Review Article
`
`JPP 2004, 56: 3–17
`ß 2004 The Authors
`Received October 2, 2003
`Accepted November 14, 2003
`DOI 10.1211/0022357022764
`ISSN 0022-3573
`
`Is nose-to-brain transport of drugs in man
`a reality?
`
`Lisbeth Illum
`
`Abstract
`
`The blood–brain barrier that segregates the brain interstitial fluid from the circulating blood
`provides an efficient barrier for the diffusion of most, especially polar, drugs from the blood to
`receptors in the central nervous system (CNS). Hence limitations are evident in the treatment of CNS
`diseases, such as Parkinson’s and Alzheimer’s diseases, especially exploiting neuropeptides and
`similar polar and large molecular weight drugs. In recent years interest has been expressed in the
`use of the nasal route for delivery of drugs to the brain, exploiting the olfactory pathway. A wealth
`of studies has reported proof of nose-to-brain delivery of a range of different drugs in animal
`models, such as the rat. Studies in man have mostly compared the pharmacological effects (e.g.
`brain functions) of nasally applied drugs with parenterally applied drugs and have shown a distinct
`indication of direct nose-to-brain transport. Recent studies in volunteers involving cerebrospinal
`fluid sampling, blood sampling and pharmacokinetic analysis after nasal, and in some instances
`parenteral administration of different drugs, have in my opinion confirmed the likely existence of a
`direct pathway from nose to brain.
`
`Introduction
`
`In the last decade increasing interest has been expressed in the possibility of circum-
`venting the blood±brain barrier for the delivery of drugs to the central nervous system
`by exploiting the potential direct transport pathway from nose to brain via the
`olfactory region. Such a pathway has been proven to exist in animal models, but it is
`still debatable whether a similar transport takes place in man. Hence,
`it is still
`debatable whether such delivery of drugs to the brain could be exploited therapeuti-
`cally for diseases of the central nervous system (Mathison et al 1998; Illum 2000;
`Pardridge 2001; Thorne & Frey 2001; Minn et al 2002). This would be especially
`beneficial for drugs that do not cross the blood±brain barrier easily due to their
`physicochemical characteristics.
`The vasculature of the central nervous system (CNS) is characterized by the
`existence of the blood±brain barrier that separates the brain interstitial fluid from
`the circulating blood. Apart from protecting the brain from agents in the blood that
`could impair neurological functions, the blood±brain barrier controls influx and efflux
`of substances to provide the brain with necessary nutrients and maintain proper
`homeostasis. The cells of the capillary epithelium in the brain are closely connected
`by complex tight junctions. These tight junctions completely encircle each endothelial
`cell like a belt and join both adjacent cells and contiguous borders of the same cell. In
`addition, each brain capillary is composed of two lipid membranes separated by
`300 nm of endothelial cytosol, the luminal membrane facing the blood and the anti-
`luminal membrane, facing the brain (Pardridge 1991).
`Lipid soluble molecules are absorbed rapidly and efficiently across the nasal mem-
`brane into the systemic blood stream via the transcellular pathway with a plasma
`profile resembling that of an intravenous injection and with a bioavailability of up to
`100%. Due to this rapid absorption such molecules do not normally show direct nose-
`to-brain transport, although this might be dependent on the site of deposition in the
`nasal cavity (Illum 2003). Once such lipophilic molecules reach the blood stream they
`can diffuse freely through the blood±brain barrier and reach the CNS. This diffusion is
`
`3
`AQUESTIVE EXHIBIT 1116 Page 0001
`
`IDentity, 19 Cavendish Crescent
`North, the Park, Nottingham
`NG7 1BA, UK
`
`Lisbeth Illum
`
`Correspondence: L. Illum,
`IDentity, 19 Cavendish Crescent
`North, the Park, Nottingham
`NG7 1BA, UK. E-mail:
`Lisbeth.illum@illumdavis.com
`
`

`

`Olfactory(cid:0)region
`(hatched(cid:0)area)
`
`Middle(cid:0)
`turbinate
`
`Superior(cid:0)
`turbinate
`
`E
`
`D
`
`Inferior(cid:0)
`turbinate
`
`C
`
`AB
`
`Internal(cid:0)
`ostium
`
`Nasal(cid:0)
`vestibule
`
`Figure 1 Schematic representation of the lateral wall of the human
`nasal cavity.
`
`vestibule, the respiratory region and the olfactory region.
`The nasal vestibule (0.6 cm2) is covered with stratified
`squamous epithelium (very similar to skin) and is the
`part of the nose one can reach with an index finger. The
`olfactory region in man is situated in the roof of the nasal
`cavity lying partly on the nasal septum and partly on the
`superior and middle turbinates. The olfactory mucosa
`covers a relatively small area of approximately 4 cm2 or
`3±5% of the area of the total nasal cavity (Morrison &
`Constanzo 1992). However,
`it has been suggested that
`the tips of the olfactory sensory neurons can stretch
`further into the nasal cavity and hence be accessible over
`a larger area (personal communication, N. Jones). As a
`comparison, in the dog the olfactory mucosa constitutes
`77% and in the rat 50% of the total nasal area (Illum
`1996).
`
`The respiratory epithelium
`The anterior part of the nasal cavity is covered with
`squamous epithelium that gradually changes posteriorly
`into the respiratory epithelium comprising a pseudostrati-
`fied columnar epithelium. The cells of the respiratory
`epithelium are covered with microvilli. These provide
`this part of the nasal cavity with a relatively high absorp-
`tive capacity, due to an increase in the surface area, and
`make this the major site for systemic drug absorption. The
`respiratory epithelium consists of four major cell types,
`namely the ciliated (approximately 15±20% of the respira-
`tory cells) and the non-ciliated columnar cells, the goblet
`cells and the basal cells. The cilia project 2±4 ·m from the
`surface of the cells, are mobile and through a co-ordinated
`movement (synchronized beating, 1000 strokes min¡1) are
`able to propel the mucous layer, covering the respiratory
`epithelium, anteriorly towards the nasopharynx. Mucus is
`mainly derived from the goblet cells, interspersed between
`the columnar cells and is the major component of the
`mucous layer. The mucous layer consists of a low viscosity
`sol layer that surrounds the cilia and a more viscous gel
`
`4
`
`Lisbeth Illum
`
`qualified by the degree of lipid solubility and molecular
`size, with smaller molecules passing through the mem-
`brane more easily than larger ones (Temsamani 2002).
`Less lipophilic or polar molecules are not as readily
`absorbed across the nasal membrane into the systemic
`circulation, with bioavailabilities being in the order of
`10% or less for low molecular weight and less than 1%
`for large molecular weight polar molecules such as peptide
`drugs (Illum 2000). Such molecules normally pass the
`nasal membrane via the paracellular pathway, through
`the tight junctions. This pathway is less efficient than the
`transcellular pathway and very dependent on the molecu-
`lar weight of the molecule. Once in the systemic circula-
`tion, the hydrophilic molecules do not pass the blood±
`brain barrier easily unless aided by some form of recep-
`tor or carrier mediated transport mechanism (Schwartz
`et al 1990), whether naturally occurring (as is the case
`for insulin) or by a specific drug delivery approach
`(Pardridge 2001). Polar molecules do not rapidly diffuse
`across the nasal membrane into the systemic circulation
`and so they have a better chance of reaching the olfactory
`mucosa and from there being transported across into the
`CNS. This has been demonstrated in many animal studies
`(Illum 2000).
`This review sets out to discuss recent relevant studies
`concerning the potential of drugs applied to the nasal
`cavity being at least partially transported via the olfactory
`pathway to the CNS. These studies have been published in
`the literature or have been provided as information at
`scientific meetings and largely concern investigations in
`man. Support in the understanding of the subject will be
`provided in the form of a brief overview of nasal morpho-
`logy and physiological function.
`
`The human nose
`
`To comprehend fully the intricacies of nasal drug delivery
`and to evaluate whether nose-to-brain transport of drugs
`is a reality, it is important to have an understanding of the
`relevant morphological
`structures and physiological
`factors affecting these functions. Comprehensive reviews
`dealing with the morphology and physiology of
`the
`nose,
`to include
`the olfactory mucosa, have been
`published (Mygind 1978; Moran et al 1982; Hilger 1989)
`and hence only limited necessary details will be given
`here.
`
`Structure and function of the human nose
`An outline of the human nose is shown in Figure 1. The
`total surface area of the nasal cavity is approximately
`150 cm2 in a man and normally less in a woman. The
`cavity is divided longitudinally into two non-connected
`parts by the nasal septum. The two cavities open ante-
`riorly to the facial site through the narrow (0.3 cm2 in
`diameter) nasal apertures or ``the nasal valve’’ at the top
`of the nostril and posteriorly to the rhinopharynx via the
`posterior nasal apertures. Each of the two nasal cavities
`are largely subdivided into three regions i.e. the nasal
`
`AQUESTIVE EXHIBIT 1116 Page 0002
`
`

`

`layer on top of the cilia. Hence, materials deposited on the
`mucous layer will gradually be cleared from the nasal
`cavity by this mucociliary clearance mechanism. For
`non-mucoadhesive materials this will generally result in a
`half-time of clearance of approximately 15±20 min (Illum
`2000).
`
`Epithelial cell barrier ± tight junctions
`The epithelial cells on the apical surface of the membrane
`are closely connected by intercellular junctions. The struc-
`tural components and specialized sites of these junctions
`are generally known as the junctional complex. They are
`composed of three regions and are, in successive order
`from the apical surface towards the basal surface, the
`zona occludens (ZO) also known as the tight junction,
`the zonola adherens and the macula adherens (Madara
`2000). These complexes create a regulatable semiperme-
`able diffusion barrier between cells. It is clear that the
`tight junction is a dynamic structure that is selectively
`permeable to certain hydrophilic molecules (ions, nutri-
`ents and drugs). The permeability of the tight junction
`varies between the epithelial tissues in the body but is
`generally limited for molecules with a hydrodynamic
`radius larger than 3.6 AÊ and negligible to molecules with
`a radius larger than 15 AÊ
`(Stevenson et al 1988). It is
`difficult to relate these sizes to exact molecular weights
`since the size of a molecule, and especially peptides and
`proteins, will be determined by the physicochemical envir-
`onment, and possible secondary and tertiary structures of
`the molecules. However, it has been shown in the litera-
`ture that for molecules of a molecular weight of approxi-
`mately 1000 Da and larger, the transport through tight
`junctions is normally very restricted (McMartin et al
`1987).
`The tight junction comprises a series of transmembrane
`and cytosolic proteins that interact not only with each
`other but also with the membrane and the cytoskeleton
`e.g. occludins, claudins and junctional adhesion molecule
`(Anderson & Van Itallie 1995; Denker & Nigram 1998)
`(Figure 2). The topology of occludin suggests that the
`amino and the carboxyl termini of this protein are situated
`in the cytoplasm of the cell with two extracellular loops
`projecting into the paracellular space between adjacent
`cells. The loops of the extracellular occludin from two
`neighbouring cells may interact in the extracellular space
`to promote sealing of the paracellular space. The cytoplas-
`mic occludin interacts with tight junction-associated pro-
`teins present in the cytoplasm (ZO-1, ZO-2 and ZO-3)
`(Ward et al 2000). For example, the N-terminal of the
`ZO-1 interacts with the C-terminal tail of occludin and
`its C-terminal interacts with F-actin of the cytoskeleton and
`thereby couples the tight junction to the scaffold of the
`cytoskeleton. The ZO-2 interacts with the C-terminal of
`the occludin and the N-terminal of ZO-1. The claudins
`have been suggested to be major structural components of
`tight junction strands in line with occludins. The third
`transmembrane protein junctional adhesion molecule is
`different structurally to the occludins and claudins, and
`is immunoglobulin-like in form.
`
`Is nose-to-brain transport of drugs in man a reality?
`
`5
`
`Figure 2 Schematic representation of the tight junction and the
`interaction of the transmembrane and cytosolic proteins (adapted
`from Ward et al (2000)).
`
`The zona occludens is closely associated with the
`zonola adherens complex. The zonola adherens complex
`holds cells close together but does not form a tight barrier.
`The zonola adherens is made up of transmembrane pro-
`teins known as cadherins. Both zona occludens and
`zonola adherens structures act to anchor cytoskeleton
`components.
`Many classical second messengers and protein kinases
`of signalling pathways such as tyrosine kinases, Ca2‡ and
`protein kinase C (PKC) influence both the barrier proper-
`ties and assembly of the tight junction. Hence increases in
`intracellular calcium can affect phosphorylation of myo-
`sin regulatory light chain contraction of perijunctional actin
`and cause increased paracellular permeability (Ward et al
`2000). PKC plays a dual role in that it initiates tight junc-
`tion synthesis under conditions that preclude tight junction
`synthesis (e.g. incubation in low calcium medium) and also
`appear to be involved in tight junction disruption in condi-
`tions that encourage tight junction formation (e.g. incuba-
`tion in normal calcium medium). Hence PKC is strongly
`involved in the highly complex signal transduction process
`that regulates the tight junction. The phosphorylation of the
`tight junction proteins or the displacement (i.e. contraction
`or relaxation) of the perijunctional actin-myosin ring is
`generally the final effect of modulation of many of these
`signalling pathways. This has been shown by the fact that a
`disruption of the tight junction integrity by ATP depletion
`
`AQUESTIVE EXHIBIT 1116 Page 0003
`
`

`

`6
`
`Lisbeth Illum
`
`induces a decrease in phosphorylation of the tight junction
`regulatory proteins. During ATP repletion the phosphory-
`lation is increased again (Tsukamoto & Nigam 1999).
`Furthermore, the same signalling pathway that induces
`phosphorylation of the tight junction proteins may also
`modulate the actin cytoskeleton, which again has been
`shown to increase the transmembrane flux of sodium and
`mannitol. Recently, it has been shown that cationic polymer
`absorption enhancers, such as poly-L-arginine and chitosan,
`which predominantly work by transiently opening epithelial
`tight junctions, initiate this mechanism by activating the
`PKC signalling pathway (Natsume et al 2003).
`
`The olfactory mucosa
`The olfactory organ is unique in the CNS, since it is the
`only part in direct contact with the environment and hence
`exposed to volatile odorants and airborne (toxic) sub-
`stances. The olfactory mucosa is located within the
`recesses of the skull, just under the cribriform plate of
`the ethmoid bone, approximately 7 cm from the nostril,
`at the top of the nasal cavity, lying partly on the nasal
`septum and partly on the superior turbinate (Figure 1).
`The olfactory region is not easily accessible anatomically
`in living human beings since to reach this area (for exam-
`ple in biopsy) an instrument must pass through a 1.5-mm
`crevasse between closely apposed nasal structures (turbi-
`nates and septum). The olfactory mucosa is above the
`normal airflow path, and hence odorants normally reach
`the sentive receptors on the neurons by diffusion. The size
`of the olfactory region in man has been quoted as 3.7 cm2
`(Jones 2001), 10 cm2 (Proctor 1977) and as 2±10 cm2
`(Morrison & Constanzo 1990). The region is much smaller
`than, for example, that found in dogs (150 cm2), indicating
`the importance of olfaction in the daily functions of dogs
`but not of man.
`The olfactory epithelium is a modified (pseudostrati-
`fied) respiratory epithelium. It comprises olfactory sen-
`sory neurons, sustentacular cells (also called supporting
`cells)
`that ensheath the receptor neurons providing
`mechanical support and maintain the normal extracellular
`potassium levels needed for neuronal activity, and basal
`cells, which are able to differentiate into neuronal receptor
`cells and replace these every 40 days (Figure 3). The
`underlying lamina propria contains olfactory nerve fasci-
`cles and the mucus secreting tubuloalveolar Bowman’s
`glands. The olfactory receptor cells are bipolar neurons
`with a round cell body. A single dendritic process extends
`from the cell body to the free apical surface where it
`terminates as a small knob-like swelling from which
`extends numerous (10±23) long and non-motile cilia. The
`olfactory sensory neurons taper into an unmyelinated
`axon which penetrates the basal membrane to join other
`axons and form large bundles in the lamina propria. The
`unbranched axons are ensheathed by glial cells, also called
`Schwann cells, and cross into the cranical cavity through
`small holes in the cribriform plate and synapse in the
`olfactory bulb. Approximately 1500 olfactory receptor
`cells on the bipolar sensory neurons converge on one
`mitral cell or tufted cell in the olfactory bulb (12.2 mm,
`range 6±16 mm, long). The mitral and the tufted cells
`
`Figure 3 The structure of the olfactory epithelium (adapted from
`Firestein (2001)).
`
`project a single primary dendrite to a single glomerolus
`and emit several dendrites within the external plexiform
`layer. From the olfactory bulb tract the main axons origi-
`nate in the mitral or tufted cells and give off striae, which
`pass to the olfactory tubercle. The projections then go to
`the amygdala, the prepyriform cortex, the anterior olfac-
`tory nucleus and the entorhinal cortex as well as the
`hippocampus, hypothalamus and thalamus.
`The olfactory epithelium is covered by a dense and
`viscous layer of mucus, which is secreted from the
`Bowman’s glands and the supporting cells. Due to the
`non-motile cilia the mucus layer in the olfactory region
`is not cleared by a mucociliary clearance mechanism as in
`the respiratory epithelium. Over-production of mucus
`results in the mucus layer slowly moving into the respira-
`tory region from where it is cleared by the normal
`mechanism of mucociliary clearance.
`At the luminal surface in the olfactory epithelium the
`membranes of the adjoining receptor cells and supporting
`cells are connected by typical junctional complexes similar
`to those described for
`the
`respiratory
`epithelium
`(Engstrom et al 1989). The olfactory region is supplied
`with blood from the anterior and posterior ethmoidal
`branches of the ophthalmic artery supply and venous
`drainage is as for the respiratory system via the spheno-
`palatine foramen into the pterygoid plexus or via the
`superior ophthalmic vein.
`
`Transport of drugs from nose to brain
`
`The CNS
`The CNS is protected against trauma by the cranium (skull)
`that encases the brain and the vertebral column that sur-
`
`AQUESTIVE EXHIBIT 1116 Page 0004
`
`

`

`Scalp
`
`Skull(cid:0)bone
`
`Arachnoid
`villus
`Arachnoid
`mater
`
`Subarachnoid
`space(cid:0)of(cid:0)brain
`
`Brain(cid:0)(cerebrum)
`
`Dura(cid:0)mater
`
`Dural(cid:0)sinus
`
`Pia(cid:0)mater
`
`Venous(cid:0)sinus
`
`Figure 4 Relationship of meninges and cerebrospinal fluid to brain
`and spinal cord. Frontal section in the region between the two
`cerebral hemispheres of the brain, depicting the meninges in greater
`detail. (Adapted from Illum (2000).)
`
`rounds the spinal cord. Three protective membranes, called
`the meninges, lie between the skull and the brain tissue
`(Pardridge 1991; Thorne & Frey 2001). Moving in the
`direction from the skull to the brain, these are the dura
`mater, the arachnoid mater and the pia mater. The dura
`mater consists of two layers, which are normally closely
`adherent. However, in some regions they are separated by
`blood-filled cavities, the dural sinuses or venous sinuses
`(Figure 4). Venous blood from the brain empties into
`these sinuses to be returned to the heart. The space between
`the arachnoid and pia mater, the subarachnoid space, is
`filled with cerebrospinal fluid (CSF) in which the brain is
`essentially suspended. Protrusions of arachnoid tissue, the
`arachnoid villi, penetrate through gaps in the overlying
`dura and project into the dural sinuses. It is across the
`surfaces of these villi that the CSF is reabsorbed into the
`blood circulating within the sinuses. The CSF is produced
`primarily by the four choroid plexi found in particular
`regions of the ventricle cavities of the brain. Once formed
`it flows through the four interconnected ventricles within
`the interior of the brain and through the spinal cord’s
`narrow central canal, which is continuous with the last
`ventricle, and escapes from this fourth ventricle at the
`base of the brain to enter the subarachnoid space. When
`the CSF reaches the upper regions of the brain, it is re-
`absorbed into the venous blood through the arachnoid villi.
`It is known also that the CSF can drain from the subarach-
`
`Is nose-to-brain transport of drugs in man a reality?
`
`7
`
`noid space through the perivascular space surrounding the
`nerve bundles in the cribriform plate, and enters the olfac-
`tory submucosa where it drains into the nasal lymphatics
`(Pardridge 1991). This drainage constitutes less than 5% of
`the CSF.
`Through the ongoing procedure of formation, circula-
`tion and re-absorption of the CSF, the entire volume of
`approximately 125±150 mL (in adults) is replaced more
`than three times a day (Sherwood 1989). In comparison
`the rat brain contains only 150 ·L CSF and is replaced
`approximately 24 times a day. These differences in CSF
`renewal between rat and man could have a significant
`impact on interpretation of nose-to-brain drug delivery
`studies and together with the other anatomical differences
`depicted in Table 1 should always be carefully considered.
`Knowledge of the manner in which drugs diffuse from
`the CSF into the brain parenchyma and the probability of
`this is important for the understanding of the significance
`of uptake of drug into the CSF after nasal application for
`treatment of CNS diseases. Unless receptors for the drug
`are present on the surface of the brain the drug will by
`necessity have to penetrate into the brain tissue. The rate
`of diffusion of drugs in the extracellular space of the brain
`can be expressed as D* ˆ D/l2, where D is the diffusion
`coefficient of the molecule in water and l is tortuosity.
`Tortuosity is a dimensionless parameter reflecting the
`restrictions placed on the diffusion of the molecule by
`cellular elements and the connectivity of the extracellular
`spaces into which the molecule has access (Nicholson &
`Sykova 1998). Values of l vary from 1.4 for small mole-
`cules to 2.5 for large molecules such as albumin. D is
`inversely related to the molecular size of the drug. It can
`be calculated that the time it takes for a small molecule
`such as glucose to diffuse 5 mm in the brain is approxi-
`mately 11.7 h and for a molecule such as albumin 4.2 days!
`The fact that there is a distinct difference between the
`bulk flow properties of the CSF and diffusional flow rates
`in the brain tissue creates a functional barrier between
`the CSF and the brain tissue (Pardridge 1991). This
`prevents complete equilibration between the two fluid
`compartments and consequently a significantly different
`drug concentration will normally exist between these two
`compartments.
`
`Transport pathways
`It is suggested in the literature that a drug administered
`nasally is able to reach the CNS (i.e. CSF and brain tissue)
`
`Table 1 The characteristics of the rat animal model vs man in relation to nose-to-brain
`delivery of drugs.
`
`The nasal cavity is approximately 180 cm2 in man and approximately 10 cm2 in rats.
`The olfactory area constitutes approximately 3% of the nasal cavity in man, but 50% in rat.
`The CSF volume is 160 mL in adult humans and 150 ·L in rats.
`The CSF volume is replaced every 5 h in man and every 1 h in rats.
`The placement of the rat on its back in most experiments with easy access to the olfactory area influences
`CSF uptake.
`
`AQUESTIVE EXHIBIT 1116 Page 0005
`
`

`

`8
`
`Lisbeth Illum
`
`Enzymatic
`degradation
`
`Mucociliary
`clearance
`
`Nasal(cid:0)cavity
`
`Trigeminal
`nerve
`receptor(cid:0)(cid:0)
`
`?
`
`Olfactory
`region(cid:0)
`
`Brain(cid:0)tissue
`
`Blood–brain(cid:0)
`barrier
`choroid
`plexus(cid:0)
`
`CSF
`
`Blood
`
`Elimination
`
`Figure 5 Suggested pathways from nose to brain.
`
`by the various transport routes shown schematically in
`Figure 5. After nasal application, drug that has escaped
`enzymatic degradation and the normal rapid clearance by
`the mucociliary clearance system may be transported across
`the nasal membrane into the systemic circulation. As men-
`tioned above, such absorption may for lipophilic drugs
`reach close to 100% (e.g. π71% for fentanyl in man), but
`is normally less. The drug is subsequently eliminated from
`the blood by the normal clearance mechanisms. However,
`once the drug is in the blood it may (if it is sufficiently
`lipophilic or by exploiting specific transport mechanisms)
`cross the blood±brain barrier and reach the brain and the
`CSF (the so-called systemic pathway). Drug present in the
`CSF or the brain tissue will also be eliminated into the
`blood and cleared. Of special interest to the present review
`is the fact that a drug may be transported directly into the
`brain tissue (e.g. olfactory bulb) or the CSF by transport
`across the olfactory region of the nasal cavity (the so-called
`olfactory pathway). Recently, preliminary evidence has
`emerged that suggests that drugs may also be transported
`to the brain via trigeminal nerve receptors present in the
`nasal cavity (Thorne et al 2000). These receptors are respon-
`sible for most chemoperception apart from olfaction.
`The various pathways that a drug can follow from the
`olfactory region of the nasal cavity to reach the CSF or
`the brain tissue have been discussed thoroughly (Mathison
`et al 1998; Dahlin 2000; Illum 2000; Thorne & Frey 2001),
`hence only a brief discussion will be given here.
`Leaving the trigeminal pathway aside, the nasal pathway
`from nose to CNS is thought to involve one or a combina-
`tion of two general mechanisms. The first is internalization
`of the drug into the primary neurons of the olfactory
`epithelium and transport by intracellular axonal transport
`to the olfactory bulb with subsequent possible distribution
`
`of the drug into more distant brain tissues. The second is
`absorption of the drug across the olfactory sustentacular
`epithelial cells, either by transcellular or paracellular
`mechanisms followed by uptake into the CSF or CNS.
`Drugs transported intracellularly in the olfactory neu-
`rons (axonal transport) are thought to enter the neurons
`by mechanisms of endocytosis or pinocytosis. They travel
`along the axon and via the nerve bundle, transverse the
`cribriform plate and reach the olfactory bulb. As
`described in The olfactory mucosa above several dendrites
`are emitted further into the CNS from the tufted cells at
`the first order synapse in the olfactory bulb.
`The existence of the axonal pathway has been described
`by several authors for transport of different materials
`from the olfactory region to the CNS e.g. gold particles
`(De Lorenzo 1970; Gopinath et al 1978), aluminium lactate
`(Perl & Good 1987) and wheat germ agglutinin± horse-
`radish peroxidase (Shipley 1985; Baker & Spencer 1986;
`Itaya 1987; Thorne et al 1995). For the last material it was
`found that uptake into the neural cell was by receptor-
`mediated endocytosis and that horseradish peroxidase
`alone was not able to reach the olfactory bulb in significant
`quantities due to a different transport pathway (Thorne
`et al 1995). It has also been shown in the above experiments
`and in others that the axonal route of transport is very slow
`and that it can take up to 24 h before the drug reaches the
`CNS (Kristensson & Olsson 1971).
`As opposed to the axonal pathway, the olfactory epithe-
`lial pathway for transport of drugs appears to be very fast,
`with drugs appearing in the CSF and in the brain a few
`minutes after nasal application. This has been shown among
`others for dihydroergotamine (Wang et al 1998), cocaine
`(Chow et al 1999), lidocaine (Chou & Donovan 1998) and
`cefalexin (Sakane et al 1991). The extracellular pathway,
`that transports polar drugs through tight junctions (see
`Epithelial cell barrier above) between sustentacular cells
`and olfactory neurons into the CSF, relies on a direct
`anatomic connection between the submucosa and the sub-
`arachnoid extensions, the perineural space surrounding the
`olfactory nerves, as they penetrate the cribriform plate
`(Figure 6) (Jackson et al 1979). The drug is thought to
`enter the perineural space either through loosely adherent
`perineural epithelium surrounding the axon (``open-cuff
`model’’), or to enter through the epithelial cell junctions if
`the perineural epithelium is closely adherent to the axon
`(``closed-cuff model’’). More lipophilic drugs passing
`though the epithelial cells transcellularly will reach the
`submucosa also and from there can likewise reach the
`perineural space. It has been shown in a rat model that
`large molecular weight drugs, such as protein nerve growth
`factor (MW 37 kDa) (Thorne & Frey 2001), insulin (MW
`6 kDa) (Gizurarson et al 1996) and vasoactive intestinal
`peptide (VIP) (MW 3.5 kDa) (Gozes et al 1996) can be
`transported rapidly into the CSF and hence are able to
`exploit the olfactory epithelial pathway in line with small
`molecular weight drugs. Since most studies in animal mod-
`els only cover limited periods of time (<4 h), it is difficult to
`determine from the literature whether drugs that are trans-
`ported initially by the olfactory epithelial pathway would
`also show exploitation of the axonal pathway.
`
`AQUESTIVE EXHIBIT 1116 Page 0006
`
`

`

`Olfactory(cid:0)nerve(cid:0)
`cell
`Bowman’s(cid:0)gland
`
`Supporting(cid:0)cell
`
`Olfactory(cid:0)epithelium
`
`Basal(cid:0)cell
`
`Lamina(cid:0)propria
`
`Schwann’s(cid:0)cell
`
`Perineural(cid:0)cell
`
`Cribriform(cid:0)plate
`
`Subarachnoid(cid:0)space
`
`Olfactory(cid:0)axons(cid:0)extending(cid:0)to(cid:0)the
`olfactory(cid:0)bulb
`
`Figure 6 Anatomical connection between the olfactory epithelium and
`the CSF in the subarachnoid space (modified from Mathieson et al
`(1998)).
`
`Transport of drugs in man
`
`Animal studies
`From the description of nose-to-brain studies in animal
`models in the literature it is evident that small molecular
`weight drugs of a suitable lipophilicity, and also larger
`hydrophilic molecules, can be transported from the nasal
`cavity into the CSF, the olfactory lobe and for some drugs
`further into the brain tissue. It is evident that if very
`lipophilic drugs, such as progesterone and estradiol, are
`administered to the nasal cavity, they will be absorbed
`rapidly and efficiently across the nasal membrane. They
`will provide a plasma concentration profile similar to that
`seen after an intravenous injection, and as such will not
`show a higher CSF or CNS uptake when given nasally as
`compared with an intravenous injection.
`As background information a selection of studies per-
`formed on various drugs and in various animal models is
`given in Table 2 together with the key results. These studies
`will not be discussed further in this review, which is focused
`primarily on a discussion of results from nose-to-brain
`studies in man.
`
`Human studies
`Pharmacologicalevidence ofnose-to-braintransport Most
`of the published studies evaluating nose-to-brain delivery
`of drugs in man do not describe the direct measurement of
`the rate and degree of transport into the CNS region but
`rather have measured indirectly the pharmacological
`
`Is nose-to-brain transport of drugs in man a reality?
`
`9
`
`effects of drugs on the CNS, e.g. the effect of the drug
`on event related brain potentials and working memory
`function. It should be mentioned that most of the pub-
`lished studies involving indirect measures originate from
`one research group in Lubeck, Germany, using peptides
`such as insulin, vasopressin and melanocortin. However,
`in the last few years studies have been published (or pre-
`sented at meetings) where the appearance of drug in the
`CSF after nasal administration has been determined.
`Furthermore, a single study has reported the evaluation
`of the transport of radiolabelled drug into the brain using
`®-scintigraphy measurements. A summary of these studies
`is given in Table 3.
`Pietrowsky et al (1996a) found evidence that after nasal
`application of 20 IU arginine-vasopressin (AVP), in a cross-
`over study in 15 volunteers, a component (P3) of an event-
`related brai