC95range referenced
by Yilmaz et al. (15 ng/ml). On the other hand, approximately
20% of the concentrations reported by Yilmaz et al. were
below the IC50of 3.2 ng/ml, compared to none in this study.
Therefore, our data appear to indicate more consistent CSF
activity than previously reported. However, the best approach
to estimating the effectiveness of antiretrovirals in the nervous
system has yet to be determined and will likely require addi-
tional in vitro and in vivo work.
In summary, we conclude that RAL achieves therapeutic
concentrations in CSF and, as a component of a combination
antiretroviral regimen or as the main antiretroviral, likely con-
tributes to the control of HIV replication in the nervous sys-
tem. Control of HIV in the nervous system should protect
individuals from HAND and support neurocognitive recovery
for those who have been diagnosed with HAND previously.
Additional work, such as a prospective clinical trial, will be re-
quired to definitively characterize the effectiveness and safety of
RAL in the CNS.
This study was supported by an investigator-initiated grant from
Merck & Co., Inc., and the U.S. National Institutes of Health
(MH62512 to I. Grant; NIH R01 MH58076 to R. Ellis).
1. Adam, P., O. Sobek, L. Ta ´borsky ´, T. Hildebrand, O. Tutterova ´, and P. Za ´cek.
2003. CSF and serum orosomucoid (alpha-1-acid glycoprotein) in patients
with multiple sclerosis: a comparison among particular subgroups of MS
patients. Clin. Chim. Acta 334:107–110.
2. Best, B. M., S. L. Letendre, E. Brigid, D. B. Clifford, A. C. Collier, B. B.
Gelman, J. C. McArthur, J. A. McCutchan, D. M. Simpson, R. Ellis, E. V.
Capparelli, and I. Grant, for the CHARTER Group. 2009. Low atazanavir
concentrations in cerebrospinal fluid. AIDS 23:83–87.
3. Capparelli, E., S. L. Letendre, R. J. Ellis, P. Patel, D. Holland, and J. A.
McCutchan. 2005. Abacavir population pharmacokinetics in plasma and
CSF. Antimicrob. Agents Chemother. 49:2504–2506.
4. Capparelli, E. V., D. Holland, C. Okamoto, B. Gragg, J. Durelle, J. Marquie-
Beck, G. van den Brande, R. Ellis, and S. Letendre; the HNRC Group. 2005.
Lopinavir concentrations in cerebrospinal fluid exceed the 50% inhibitory
concentration for HIV. AIDS 19:949–952.
5. Cysique, L. A., F. Vaida, S. Letendre, S. Gibson, M. Cherner, S. P. Woods,
J. A. McCutchan, R. K. Heaton, and R. J. Ellis. 2009. Dynamics of cognitive
change in impaired HIV-positive patients initiating antiretroviral therapy.
6. Danovich, R., Y. Ke, H. Wan, B. Y. Nguyen, H. Teppler, W. Schleif, D.
Hazuda, and M. Miller, for the BENCHMRK-1 and BENCHMRK-2 Study
Groups. 2008. Raltegravir has similar in vitro antiviral potency, clinical
efficacy, and resistance patterns in B subtype and non-B subtype HIV-1,
abstr. TUAA0302, p. 305. Abstract book, vol. 1. XVII Int. AIDS Conf.,
Mexico City, Mexico, 3 to 8 August 2008. http://www.aids2008-abstracts.org
7. Ellis, R. J., M. E. Childers, J. D. Zimmerman, S. D. Frost, R. Deutsch, J. A.
McCutchan, and the HIV Neurobehavioral Research Center Group. 2003.
Human immunodeficiency virus-1 RNA levels in cerebrospinal fluid exhibit
a set point in clinically stable patients not receiving antiretroviral therapy.
J. Infect. Dis. 187:1818–1821.
8. Ellis, R. J., A. C. Gamst, E. Capparelli, S. A. Spector, K. Hsia, T. Wolfson,
I. Abramson, I. Grant, and J. A. McCutchan. 2000. Cerebrospinal fluid HIV
RNA originates from both local CNS and systemic sources. Neurology 54:
9. Ellis, R. J., K. Hsia, S. A. Spector, J. A. Nelson, R. K. Heaton, M. R. Wallace,
I. Abramson, J. H. Atkinson, I. Grant, and J. A. McCutchan. 1997. Cere-
brospinal fluid human immunodeficiency virus type 1 RNA levels are ele-
vated in neurocognitively impaired individuals with acquired immunodefi-
ciency syndrome. HIV Neurobehavioral Research Center Group. Ann.
10. Ellis, R. J., D. J. Moore, M. E. Childers, S. Letendre, J. A. McCutchan, T.
Wolfson, S. A. Spector, K. Hsia, R. K. Heaton, and I. Grant. 2002. Progres-
sion to neuropsychological impairment in human immunodeficiency virus
infection predicted by elevated cerebrospinal fluid levels of human immu-
nodeficiency virus RNA. Arch. Neurol. 59:923–928.
11. Haas, D. W., B. Johnson, J. Nicotera, V. L. Bailey, V. L. Harris, et al. 2003.
Effects of ritonavir on indinavir pharmacokinetics in cerebrospinal fluid and
plasma. Antimicrob. Agents Chemother. 47:2131–2137.
12. Letendre, S., R. Ellis, R. Heaton, I. Grant, and A. McCutchan. 2009. Neu-
rocognitive complications of HIV and their management, abstr. 181. 16th
Conf. Retrovir. Opportunistic Infect., Montreal, Quebec, Canada, 8 to 11
13. Letendre, S. L., E. V. Capparelli, R. J. Ellis, J. A. McCutchan, and the HIV
Neurobehavioral Research Center Group. 2000. Indinavir population phar-
VOL. 54, 2010RALTEGRAVIR IN CSF EXCEEDS WILD-TYPE HIV-1 IC50
macokinetics in plasma and cerebrospinal fluid. Antimicrob. Agents Che-
14. Letendre, S. L., J. A. McCutchan, M. E. Childers, S. P. Woods, D. Lazza-
retto, R. K. Heaton, I. Grant, and R. J. Ellis; HNRC Group. 2004. Enhancing
antiretroviral therapy for human immunodeficiency virus cognitive disorders.
Ann. Neurol. 56:416–423.
15. Letendre, S. L., G. van den Brande, A. Hermes, S. P. Woods, J. Durelle, J. M.
Beck, J. A. McCutchan, C. Okamoto, R. J. Ellis, and the HIV Neurobehav-
ioral Research Center Group. 2007. Lopinavir with ritonavir reduces the
HIV RNA level in cerebrospinal fluid. Clin. Infect. Dis. 54:1511–1517.
16. Marra, C. M., Y. Zhao, D. B. Clifford, S. Letendre, S. Evans, K. Henry, R. J.
Ellis, B. Rodriguez, R. W. Coombs, G. Schifitto, J. C. McArthur, K. Robert-
son, and the AIDS Clinical Trials Group 736 Study Team. 2009. Impact of
combination antiretroviral therapy on cerebrospinal fluid HIV RNA and
neurocognitive performance. AIDS 23:1359–1366.
17. Masliah, E., R. K. Heaton, T. D. Marcotte, R. J. Ellis, C. A. Wiley, M.
Mallory, C. L. Achim, J. A. McCutchan, J. A. Nelson, J. H. Atkinson, and I.
Grant. 1997. Dendritic injury is a pathological substrate for human immu-
nodeficiency virus-related cognitive disorders. HNRC Group. The HIV Neu-
robehavioral Research Center. Ann. Neurol. 42:963–972.
18. McArthur, J. C., D. R. McClernon, M. F. Cronin, T. E. Nance-Sproson, A. J.
Saah, M. St Clair, and E. R. Lanier. 1997. Relationship between human
immunodeficiency virus-associated dementia and viral load in cerebrospinal
fluid and brain. Ann. Neurol. 42:689–698.
19. Merschman, S. A., P. T. Vallano, L. A. Wenning, B. K. Matuszewski, and
E. J. Woolf. 2007. Determination of HIV integrase inhibitor, MK-0518
(raltegravir) in human plasma using 96-well liquid-liquid extraction and
HPLC-MS/MS. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 857:
20. Murray, J. M., S. Emery, A. D. Kelleher, M. Law, J. Chen, D. J. Hazuda,
B. Y. Nguyen, H. Teppler, and D. A. Cooper. 2007. Antiretroviral therapy
with the integrase inhibitor raltegravir alters decay kinetics of HIV, signifi-
cantly reducing the second phase. AIDS 21:2315–2321.
21. Simpson, D. M. 1999. Human immunodeficiency virus-associated dementia:
review of pathogenesis, prophylaxis, and treatment studies of zidovudine
therapy. Clin. Infect. Dis. 29:19–34.
22. Spudich, S., N. Lollo, T. Liegler, S. G. Deeks, and R. W. Price. 2006.
Treatment benefit on cerebrospinal fluid HIV-1 levels in the setting of
systemic virological suppression and failure. J. Infect. Dis. 194:1686–1696.
23. Tozzi, V., P. Balestra, M. F. Salvatori, C. Vlassi, G. Liuzzi, M. L. Giancola,
M. Giulianelli, P. Narciso, and A. Antinori. 2009. Changes in cognition
during antiretroviral therapy: comparison of 2 different ranking systems to
measure antiretroviral drug efficacy on HIV-associated neurocognitive dis-
orders. J. Acquir. Immune Defic. Syndr. 52:56–63.
24. Wenning, L. A., A. S. Petry, J. T. Kost, B. Jin, S. A. Breidinger, I. DeLepe-
leire, E. J. Carlini, S. Young, T. Rushmore, F. Wagner, N. M. Lunde, F.
Bieberdorf, H. Greenberg, J. A. Stone, J. A. Wagner, and M. Iwamoto. 2009.
Pharmacokinetics of raltegravir in individuals with UGT1A1 polymorphisms.
Clin. Pharmacol. Ther. 85:623–627.
25. Wiley, C., R. Schrier, J. Nelson, P. W. Lampert, and M. B. Oldstone. 1986.
Cellular localization of human immunodeficiency virus infection within the
brains of acquired immune deficiency syndrome patients. Proc. Natl. Acad.
Sci. U. S. A. 83:7089–7093.
26. Yilmaz, A., M. Gisslen, S. Spudich, E. Lee, A. Jayewardene, F. Aweeka, and
R. W. Price. 2009. Raltegravir cerebrospinal fluid concentrations in HIV-1
infection. PLoS One 4:e6877.
5160 CROTEAU ET AL.ANTIMICROB. AGENTS CHEMOTHER.