J Pharm Pharmaceut Sci (www.ualberta.ca/~csps) 6(2):252-273, 2003
`
`Drug delivery to the central nervous system: a review.
`
`Ambikanandan Misra, Ganesh S., Aliasgar Shahiwala
`Pharmacy Department, M.S. University of Baroda
`
`Shrenik P. Shah
`Sun Pharma Advanced Research Center, Baroda
`
`Received 16 June 2003, Revised 26 June 2003, Accepted 5 August 2003
`
`Abstract The brain is a delicate organ, and evolution built
`very efficient ways to protect it. Unfortunately, the same
`mechanisms that protect it against intrusive chemicals can
`also frustrate therapeutic interventions. Many existing
`pharmaceuticals are rendered ineffective in the treatment
`of cerebral diseases due to our inability to effectively
`deliver and sustain them within the brain. General meth-
`ods that can enhance drug delivery to the brain are, there-
`fore, of great interest. Despite aggressive research, patients
`suffering from fatal and/or debilitating central nervous
`system (CNS) diseases, such as brain tumors, HIV enceph-
`alopathy, epilepsy, cerebrovascular diseases and neurode-
`generative disorders, far outnumber those dying of all
`types of systemic cancer or heart disease. The clinical fail-
`ure of much potentially effective therapeutics is often not
`due to a lack of drug potency but rather to shortcomings
`in the method by which the drug is delivered. Treating
`CNS diseases is particularly challenging because a variety
`of formidable obstacles often impede drug delivery to the
`brain and spinal cord. By localizing drugs at their desired
`site of action one can reduce toxicity and increase treat-
`ment efficiency. In response to the insufficiency in con-
`ventional delivery mechanisms, aggressive research efforts
`have recently focused on the development of new strate-
`gies to more effectively deliver drug molecules to the CNS.
`This review intends to detail the recent advances in the
`field of brain-targeting, rational drug design approach and
`drug delivery to CNS. To illustrate the complexity of the
`problems that have to be overcome for successful brain
`targeting, a brief intercellular characterization of the
`blood–brain barrier (BBB) is also included.
`
`INTRODUCTION
`Despite enormous advances in brain research, brain and
`
`Corresponding Author: Ambikanandan Misra, Pharmacy Depart-
`ment, Faculty of Technology & Engineering, M.S.University of Baroda,
`Kalabhavan, Vadodara – 390001. Gujarat. misraan@satyam.net.in,
`misraan@hotmail.com
`
`central nervous system disorders remain the world's lead-
`ing cause of disability, and account for more hospitaliza-
`tions and prolonged care than almost all other diseases
`combined. The major problem in drug delivery to brain is
`the presence of the BBB. Drugs that are effective against
`diseases in the CNS and reach the brain via the blood
`compartment must pass the BBB. In order to develop
`drugs which penetrate the BBB well to exhibit the
`expected CNS therapeutic effects, it is of great importance
`to understand the mechanisms involved in uptake into and
`efflux from the brain. The function of the BBB is dynami-
`cally regulated by various cells present at the level of the
`BBB (1). This realization implies better understanding of
`the relationship of transport at the BBB to drug structure
`and physicochemical properties.
`
`Despite successful examples of drug delivery to the CNS,
`but only some have reached the phase where they can pro-
`vide safe and effective human applications. As pharmaco-
`logical strategies improve, there will be less need for
`invasive procedures for treating CNS diseases. Consider-
`able strides have been made in intravascular delivery and
`neurosurgical invasive procedures to deliver therapeutic
`substances into the brain.
`
`This review will prove invaluable to researchers interested
`in the fundamental function of the BBB and those in the
`pharmaceutical industry interested in rational drug design
`directed at delivering drugs to the brain.
`
`BARRIERS TO CNS DRUG DELIVERY
`The failure of systemically delivered drugs to effectively
`treat many CNS diseases can be rationalized by consider-
`ing a number of barriers that inhibit drug delivery to the
`CNS.
`
`Blood-Brain Barrier
`It is now well established that the BBB is a unique mem-
`branous barrier that tightly segregates the brain from the
`
`252
`
`Auspex Exhibit 2020
`Apotex v. Auspex
`IPR2021-01507
`Page 1
`
`

`

`J Pharm Pharmaceut Sci (www.ualberta.ca/~csps) 6(2):252-273, 2003
`
`circulating blood (2, 3). The CNS consist blood capillaries
`which are structurally different from the blood capillaries
`in other tissues; these structural differences result in a per-
`meability barrier between the blood within brain capillaries
`and the extracellular fluid in brain tissue. Capillaries of the
`vertebrate brain and spinal cord lack the small pores that
`allow rapid movement of solutes from circulation into
`other organs; these capillaries are lined with a layer of spe-
`cial endothelial cells that lack fenestrations and are sealed
`with tight junctions. Tight epithelium, similar in nature to
`this barrier, is also found in other organs (skin, bladder,
`colon, and lung) (4).This permeability barrier, comprising,
`the brain capillary endothelium, is known as the BBB.
`Ependymal cells lining the cerebral ventricles and glial cells
`are of three types. Astrocytes form the structural frame
`work for the neurons and control their biochemical envi-
`ronment. Astrocytes foot processes or limbs that spread
`out and abutting one other, encapsulate the capillaries are
`closely associated with the blood vessels to form the BBB.
`Oligodendrocytes are responsible for the formation and
`maintenance of the myelin sheath, which surrounds axons
`and is essential for the fast transmission of action poten-
`tials by salutatory conduction. Microglias are blood derived
`mononuclear macrophages. The tight junctions between
`endothelial cells results in a very high trans-endothelial
`electrical resistance of 1500-2000 Ω.cm2 compared to 3-33
`Ω.cm2 of other tissues which reduces the aqueous based
`para-cellular diffusion that is observed in other organs (5,
`6).
`
`Micro-vessels make up an estimated 95% of the total sur-
`face area of the BBB, and represent the principal route by
`which chemicals enter the brain. Vessels in brain were
`found to have somewhat smaller diameter and thinner wall
`than vessels in other organs. Also, the mitochondrial den-
`sity in brain micro-vessels was found to be higher than in
`other capillaries not because of more numerous or larger
`mitochondria, but because of the small dimensions of the
`brain micro-vessels and consequently, smaller cytoplasmic
`area. In brain capillaries, intercellular cleft, pinocytosis, and
`fenestrae are virtually nonexistent; exchange must pass
`trans-cellularly. Therefore, only lipid-soluble solutes that
`can freely diffuse through the capillary endothelial mem-
`brane may passively cross the BBB. In capillaries of other
`parts of the body, such exchange is overshadowed by other
`nonspecific exchanges. Despite the estimated total length
`of 650km and total surface area of 12 m2 of capillaries in
`human brain, this barrier is very efficient and makes the
`brain practically inaccessible for lipid- insoluble com-
`
`pounds such as polar molecules and small ions. As a con-
`sequence, the therapeutic value of many promising drugs
`is diminished, and cerebral diseases have proved to be
`most refractory to therapeutic interventions. Given the
`prevalence of brain diseases alone, this is a considerable
`problem. Practically all drugs currently used for disorders
`of the brain are lipid-soluble and can readily cross the BBB
`following oral administration. Although antimicrobial b-
`lactam antibiotics, when administered intracerebroventric-
`ularly, cause severe convulsion, fortunately these antibiot-
`ics, when administered intravenously or orally, do not
`cause such central nervous system (CNS) side effect
`because their limited transport across the blood–brain bar-
`rier (BBB). Further, in spite of being well distributed into
`various tissues, a lipophilic new quinolone antimicrobial
`agent, grepafloxacin, cannot enter the brain, resulting in
`the avoidance of CNS side effects such as headache and
`dizziness due to the displacement of g-aminobutyric acid
`(GABA) from the GABA receptor binding sites. On the
`other hand, benzodiazepines such as diazepam have been
`used as sedative-hypnotic agents, because these lipophilic
`drugs readily cross the BBB. However, the BBB transport
`of an immunosuppressive agent, cyclosporin A, which is
`more lipophilic than diazepam, is highly restricted. Simi-
`larly, almost all of the lipophilic anticancer agents such as
`doxorubicin, epipodophylotoxin and Vinca alkaloids (e.g.,
`vincristine and vinblastine) hardly enter the brain, causing
`difficulty in the treatment of brain tumors. Although
`levodopa, which is useful for treatment of Parkinson’s dis-
`ease, is very hydrophilic, it can readily penetrate the BBB.
`What mechanisms underlie these diverse BBB transport
`characteristics of drugs which are apparently structurally
`and pharmacologically unrelated? In order to avoid over-
`lap with this section, the drug transport across the BBB of
`small-molecular drugs by carrier-mediated transport and
`of peptide drugs by the adsorptive-mediated transcytosis
`are discussed in section 7.1.4 and 7.1.5 respectively.
`
`Some regions of the CNS do not express the classical BBB
`capillary endothelial cells, but have micro-vessels similar to
`those of the periphery. These areas are adjacent to the ven-
`tricles of the brain and are termed the circumventricular
`organs (CVOs). The CVOs include the choroid plexus, the
`median
`eminence, neurohypophysis, pineal gland,
`organum vasculosum of the lamina terminalis, subfornical
`organ, subcommisaral organ and the area postrema.
`Though in the CVO brain regions the capillaries are more
`permeable to solutes, the epithelial cells of the choroid
`plexus and the tanycytes of other regions form tight junc-
`
`253
`
`Auspex Exhibit 2020
`Apotex v. Auspex
`IPR2021-01507
`Page 2
`
`

`

`J Pharm Pharmaceut Sci (www.ualberta.ca/~csps) 6(2):252-273, 2003
`
`tions to prevent transport from the abluminal extracellular
`fluid (ECF) to the brain ECF. The choroid plexus may be
`of importance when considering the transport of peptide
`drugs, because it is the major site of cerebrospinal-fluid
`(CSF) production, and both the CSF and brain ECF freely
`exchange (7).
`
`The BBB also has an additional enzymatic aspect. Solutes
`crossing the cell membrane are subsequently exposed to
`degrading enzymes present in large numbers inside the
`endothelial cells that contain large densities of mitochon-
`dria, metabolically highly active organelles. BBB enzymes
`also recognize and rapidly degrade most peptides, includ-
`ing naturally occurring neuropeptides (8, 9).
`
`Finally, the BBB is further reinforced by a high concentra-
`tion of P-glycoprotein (Pgp), active –drug-efflux-trans-
`porter protein in the luminal membranes of the cerebral
`capillary endothelium. This efflux transporter actively
`removes a broad range of drug molecules from the endot-
`helial cell cytoplasm before they cross into the brain paren-
`chyma. Figure-1 gives a schematic representation of all
`these BBB properties using a comparison between brain
`and general capillaries.
`
`Figure 1: Schematic comparison between general (left)
`and brain (right) capillaries.
`
`Blood-Cerebrospinal Fluid Barrier
`The second barrier that a systemically administered drug
`encounters before entering the CNS is known as the
`blood-cerebrospinal fluid barrier (BCB). Since the CSF
`can exchange molecules with the interstitial fluid of the
`brain parenchyma, the passage of blood-borne molecules
`into the CSF is also carefully regulated by the BCB. Physio-
`logically, the BCB is found in the epithelium of the chor-
`oids plexus, which are arranged in a manner that limits the
`passage of molecules and cells into the CSF. The choroid
`
`254
`
`plexus and the arachnoid membrane act together at the
`barriers between the blood and CSF. On the external sur-
`face of the brain the ependymal cells fold over onto them-
`selves to form a double layered structure, which lies
`between the dura and pia, this is called the arachnoid
`membrane. Within the double layer is the subarachnoid
`space, which participates in CSF drainage. Passage of sub-
`stances from the blood through the arachnoid membrane
`is prevented by tight junctions (10). The arachnoid mem-
`brane is generally impermeable to hydrophilic substances,
`and its role is forming the Blood-CSF barrier is largely pas-
`sive. The choroid plexus forms the CSF and actively regu-
`lates the concentration of molecules in the CSF. The
`choroid plexus consist of highly vascularized, "cauliflower-
`like" masses of pia mater tissue that dip into pockets
`formed by ependymal cells. The preponderance of choroid
`plexus is distributed throughout the fourth ventricle near
`the base of the brain and in the lateral ventricles inside the
`right and left cerebral hemispheres. The cells of the chor-
`oidal epithelium are modified and have epithelial charac-
`teristics. These ependymal cells have microvilli on the CSF
`side, basolateral interdigitations, and abundant mitochon-
`dria. The ependymal cells, which line the ventricles, form a
`continuous sheet around the choroid plexus. While the
`capillaries of the choroid plexus are fenestrated, non-con-
`tinuous and have gaps between the capillary endothelial
`cells allowing the free-movement of small molecules, the
`adjacent choroidal epithelial cells form tight junctions pre-
`venting most macromolecules from effectively passing
`into the CSF from the blood (11). However, these epithe-
`lial-like cells have shown a low resistance as compared the
`cerebral endothelial cells, approximately 200 Ω.cm2,
`between blood and CSF (12).
`
`In addition, the BCB is fortified by an active organic acid
`transporter system in the choroids plexus capable of driv-
`ing CSF-borne organic acids into the blood. As a result a
`variety of therapeutic organic acids such as the antibiotic
`penicillin, the anti-neoplastic agent methotrexate, and the
`antiviral agent zidovudine are actively removed from the
`CSF and therefore inhibited from diffusing into the brain
`parenchyma. Furthermore, substantial
`inconsistencies
`often exist between the composition of the CSF and inter-
`stitial fluid of the brain parenchyma, suggesting the pres-
`ence of what is sometimes called the CSF-brain barrier
`(13). This barrier is attributed to the insurmountable diffu-
`sion distances required for equilibration between the CSF
`and the brain interstitial fluid. Therefore, entry into the
`CSF does not guarantee a drug’s penetration into the brain.
`
`Auspex Exhibit 2020
`Apotex v. Auspex
`IPR2021-01507
`Page 3
`
`

`

`J Pharm Pharmaceut Sci (www.ualberta.ca/~csps) 6(2):252-273, 2003
`
`Blood-Tumor Barrier
`Intracranial drug delivery is even more challenging when
`the target is a CNS tumor. The presence of the BBB in the
`microvasculature of CNS tumors has clinical conse-
`quences. For example, even when primary and secondary
`systemic tumors respond to chemotherapeutic agents
`delivered via
`the cardiovascular system,
`intracranial
`metastases often continue to grow. In CNS malignancies
`where the BBB is significantly compromised, a variety of
`physiological barriers common to all solid tumors inhibit
`drug delivery via the cardiovascular system. Drug delivery
`to neoplastic cells in a solid tumor is compromised by a
`heterogeneous distribution of microvasculature through-
`out the tumor interstitial, which leads to spatially inconsis-
`tent drug delivery. Furthermore, as a tumor grows large,
`the vascular surface area decreases, leading to a reduction
`in trans-vascular exchange of blood-borne molecules. At
`the same time, intra-capillary distance increases, leading to
`a greater diffusional requirement for drug delivery to neo-
`plastic cells and due to high interstitial tumor pressure and
`the associated peri-tumoral edema leads to increase in
`hydrostatic pressure in the normal brain parenchyma adja-
`cent to the tumor. As a result, the cerebral microvascula-
`ture in these tumor adjacent regions of normal brain may
`be even less permeable to drugs than normal brain endot-
`helium, leading to exceptionally low extra-tumoral intersti-
`tial drug concentrations (14). Brain tumors may also
`disrupt BBB, but these are also local and nonhomoge-
`neous disruptions (15).
`
`In conclusion, the delivery of drugs to the CNS via the
`cardiovascular system is often precluded by a variety of
`formidable barriers including the BBB, the BCB, and the
`BTB.
`
`EFFLUX MECHANISMS IN DRUG TRANSPORT TO THE
`BRAIN
`A detailed understanding of the uptake and efflux mecha-
`nisms at the BBB would be very helpful for targeting drugs
`to the brain to provide the expected CNS pharmacological
`effect or for the reduction of BBB penetration of drugs in
`order to minimize side effects in the CNS. Most in-vivo
`experimental methods describing drug uptake into brain
`will automatically incorporate any activity of CNS efflux
`into their apparent determination of brain penetration.
`Within the CNS are a number of efflux mechanisms that
`will influence drug concentrations in the brain. Some of
`these mechanisms are passive while others are active.
`
`Active efflux from the CNS via specific transporters may
`often reduce the measured penetration of drug at the BBB
`to levels that are lower than might be predicted from the
`physicochemical properties of the drug, for example, its
`lipid solubility. The activity of these efflux mechanisms
`influence the concentration in brain extracellular fluid of
`free drugs that are available to interact with drug receptor
`sites. Recently much attention has been focused on the so-
`called multi-drug transporters; multi-drug resistance pro-
`tein (MRP), P-glycoprotein (Pgp) and the multi-specific
`organic anion transporter (MOAT), which belong to the
`members of the ABC cassette (ATP-binding cassette) of
`transport protein (16, 17). The MRP in humans appears to
`be five isoforms, and there are different levels of expres-
`sion of these various isoforms in different tissues. Pgp is
`the product of the multidrug resistance (MDR) gene in
`humans and accepts a wide range of lipid-soluble sub-
`strates and will actively efflux these from cells expressing
`the gene product. The MOAT in the choroid plexus shows
`some similarity in its substrate preferences with MRP.
`Noticeably, brain exposure can be increased not only by
`enhancing influx, but by restricting efflux through the
`BBB as well. Hence, strategies directed at increasing brain
`uptake of drugs that are substrates for specific efflux
`mechanisms need to be focused on designing reactivity
`with a transporter out of a drug molecule or by examining
`ways of inhibiting the activity of an efflux mechanism by
`co-administering a competitive or noncompetitive inhibi-
`tor of the efflux pump together with the desired drug. For
`example, for certain Pgp substrates, coadministeration of a
`Pgp inhibitor can increase not only oral absorption, but
`also BBB permeability (18, 19). Coadministration of the
`Pgp blocker valspodar has recently been shown to not
`only increase the brain levels pf paclitaxel, but also to con-
`siderably improve its therapeutic effect on tumor volume
`in mice (20). On the contrary, among the brain drug deliv-
`ery strategies to be discussed later, chemical drug delivery
`systems (CDDS) are the only ones attempting to not only
`increase influx, but also to decrease efflux. This strategy is
`done by exploiting a sequential metabolic approach that
`first
`increases
`influx by passive diffusion
`through
`increased lipophilicity and then decreases efflux by a ‘lock-
`in’ mechanism.
`
`PHYSICOCHEMICAL FACTORS THAT INFLUENCE BRAIN
`UPTAKE
`Brain penetration, brain uptake, and ability to cross the
`BBB need to be defined exactly to understand concepts
`
`255
`
`Auspex Exhibit 2020
`Apotex v. Auspex
`IPR2021-01507
`Page 4
`
`

`

`J Pharm Pharmaceut Sci (www.ualberta.ca/~csps) 6(2):252-273, 2003
`
`involved in brain uptake. Hence, the various ways in which
`transfer across the BBB are defined in table-1.
`
`uptake index (BUI) is defined in equation-1:
`
`Table 1: Measures of “Brain Uptake”.
`
`where the BUI for water is 100. Although, the BUI is use-
`ful as a rank order index of brain uptake, is not easily ame-
`nable to analysis by physicochemical methods.
`
`A more well-defined measure of rapid brain uptake is the
`permeability, expressed either as a permeability-surface
`area product (PS) or as a permeability coefficient (PC),
`obtained by intravenous injection and measurement of the
`drug profile in arterial blood. Both the PS product and PC
`are quantitative measures of the rate of transport obtained
`by in-situ vascular perfusion technique (27) and so are
`amenable to analysis through standard physicochemical
`procedures. An advantage of the perfusion technique as a
`measure of brain uptake is that the time scale for determi-
`nation of PS products is very short, so that back transport
`and biological degradation are minimized. Although there
`are numerous physicochemical studies on brain perfusion,
`it is not possible to reach any general conclusions.
`
`Following systemic drug administration, uptake from the
`circulation into parenchyma by a specific organ of interest
`will be determined by the following factors: (a) blood flow
`to the organ, (b) permeability of the micro-vascular wall,
`and (c) the amount of drug available for uptake, which is
`inversely related to systemic clearance and is represented
`by the area under the plasma concentration-time curve
`(AUC). For the quantification of brain tissue accumulation
`(Cbrain) at time T during the phase of unidirectional
`uptake, the following equation-2 holds:
`
`where PS is the brain capillary permeability surface area
`product, an expression equivalent to the organ clearance
`and AUC is the area under the plasma concentration time
`curve. It should be mentioned that this equation does not
`take into account efflux of either intact drug or metabo-
`lism and efflux of degradation products from the brain.
`Measurement of efflux is covered in section 6 of this
`review.
`
`Based on the relationship between the octanol / water par-
`tition coefficient (PC) divided by the square root of the
`molecular weight (PC/ Mw1/2) and the BBB permeability
`coefficient (PS), one can classify at least three different
`groups: (a) substrates exhibiting a good correlation, (b)
`
`Biological activity is a general measure of brain uptake.
`The hypnotic activity of a number of congeneric series of
`CNS depressants reached a maximum when log octanol–
`water partition coefficient (log Po/w) was near to 2. Vari-
`ous researchers confirmed this finding and the “rule of 2”
`became generally accepted (21). But the difficulty here is
`that the biological activity will depend on at least two fac-
`tors:
`
`•
`
`rate of transfer from blood to brain, or distribution between
`blood and brain; and
`interaction between drug and some receptors in the brain.
`•
`If these two factors cannot be distinguished, then it is
`impossible to use biological activity as a measure of either
`rate or equilibrium transfer.
`
`The log Po/w probably still represents the most informative
`physicochemical parameter used in medicinal chemistry
`and countless examples where it proved as useful descrip-
`tors are available in the literature (22). On the other hand,
`increasing lipophilicity with the intent to improve mem-
`brane permeability might not only make chemical handling
`difficult, but also increase the volume of distribution in
`particular plasma protein binding and tends to affect all
`other pharmacokinetic parameters (23, 24). Furthermore,
`increasing lipophilicity tends to increase the rate of oxida-
`tive metabolism by cytochromes P450 and other enzymes
`(23, 25). Hence, to improve bioavailability, the effects of
`lipophilicity on membrane permeability and first pass
`metabolism have to be balanced.
`
`The brain uptake index (26) is a more rigorous measure of
`brain uptake in which there is a relative measure of brain
`uptake by intra-carotid injection of a mixture of 14C-
`labeled compound and 3H-labeled water (i.e. a saline solu-
`tion in 3H-labeled water). The radioactivity in brain tissue
`is recorded 15 seconds after administration, and a brain
`
`256
`
`Auspex Exhibit 2020
`Apotex v. Auspex
`IPR2021-01507
`Page 5
`
`

`

`J Pharm Pharmaceut Sci (www.ualberta.ca/~csps) 6(2):252-273, 2003
`
`substrates exhibiting a significantly greater PS value than
`indicated by their lipophilicity, and (c) substrates exhibiting
`a significantly smaller PS value than indicated by their lipo-
`philicity. The transport mechanism for groups (a) and (b)
`is passive diffusion and facilitated transport, respectively
`(27). The molecular weight of the compounds in group (c)
`is greater than 400 Da., the absolute cut-off for significant
`BBB passage regardless of lipophilicity. This molecular
`weight threshold hypothesis was proposed to explain the
`mechanism operating in the case of group (c) (28).
`
`Brain uptake can be positively correlated with lipid solubil-
`ity or negatively correlated with hydrogen bonding (29).
`The extent to which a compound forms hydrogen bonds
`is vital for its ability to permeate endothelial cell mem-
`branes. The higher the hydrogen bonding potential, lower
`the uptake into the brain. By reducing the hydrogen bond-
`ing potential for a congeneric series of steroid hormones,
`there was a log increase in uptake with each removal of
`hydrogen bond pairs. The correlation of blood-brain dis-
`tribution coefficients (as log BB in-vivo and in-vitro values)
`using hydrogen bonding descriptors are available (30) but
`are not very similar to correlations for log PS. Hence the
`factors that influence blood-brain distribution are not
`quantitatively the same as those that influence brain perfu-
`sion. So it is vitally important when discussing brain
`uptake to specify what measure of brain uptake is being
`used. A variety of in silico models (31) and in vitro perme-
`ability assays (32) have been developed in an attempt to
`characterize and predict BBB permeability and integrate
`such prediction in the early phases of drug development,
`together with various other considerations (33-35).
`
`IN VIVO AND IN VITRO MODELS TO STUDY DRUG
`TRANSPORT ACROSS THE BLOOD-BRAIN AND BLOOD-
`CSF BARRIERS
`The pharmacokinetics and pharmacodynamics of drugs in
`the CNS are understood by their unbound concentrations
`in the extracellular fluid of the brain. Various in-vivo and
`in-vitro techniques are available to study this property. The
`in-vivo techniques include the brain uptake index (BUI)
`(26), the brain efflux index (BEI) (36), brain perfusion
`(37), the unit impulse response method (38) and micro-
`dialysis (39).
`
`The efflux transport across the BBB is a very important
`process for explaining the mechanism of the apparent
`restricted cerebral distribution of drugs after their systemic
`administration. In order to examine the BBB efflux trans-
`
`port mechanism under in-vivo conditions, the intracere-
`bral microinjection technique has been developed and
`recently established as the BEI. The BEI value is defined
`as the relative percentage of drug effluxed from the ipsilat-
`eral (that is, they do not cross to the opposite hemisphere)
`cerebrum to the circulating blood across the BBB com-
`pared with the amount of drug injected into the cerebrum,
`i.e.:
`
`The advantages of the BEI method are its ability to allow
`determination of the apparent in vivo drug efflux rate con-
`stant across the BBB, monitoring the concentration
`dependency of the test drug and the performance of inhi-
`bition studies. By contrast, the limitations of the BEI
`method are that only one data point can be obtained for a
`single intracerebral microinjection. The drug concentra-
`tion in the cerebrum cannot be accurately determined. In
`other words, at the present time, the drug concentration in
`the brain is estimated by using the dilution factor, i.e. 30.3-
`to 46.2-fold dilution (36).
`
`The brain interstitial fluid (ISF) concentration is a determi-
`nant for the effect of a drug in the CNS in-vivo. If the drug
`would cross the BBB in significant quantities by passive
`diffusion, the brain ISF concentration will equal the
`plasma unbound drug concentration after its administra-
`tion. In this case, the plasma unbound drug concentration
`will be very important in predicting the CNS effect. How-
`ever, if the brain ISF concentration of a drug is signifi-
`cantly lower than the plasma unbound drug concentration,
`it will be very important to identify the mechanism
`involved. For the direct measurement of brain ISF drug
`concentration, many researchers have found brain micro-
`dialysis to be a useful technique (40, 41). Micro-dialysis is a
`method of choice in the study of in-vivo drug transport
`across the BBB, based on brain’s physiological and ana-
`tomical characteristics considering it to be a non-homoge-
`neous compartment. In addition, drug disposition in the
`brain is determined by protein binding, blood flow, BBB
`transport, and the exchange between brain extracellular
`fluid (ECF) and brain cells. Nevertheless, intra-cerebral
`micro-dialysis is an invasive technique: it involves the
`implantation of a probe, which may cause tissue trauma,
`and hence may have consequences for BBB function.
`Therefore it is necessary to determine whether intra-cere-
`bral micro-dialysis provides meaningful data on drug
`transport across the BBB and drug disposition in the
`
`257
`
`Auspex Exhibit 2020
`Apotex v. Auspex
`IPR2021-01507
`Page 6
`
`

`

`J Pharm Pharmaceut Sci (www.ualberta.ca/~csps) 6(2):252-273, 2003
`
`brain.
`
`Since thousands of new therapeutic compounds will have
`to be tested in the near future; alternatives to in-vivo test
`systems must be developed. Thus, in-vitro models that
`closely mimic the in-vivo system, at least with respect to
`barrier properties, are in high demand. Blood-brain barrier
`models now available make use of cerebral capillary endot-
`helium (porcine brain capillary endothelial cells) or chor-
`oid plexus epithelial cells (porcine choroid plexus) (42, 43).
`Both cell types need serum in the growth medium to pro-
`liferate. Serum, however, inhibits the formation of tight
`cell-cell contacts. Withdrawal of serum favors cellular
`polarity and increases the barrier properties drastically.
`Electrical resistance is an easy measure of junctional tight-
`ness (44). A very sophisticated but highly reliable and
`reproducible new method is impedance spectroscopy (IS)
`(45), in which AC potentials are applied over a wide fre-
`quency range. At a single fixed frequency, AC potentials
`may be applied and analyzed if only relative changes after
`substrate application are expected. IS yields information
`about both conductivity and dielectric constant (capaci-
`tance) of the interfacial region of the cell monolayer.
`Essentially three types of brain capillary endothelial cell
`culture are currently used by researchers: primary cultures,
`cell lines and co-culture systems. The limitation of primary
`cultures has been their higher para-cellular permeability,
`reflected by the measurement of the electrical resistance
`across the monolayer. Later developments led to the gen-
`eration of rat, bovine and human immortalized endothelial
`cells and their use as a replacement for primary cells in in-
`vitro BBB models (46). However, these cell systems have
`not been characterized to the same extent as either pri-
`mary or passaged cells. The in-vitro BBB model, consist-
`ing of a co-culture of brain capillary endothelial cells on
`one side of a filter and astrocytes on the other, is currently
`used. The strong correlation between the in-vivo and in-
`vitro values demonstrated that this in-vitro system is an
`important tool for the investigation of the role of the BBB
`in the delivery of nutrients and drugs to the CNS (47). The
`main advantage of this model is the possible rapid evalua-
`tion of strategies for achieving drug targeting to the CNS
`or to appreciate the eventual central toxicity of systemic
`drug and to elucidate the molecular transport mechanism
`of substances across the BBB.
`
`STRATEGIES FOR ENHANCED CNS DRUG
`
`DELIVERY
`To circumvent the multitude of barriers inhibiting CNS
`penetration by potential therapeutic agents, numerous
`drug delivery strategies have been developed (6, 9, 15, 48-
`50). These strategies generally fall into one or more of the
`following three categories: manipulating drugs, disrupting
`the BBB and finding alternative routes for drug delivery.
`
`Drug Manipulations
`Lipophilic Analogs
`CNS penetration is favored by low molecular weight, lack
`of ionization at physiological pH, and lipophilicity (13).
`Delivery of poorly lipid-soluble compounds to the brain
`requires some way of getting past the BBB. There are sev-
`eral possible strategies, such as transient osmotic opening
`of the BBB, exploiting natural chemical transport