![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 303, Issue 3, 928-936, December 2002
Laboratory of Biopharmacy and Pharmaceutical Technology, UMR Centre National de la Recherche Scientifique 8612, Faculty of Pharmacy, Châtenay-Malabry, France (I.B., H.C., P.C.), Laboratory of Pharmacology and New Treatments of Cancers, UPRES EA, Institut Gustave-Roussy, Villejuif, France (J.M., G.A., G.V.), Department of Anatomopathology, Institut Gustave-Roussy, Villejuif, France (M-J.T-L.), and Department of Pediatric Oncology, Institut Gustave-Roussy, Villejuif, France (G.V.)
 |
Abstract |
|---|
 Top
 Abstract
 Introduction
Materials and Methods
Results
Discussion
References
|
|---|
The aim of the present study was to evaluate the tumor accumulation of radiolabeled long-circulating poly(ethylene glycol) (PEG)-coated hexadecylcyanoacrylate nanospheres and non-PEG-coated hexadecylcyanoacrylate nanospheres (used as control), after intravenous injection in Fischer rats bearing intracerebrally well established 9L gliosarcoma. Both types of nanospheres showed an accumulation with a retention effect in the 9L tumor. However, long-circulating nanospheres concentrated 3.1 times higher in the gliosarcoma, compared with non-PEG-coated nanospheres. The tumor-to-brain ratio of pegylated nanospheres was found to be 11, which was in accordance with the ratios reported for other carriers tested for brain tumor targeting such as long-circulating liposomes or labels for magnetic resonance imaging. In addition, a 4- to 8-fold higher accumulation of the PEG-coated carriers was observed in normal brain regions, when compared with control nanospheres. Using a simplified pharmacokinetic model, two different mechanisms were proposed to explain this higher concentration of PEG-coated nanospheres in a tumoral brain. 1) in the 9L tumor, the preferential accumulation of pegylated nanospheres was attributable to their slower plasma clearance, relative to control nanospheres. Diffusion/convection was the proposed mechanism for extravasation of the nanospheres in the 9L interstitium, across the altered blood-brain barrier. 2) In addition, PEG-coated nanospheres displayed an affinity with the brain endothelial cells (normal brain region), which may not be considered as the result of a simple diffusion/convection process. The exact underlying mechanism of such affinity deserves further investigation, since it was observed to be as important as specific interactions described for immunoliposomes with the blood-brain barrier.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Clinical
studies have shown that many brain tumors, especially primary tumors,
are among the most resistant to chemotherapy, presumably due to the
presence of a tight blood-brain barrier (Tamai and Tsuji, 1996
).
However, the blood-brain barrier can be selectively disrupted at the
tumor site, which opens the possibility for carrier-mediated
chemotherapy with more active compounds than the widely used nitrosourea.
Indeed, the vasculature of gliomas, like that of other solid tumors,
may present some peculiarities, which all form the basis of an
increased microvascular permeability. Some of these peculiarities include open endothelial gaps (interendothelial junctions and transendothelial channels; diameter of about 0.3 µm), fenestrations (maximum channel width of 5.5 nm), and cytoplasmic vesicles such as
caveolae (diameter of 50-70 nm) and vesicular vacuolar
organelles (diameter of 108 ± 32 nm). An increase in
vessel wall thickness has also been reported, which was attributed to
endothelial cell hyperplasia, reflecting an increase in nonselective
transendothelial transport. All these characteristics, which may be due
in major part to the secretion of the vascular permeability factor
(vascular endothelial growth factor) and which can lead to a loss of
blood-brain barrier function in the case of cerebral malignancies,
would be beneficial for transvascular transport of drugs (for a review, see Vajkoczy and Menger, 2000).
Another challenge lies in the fact that, despite being highly
vascularized (Vajkoczy and Menger, 2000), brain tumors are
exposed to subtherapeutic drug concentrations, as many compounds
possess a short plasma half-life, giving them slim chances to circulate through the tumor vascular bed and to permeate in the interstitium.
An emerging approach for improving systemic chemotherapy is the use of
surface-modified carriers. Coating the carriers with hydrophilic
polymers such as polyethylene glycol (PEG) gives them a form of steric
barrier against interactions with plasma proteins, such as opsonins and
lipoproteins. As a result, PEG-coated carriers evade capture by the
mononuclear phagocyte system (MPS), resulting in a longer circulation
time in the blood compartment (Storm et al., 1995; Peracchia et al.,
1999). This prolonged plasma half-life will increase the chances for
the sterically stabilized carriers to passively extravasate in the
tumor interstitium, across the leaky and hyperpermeable blood-brain
barrier. This approach should also permit the selective delivery of
chemotherapeutics at the pathological site and thus to enhance its
effectiveness, while sparing the other tissues from drug exposure
(Gabizon, 1995).
This strategy has been conducted in many solid tumors, including
glioma; for example, intracerebral implanted 9L tumors appeared hyperintense relative to brain on T1-weighted magnetic resonance images, after intravenous injection of long-circulating dextran-coated iron oxide particles to rats (Moore et al., 2000). In much the same
fashion, Sinerem, another long-circulating label for magnetic resonance
imaging, was successfully used for the detection of brain tumors in
patients (Enochs et al., 1999).
As far as tumor therapy is concerned, a study related that the use of
long-circulating liposomes as cytotoxic drug carriers resulted in
enhanced drug exposure and improved therapeutic activity, when compared
with the nonencapsulated drug used on the same animal brain tumor model
(T-749 in syngeneic Fischer rats) (Siegal et al., 1995). These results
were very much in line with those reported in a rat 9L gliosarcoma
tumor (Sharma et al., 1997). Finally, in patients, the radiolabeled
Stealth pegylated carrier Caelyx (liposomal doxorubicin) showed a high
intratumoral accumulation in glioblastoma and in metastatic brain
tumors (Koukourakis et al., 2000). Caelyx also made possible a
long-term stabilization in patients with recurrent malignant glioma
(Fabel et al., 2001).
As an alternative to sterically stabilized liposomes or labeled ferrous
nanoparticles, this work proposes new polymeric nanoparticles as
carriers for cytotoxic drugs against glioma. The advantage of
nanoparticles over liposomes arises from the possibility of controlling
drug release depending on the polymer used. Moreover, nanoparticles are
known to be more stable in the bloodstream, due to reduced interactions
and exchanges with blood components (Lasic, 1998). Finally,
cyanoacrylate nanoparticles were found to be able to reverse
P-glycoprotein-mediated multidrug resistance by a specific mechanism
(Colin de Verdière et al., 1997).
The above characteristics make long-circulating poly(cyanoacrylate) nanoparticles a potentially advantageous delivery system for brain tumor chemotherapy. Hence, this study aimed to evaluate the ability of these nanoparticles to accumulate in a rodent cerebral gliosarcoma model. Moreover, a pharmacokinetic model was proposed, to understand the mechanisms involved in the uptake of PEG-coated nanoparticles in both normal and tumoral brains.
| |
Materials and Methods |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Materials.
Control nanospheres and sterically stabilized
nanospheres were prepared with poly(hexadecylcyanoacrylate) (PHDCA) and
poly(MePEG2000cyanoacrylate-co-hexadecylcyanoacrylate) 1:4 (PEG-PHDCA), respectively. These polymers and copolymers were synthesized as described before (Brigger et al., 2000). Their 14C-radiolabeled counterparts were prepared at
the Commissariat à l'Energie Atomique (Saclay, France) according
to the previously described protocols (Peracchia et al., 1999; Brigger
et al., 2000). [14C]PHDCA had a specific
activity of 5.8 µCi/mg, whereas the activity of
[14C]PEG-PHDCA was 1.5 µCi/mg. Pluronic F68,
a nontoxic tensio-active agent used in the preparation of the pegylated
nanospheres, was purchased from BASF Corp. (Worcester, MA). Glucose 5%
(Laboratoires Aguettant, Lyon, France) was used as the
dispersion medium for the purified nanospheres. The hydrophilic tracer
[3H]sucrose (specific activity, 16.1 mCi/mg;
radioactive concentration, 1.0 mCi/ml) was obtained from Amersham
Pharmacia Biotech (Les Ulis, France). All the solvents and other
reagents used were of analytical grade.
) was kindly offered by Dr. V. W. van
Beusechem (Division of Gene Therapy, Department of Medical Oncology,
Vrije Universiteit Medical Center Amsterdam, Amsterdam, The
Netherlands). Cell culture medium and additional products were
purchased from Invitrogen (Cergy Pontoise, France).
Biodegradable Cyanoacrylate Nanosphere Preparation and
Characterization.
Biodegradable [14C]PHDCA
and [14C]PEG-PHDCA nanospheres were obtained by
the method of nanoprecipitation (Fessi et al., 1989). Typically, for
the control nanospheres, [14C]PHDCA (18.1 mg
corresponding to 105 µCi) and PHDCA (281.9 mg; isotopic dilution,
1:15.6) were dissolved under slight warming in acetone (15 ml). This
organic phase was then rapidly poured through a syringe in milliQ water
(30 ml) (Millipore), under magnetic stirring. Nanosphere precipitation
occurred instantaneously. Directly afterward, acetone was evaporated
under reduced pressure. The colloidal suspension was purified by
ultracentrifugation (145,000g, 1.5 h, 4°C; Beckman
L7-55 ultracentrifuge; Beckman Coulter, Inc.), and the resulting
nanosphere pellet was suspended in a solution of 5% glucose (12.0 ml),
to obtain the following relation: 25 mg of radiolabeled
nanospheres/8.75 µCi/ml. The suspension was finally filtered on a
sintered glass membrane (Millex AP 20; Millipore, St-Quentin en
Yvelines, France). For the PEG-coated nanospheres, only 50% (w/w) of
the PEG-PHDCA copolymer precipitated as nanospheres (the remaining
fraction had a high PEG content and was thus soluble in the dispersion
medium) (Brigger et al., 2000). Hence, the organic phase consisted of
[14C]PEG-PHDCA (140.0 mg corresponding to 210 µCi) and PEG-PHDCA (460.0 mg; isotopic dilution, 1:3.3) which were
both dissolved in warm acetone (30 ml).
14C-pegylated nanospheres were then obtained in
the same way as the radiolabeled control nanospheres, in 60 ml of
milliQ water containing Pluronic F68 (1%, w/v). Purification and
resuspension steps were equivalent, as for
[14C]PHDCA nanospheres.
45 mV), which was found to be partially shielded when the PHDCA
polymer was replaced by its pegylated counterpart (surface potential of
PEG-coated nanospheres, 20 mV). In this latter case, the
tensio-active Pluronic F68 may also have contributed to raising the
surface charge.
Finally, it was important for the validation of the biodistribution and
pharmacokinetic studies to check whether the 14C
label remained stably bound to purified PHDCA or PEG-PHDCA carriers during in vivo experiments. To answer this question,
14C-radiolabeled PHDCA or PEG-PHDCA nanospheres
were incubated in vitro, in Fischer rat plasma, at 37°C (experiment
realized in triplicate). The nanospheres/plasma ratio was the same as
that obtained in vivo (see below). At predetermined time intervals (5 min, 30 min, 4 h, and 8 h), an aliquot of the plasma medium was withdrawn and ultrafiltrated (Ultrafree-MC, 100-kDa cutoff; Millipore) at 10,000g during 20 min. The radioactivity of
the ultrafiltrate, containing soluble polymer or degradation products, was then measured by liquid scintillation counting. The results were
expressed as a percentage of the total measured radioactivity in the
plasma medium.
Animals. Male CD 344 Fischer rats, weighing 250 to 300 g, were purchased from Charles River Laboratories (L'Arbresles, France). The animals were housed in the Animal Experiment Unit of the Institut Gustave-Roussy and had free access to food and water. Experiments were carried out in compliance with the conditions established by the European Community (Directive n°86/609/CEE).
Brain Tumor Model. 9L gliosarcoma cells were cultured in a 5% CO2 incubator at 37°C, in Dulbecco's modified Eagle's medium with Glutamax, supplemented with 10% inactivated fetal calf serum, penicillin (100 U/ml), streptomycin (100 µg/ml), and amphotericin B (1.5 µg/ml). A minimum of 2 weeks of in vitro culture was allowed before in vivo engraftment of the 9L cells. The rats were anesthetized by an intraperitoneal administration of xylazine (11 mg/kg)-ketamine (73 mg/kg). Through a small burr hole, 10 µl containing 2 × 105 9L cells in Dulbecco's modified Eagle's medium were inoculated with a Hamilton syringe in the right hemisphere, at the following coordinates: 2 mm lateral from the midline, 2 mm anterior to bregma, and 3 mm deep. Control animals received an injection of an equal volume of 0.9% NaCl, without tumor cells. Preliminary experiments with intravenously injected Blue Evans revealed that the blood-brain barrier was locally disrupted at the pathological site 7 days after tumor implantation.
Tissue Distribution Study. The tissue distribution study was performed 7 days after intracerebral inoculation of tumor cells or NaCl (control group) in rats. Tumor-bearing or control animals received an intravenous bolus injection (450 µl) of [14C]PHDCA or [14C]PEG-PHDCA nanospheres (400 µl, 3.5 µCi, 40 mg/kg), with [3H]sucrose (50 µl, 15 µCi) added in the dispersion medium of the nanospheres (5% glucose). After 3 min (one tumor-bearing rat), 5 min (four tumor-bearing rats), 30 min, and 4 h (four control and four tumor-bearing rats), animals were anesthetized by intraperitoneal administration of xylazine (11 mg/kg) and ketamine (73 mg/kg), and blood was obtained by cardiac puncture and collected into tubes containing anticoagulant (lithium-heparin). The rats were then subjected to a transcardiac perfusion of heparin-containing saline (150 ml/10 min), to clear the intravascular compartment from any residual circulating marker.
An aliquot of the blood sample was centrifuged to separate the cellular components from the plasma fraction. The radioactivity contained in the blood, plasma, and, if possible, the urine was finally analyzed by routine liquid scintillation counting. Various organs including liver, lung, spleen, kidney, and muscle were dissected and weighed, and their activity was measured by liquid scintillation after dissolution with Soluène 350 (PerkinElmer Life Sciences, B.V. Groningen, The Netherlands). The brain and cerebellum were also removed for analysis and the following regions were dissected: the visible 9L tumor mass or the intracerebral injection site (control animals), the 2- to 4-mm brain area surrounding the tumor or injection site, and the matching contralateral brain region. In the calculation of the radioactivity distribution, muscle and blood volume were assumed to be, respectively, 45 and 6.5% of the total body weight (Ritschel, 1987). For blood partition calculation, the plasma
volume was estimated to be 45% of the total blood volume (Ritschel,
1987).
Pharmacokinetic Modelization and Tumor Uptake Rate.
To
understand how the nanospheres accumulated in the normal or tumoral
brain, a pharmacokinetic model was built up. A simplified two-compartment open model allowed separation of normal or tumoral brain uptake from biodistribution and excretion phenomena. Thus, the
pharmacokinetic model consisted of a central compartment (blood pool)
and the normal or tumoral brain. The normal (k1) or tumoral (k1') brain uptake rates as well as the rate constant
"blood elimination" (ke), which included all
processes that lead to the removal of the carrier from the blood, such
as excretion, phagocytic clearance, and biodistribution into organs and
tissues other than the brain, were assumed to be unidirectional.
k1, k1', and ke
(min
1) were the arithmetic means
obtained for each rat at the time points mentioned earlier (see
Tissue Distribution Study).
Qblood and
Qbrain were the amount of nanospheres
[µg/organ weight (g)] present in the bloodstream and in the brain,
respectively. Hence, the pharmacokinetic model represented in Fig.
1 could be described by the following
equations:
|
(1) |
|
(2) |
|
(3) |
|
|
(4) |
Statistical Analysis. Data were represented as means ± S.E. Sets of data were then compared with the nonparametric Mann-Whitney U test. Differences were considered significant at p < 0.05.
| |
Results |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Plasma Stability of PHDCA and PEG-PHDCA Nanospheres.
It is now
well established that cyanoacrylate nanoparticles are biodegradable.
Their degradation mechanism involves an enzymatic degradation with
hydrolysis of the ester function, leading to the formation of the
soluble cyanoacrylic acid and the corresponding alcohols (i.e.,
hexadecanol for PHDCA and PEG-PHDCA, as well as polyethylene glycol for
PEG-PHDCA) (Lenaerts et al., 1984). In addition, the biodegradation
rate of poly(cyanoacrylate) depends on the length of the alkyl chain
(Couvreur et al., 1979), with hexadecyl being more slowly degraded
compared with isohexyl and isobutyl. As shown in Fig.
2, representing the biodegradation profile of the poly(hexadecylcyanoacrylate) nanospheres in vitro (Fischer rat plasma), PEG-coated nanospheres were more rapidly degraded
when compared with noncoated nanospheres. Since PEG-coating inhibits
protein adsorption including enzymes, a rapid partial (7%)
solubilization rather than an enzymatic degradation was likely to occur
within the first 30 min. In our case, this solubilized fraction could
correspond to the more PEG-containing and hydrophilic oligomers present
in the PEG-PHDCA nanosphere dispersion (Brigger et al., 2000). After 30 min, the PEG-PHDCA nanosphere biodegradation rate (probably by an
enzymatic reaction) became quite similar to that of PHDCA nanospheres.
|
). Moreover, when the
solubilized nanospheres were injected intravenously, a very low level
of radioactivity was recovered in the plasma as soon as 1 min after
administration (Peracchia et al., 1999). Thus, the measured
radioactivity in vivo was likely to represent the biodistribution and
pharmacokinetics of the insoluble cyanoacrylate polymers only.
Tissue Distribution Study.
The biodistribution of
[14C]PHDCA and
[14C]PEG-PHDCA nanospheres in the blood, the
plasma, and the main organs of 9L tumor-bearing rats is presented in
Table 1. No significant difference in
biodistribution was observed between 9L tumor-bearing rats and control
animals (data not shown) for both types of nanospheres. Control PHDCA nanospheres had a very short plasma residence time, with a rapid and
massive uptake by the MPS, especially the liver and the spleen. On the
other hand, sterically stabilized PEG-PHDCA nanospheres displayed a
longer circulation time. Uptake of the pegylated nanoparticles by the
MPS also took place, but it was both delayed and less important than
for control nanospheres, except for the lungs. For the blood partition
of PHDCA and PEG-PHDCA nanospheres, it can be seen in Table 1 that all
of the 14C radioactivity was contained in the
plasma fraction, which suggested that none of the circulating
radioactivity was associated with blood cells.
|
|
), the
hydrophilic tracer [3H]sucrose is very slightly
permeant at the normal blood-brain barrier and is not metabolized when
administered intravenously to rats. Consequently, it was added to the
dispersion medium of both [14C]PHDCA and
[14C]PEG-PHDCA nanospheres to assert the state
of the blood-brain barrier alteration (Fig.
4). The biodistribution profiles of
sucrose in the main organs were, for all purposes except brain tumors, independent of the type of nanospheres dispersed
([14C]PHDCA or
[14C]PEG-PHDCA), as well as the group the rats
belonged to: tumor-bearing or control group (data not shown). In the
main organs, this compound featured a biodistribution profile which was
essentially reduced to the blood compartment and the kidney; high
concentrations were also found in the urine (data not shown). Plasma
concentrations dropped rapidly to 0.3%/g of organ after 30 min and to
0.01%/g of organ after 240 min, due to substantial elimination.
|
Pharmacokinetic Modelization and Tumor Uptake Rate.
Based on
the biodistribution study, the question remains whether the higher
concentration of PEG-coated nanospheres observed in all the brain
regions could be attributed to their long-circulating properties and/or
to an affinity with brain tissue and tumor. Therefore, to gain insight
into how the pegylated nanospheres accumulated more easily in the brain
when compared with control nanospheres, a simplified two-compartment
open pharmacokinetic model was developed (as described earlier and
represented in Fig. 1). Table 2
shows the experimental values of the rate constants k1,
k1', and ke, upon which our model was based.
First, kePHDCA and
kePEG-PHDCA, when calculated for the
tumor-bearing group (case A) or the control group (case B), were always
approximately 1,000 to 10,000 times higher than their respective
k1' (case A) or k1 (case B), indicating
that the rest of the body was predominantly responsible for nanosphere
blood clearance and excretion. Moreover, even if PEG-PHDCA nanospheres
were degraded and excreted faster, kePHDCA was about 2 to 7 times higher
than kePEG-PHDCA. The control nanospheres were captured more rapidly and massively by the MPS than
the PEG-coated nanospheres, resulting in a shorter circulation time.
Then, each type of nanospheres featured ke values that were approximately equivalent in both cases (A and B). This was not the case
for the k1PEG-PHDCA and
k1PHDCA values, which were 2- to
7-fold higher in a tumoral brain (k1', case A), when
compared with a normal brain (k1, case B). This result
agrees with the fact that the nanospheres more easily cross the
blood-brain barrier in the presence of a 9L gliosarcoma, by
extravasation across its altered and leaky blood-brain barrier.
Finally, the
k1PEG-PHDCA/k1PHDCA ratios were greater than 2 in both cases (k1' ratio for case
A and k1 ratio for case B), showing an affinity of the
pegylated nanospheres for the brain with a normal or an altered
blood-brain barrier.
|
6
min1 and 6.5*106
min1 for PHDCA and PEG-PHDCA nanospheres,
respectively. The fact that the mean rate constants k1'',
which express the extravasation of the carriers in the tumor
interstitium, were approximately identical suggested that the
intratumoral accumulation mechanism was the same for control or
pegylated nanospheres.
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
The successful transvascular delivery of optimal quantities of
blood-borne agents (particles or molecules) into tumors is critical to
the effectiveness of chemotherapy. Now, as a result of structural and
functional anomalies, the blood vessels in a tumor are generally leaky.
This can be taken advantage of when targeting tumors passively with
long-circulating particulate carriers (Gerlowski and Jain, 1986;
Yuan et al., 1995).
As expected, PHDCA nanospheres had a very short circulation time in
vivo, mainly due to a rapid and massive uptake by the MPS (Storm et
al., 1995). This result was also mirrored by mean rate constants
ke, which expressed elimination from the blood compartment
either by capture by organs or by urinary excretion, which comprised
between 9.95 × 102 and 3.13 × 101 min1. These mean
values were about 3 to 7 times higher than those of the sterically
stabilized PEG-PHDCA nanospheres. However, mean kePEG-PHDCA values were still 500 to
2000 times higher than the rate constants expressing brain accumulation
(k1, normal brain and k1', tumoral brain), which
demonstrated that, although biodegradable pegylated nanospheres are
engineered to evade the MPS, there was still a significant amount of
clearance via this route. This is the common drawback for brain tumor
targeting with intravascularly administered compounds: the body acts as
a sink, and only a small fraction of administered compound actually
reaches the tumor (Groothuis, 2000). However, the longer
circulation time of the PEG-PHDCA nanospheres, as compared with PHDCA
ones, allowed us to enhance importantly the fraction of carrier
redirected to non-MPS sites, for instance, brain tumors.
The delivery of carriers to brain tumors may be affected not only by
their blood clearance by MPS or by excretion, but also by physiological
barriers limiting tumor accumulation (Yuan et al., 1994; Jain, 1998,
2001).
Nevertheless, PHDCA and PEG-PHDCA nanospheres were able to selectively
extravasate across the blood-brain barrier in the 9L gliosarcoma
interstitium. However, this effect was significantly increased for the
pegylated nanospheres. Their maximal tumor-to-brain ratio, measured
4 h after intravenous injection, was found to be 11. This ratio
was quite consistent with the results obtained 2 h after
intravenous injection of radiolabeled Caelyx to patients, where the
carrier accumulation was 13 to 19 times higher in the glioblastoma and
7 to 13 times higher in the metastatic lesions compared with normal
brain (Koukourakis et al., 2000). In the same way, Vescan liposomes,
characterized by an improved plasma stability, showed at best a
tumor-to-brain ratio of 7.5 in patients (Khalifa et al., 1997). For a
final comparison, tumor uptake of radiolabeled long-circulating
dextran-coated iron oxide nanoparticles exceeded the uptake in any of
the normal brain tissues by a factor of 10, 24 h after their
intravenous injection (Moore et al., 2000).
The PHDCA or PEG-PHDCA nanosphere accumulation in the 9L gliosarcoma
correlated well with the alteration of the blood-brain barrier at the
tumor site, as unequivocally evidenced by the measurement of the
hydrophilic tracer sucrose penetration. Like many solid tumors
(Gerlowski and Jain, 1986; Yuan et al., 1995), 7 days after intracerebral engraftment, 9L tumor vessels were thus hyperpermeable, when compared with normal blood vessels, even to blood-borne particles with a diameter approximating 200 nm. This value was 2 times higher than the pore cutoff size reported for the U87 human glioblastoma implanted intracerebrally in mice (7-100 nm) (Hobbs at al., 1998).
As far as the mechanism for tumor accumulation in the 9L gliosarcoma
was concerned, our simplified pharmacokinetic model allowed the
calculation of k1'' for the PHDCA and PEG-PHDCA nanospheres, which is the constant rate governing the passage from the bloodstream to the malignant tissue. Because the obtained k1'' values
were approximately the same for both types of nanospheres (PHDCA
nanospheres, 5 × 106
min1 and PEG-PHDCA nanospheres, 6.5 × 106 min1), this
suggested that the intratumoral accumulation mechanism was the same,
probably by diffusion/convection across the altered and hyperpermeable
tumoral endothelium (Yuan, 1998). Thus, the higher accumulation of the
pegylated nanospheres in the gliosarcoma was mostly due to their longer
blood circulation time, which gave them more chances to extravasate in
the interstitium.
Finally, both carrier systems displayed a retention effect in the
glioma, even if their plasma concentration was lower than that in the
tumor. On the contrary, such retention was not observed for the
hydrophilic tracer sucrose, which diffused out of the tumor during the
time of analysis (4 h). Under physiological conditions, components of
the interstitial fluid can either reenter the bloodstream directly via
the postcapillary venules, or they can enter the lymphatic system,
depending on their size. Smaller molecules are better suited to reenter
the veins, whereas larger molecules tend to enter the lymphatic system
(Seymour, 1992). Hence, the observed extravasation and retention of the
nanospheres within the 9L tumor interstitium accounted for the
leakiness and the coexisting lack of a lymphatic system in the brain.
This phenomenon was described as the EPR effect: enhanced permeability
and retention effect (Maeda, 2001). Conversely, because sucrose
displayed a small molecular size as well as a rapid elimination by
renal clearance, this molecule could diffuse out of the tumor in the
bloodstream, following gradient concentration equilibrium.
An interesting result was the 4- to 8-fold higher concentration of the more hydrophilic PEG-coated nanospheres observed in normal brain regions (contralateral hemisphere in the tumor-bearing and control groups: stereotactic injection site and surrounding tissue in the control group), when compared with the PHDCA nanospheres. This higher accumulation could neither be explained by residual radioactivity in blood due to their longer circulation time, since a transcardiac perfusion with saline was systematically performed, nor by a specific blood-brain barrier permeabilization due to a toxic surfactant effect of the amphiphilic PEG-coated carriers. Indeed, in the last case, an identical and negligible brain sucrose penetration was observed for both types of nanospheres, indicating the same state of the blood-brain barrier.
Our simplified pharmacokinetic model supposed that the concentration of
the nanospheres in the brain depended on their blood clearance,
especially by MPS organs, represented by ke, and on their
absorption in the brain, expressed by k1
[Qbrain = Q0/(1 + ke/k1)]. This points out that the accumulation
in the brain was improved when using a long-circulating carrier with
reduced MPS clearance. Besides, PEG-coated carriers displayed a 2.5 times higher mean k1 than did PHDCA nanospheres, suggesting
that PEG-coated carriers had a stronger affinity for the blood-brain
barrier than did their nonpegylated counterparts. A specific
interaction of the PEG or Pluronic coating with the blood-brain barrier
endothelial cells could explain this result. The mean concentration of
the PEG-PHDCA carriers reached 0.027% injected dose/g of brain tissue at 30 min (Fig. 3D), which was quite in line with the accumulation obtained for long-circulating OX26 monoclonal antibody immunoliposomes directed to the rat transferrin receptor, which featured an average brain delivery of 0.030% injected dose/g of tissue 1 h after
their intravenous injection (Huwyler et al., 1996).
However, the exact uptake mechanism across the normal blood-brain
barrier could not be explained for the pegylated nanospheres. Among
possible mechanisms of transport to the brain are transcytosis or
inhibition of an efflux system, especially P-glycoprotein (Batrakova et
al., 1999; Kreuter, 2001). Another hypothesis could be the adsorption
of apolipoprotein E on the PEG-coated nanospheres. This overcoating
with apolipoprotein E could mimic low-density lipoprotein particles,
leading to an uptake of the nanospheres by an endocytic process, after
interaction with the low-density lipoprotein receptor (Kreuter, 2001).
Finally, adhesion of the nanospheres to the vessel walls of the brain,
rather than a real uptake across the blood-brain barrier, is also
possible. This point deserves further investigations, but it seems that
the density of the hydrophilic coating has an importance for brain
targeting: it should be a compromise between the long-circulating
properties of the carrier and appropriate surface characteristics to
still permit interaction with brain endothelial cells and macrophages (Hong et al., 1999; Calvo et al., 2001). For example, the same PEG-coated liposomes as previously reported, but without the grafted monoclonal antibody OX26, displayed a greatly increased plasma half-life, when compared with conventional liposomes, OX26
PEG-immunoliposomes, and also PEG-PHDCA nanospheres; however, the brain
uptake of these pegylated liposomes was zero (Huwyler et al., 1996).
In conclusion, although PHDCA and PEG-PHDCA nanospheres were
cleared by the MPS organs, their accumulation, probably by
diffusion/convection across the altered and leaky blood-brain barrier,
and retention in the 9L brain tumor was found to be possible. In
addition, due to their higher concentration in the malignant tissue and
their lower uptake by MPS organs, the sterically stabilized PEG-coated nanospheres showed a biodistribution profile that was better suited for
brain tumor targeting than that of the control PHDCA nanospheres. Finally, contrary to the long-circulating pegylated liposomes from
Huwyler et al. (1996), our sterically stabilized PEG-PHDCA nanospheres
displayed an affinity for brain regions protected by a normal
blood-brain barrier. This can be of benefit for the treatment of
infiltrating brain tumors.
| |
Acknowledgments |
|---|
We thank Patrice Ardouin and the staff of the animal unit at the Institut Gustave-Roussy for their care of the animals. The revision of the manuscript by Eric Guilbert was also greatly appreciated.
| |
Footnotes |
|---|
Accepted for publication August 9, 2002.
Received for publication May 30, 2002.
This work was supported by Grant ARC 5837 from the Association pour la Recherche sur le Cancer, France.
DOI: 10.1124/jpet.102.039669
Address correspondence to: Prof. Patrick Couvreur, UMR CNRS 8612, Faculty of Pharmacy, 5, rue J.B. Clément, 92296 Châtenay-Malabry, France. E-mail: patrick.couvreur{at}cep.u-psud.fr
| |
Abbreviations |
|---|
PEG, poly(ethylene glycol); MPS, mononuclear phagocyte system; PHDCA, poly(hexadecylcyanoacrylate); PEG-PHDCA, poly(MePEG2000cyanoacrylate-co-hexadecylcyanoacrylate).
| |
References |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  H. S. SHARMA, S. F. ALI, W. DONG, Z. R. TIAN, R. PATNAIK, S. PATNAIK, A. SHARMA, A. BOMAN, P. LEK, E. SEIFERT, et al. Drug Delivery to the Spinal Cord Tagged with Nanowire Enhances Neuroprotective Efficacy and Functional Recovery following Trauma to the Rat Spinal Cord Ann. N.Y. Acad. Sci., December 1, 2007; 1122(1): 197 - 218. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
(p < 0.05), and statistical
differences between the tumor and control group are expressed by 
(p < 0.05) (nonparametric Mann-Whitney
U test).
