`
`E'Ls EVIR
`
`European Journal of Pharmaceutical Sciences 11 (2000) 1—18
`
`Review
`
`
`IUIOPIAI JOIIIIAL OI
`
`PHARMACEUTICAL
`SCIENCES
`
`www.clseviernl / locate/ ejps
`
`Transport of drugs from the nasal cavity to the central nervous system
`
`Lisbeth Illum B.Sc. Ph.D., D.Sc.*
`West Pharmaceutical Services Drug Delivery and Clinical Research Centre Ltd., Albert Einstein Centre, Nottingham Science and Technology Park,
`Nottingham NG7 ZTN, UK
`
`Received 20 December 1999; received in revised form 9 March 2000; accepted 10 March 2000
`
`Keywords: Nasal cavity; Drug transport; CNS
`
`1. Introduction
`
`The blood—brain barrier that segregates the brain inter-
`stitial fluid from the circulating blood consists of two
`plasma membranes in series,
`the lumenal and the anti-
`lumenal membranes of the brain capillary epithelium,
`separated by about 0.3 mm of endothelial cytosol. The
`cells of the capillary endothelium are closely connected via
`intercellular connections; the tight junctions that act as zips
`closing the inter-endothelial pores that normally exist in
`endothelial membranes. This makes the blood—brain bar-
`rier resistant to the free diffusion of molecules across the
`
`membrane and prevents most molecules from reaching the
`central nervous system from the blood stream. Several
`different approaches have been attempted in order
`to
`circumvent the blood—brain barrier and to deliver drugs
`efficiently to the brain for therapeutic or diagnostic appli—
`cations. Since lipid soluble drugs, with a molecular weight
`less than 600 Da, will readily diffuse through the blood—
`brain barrier a normal approach to increase the permeabili-
`ty of a drug is to create a more lipophilic molecule, often
`in the form of a prodrug that is converted to the parent
`drug once in the brain. Other approaches include the
`binding of drugs to carrier molecules such as transferrin or
`to a polycationic molecule such as cationised proteins that
`will bind preferentially to the negatively charged endo-
`thelial surface (Pardridge, 1991).
`In the last decade, much interest has been given to the
`exploitation of the nasal route for delivery of drugs to the
`brain via a specific site, the olfactory region (Pardridge,
`1991; Thome et al., 1995). Recent times have also seen
`review articles dealing with this topic (Pardridge, 1991;
`Mathison et al., 1998). Indeed, it was realised early in the
`last century that the olfactory region of the nose can be a
`major site for entry of viruses into the brain. Hence, it has
`been shown by several groups that the nasal route can be
`
`*Tel.: +44—115-925-3789; fax: +44-115-925-0351.
`
`involved in the contraction of the neurotropic poliomyelitis
`virus and that the virus reached both the olfactory lobe of
`the brain and the cerebrospinal fluid (CSF) (Landsteiner
`and Levaditi, 1910; Flexner, 1912; Fairbrother and Hurst,
`1930; Faber and Gebhardt, 1933; Sabin and Olitsky, 1936,
`1937a, 1938; Faber, 1938). In addition to poliomyelitis
`virus, Sabin and Olitsky (1937b) also showed that
`the
`vesicular stomatitis virus could enter the CNS via the nasal
`
`cavity. It was later demonstrated by Bodian and Howe
`(1940) that viruses can move from the nose to the brain via
`the olfactory neurons. This was confirmed in a study by
`Reiss et a1. (1998) in mice who showed that vesicular
`stomatitis virus, applied to the nasal neuroepithelium,
`initially replicated in the olfactory receptor neurons and
`was transmitted along the olfactory nerve to the CNS
`within 12 h. The virus replicated invasively in the olfactory
`bulb and reached the olfactory ventricle by 4—5 days post
`infection and the hindbrain by day 8.
`Studies in animals have reported that tracer materials,
`such as potassium ferrocyanide and iron ammonium citrate
`(Faber, 1937), albumin (Kristensson and Olsson, 1971),
`horseradish peroxidase (Stewart, 1985),
`the conjugate
`wheat germ agglutinin—horseradish peroxidase (Shipley,
`1985; Baker
`and Spencer,
`1986)
`and colloid gold
`(Gopinath et al., 1978), are also able to be transported
`across the olfactory epithelium into the CNS. A number of
`studies have also been published on the transport of heavy
`metals from the nasal mucosa into the CNS via olfactory
`pathways (Evans and Hastings, 1992; Tjalve et al., 1996;
`Gianutsos et al., 1997). Similarly, a large number of
`studies have been performed where low molecular weight
`drugs and peptides
`(such as estradiol, progesterone,
`cephalexin, dihydroergotamine and cocaine) have been
`shown to reach the CSF, the olfactory bulb and in some
`cases other parts of the brain after nasal administration
`(Anand Kumar et al., 1974, 1976, 1979, 1982; Hussain et
`al., 1981; Sakane eta1., 1991a,b, 1994, 1995; Wang et al.,
`1998; Chou and Donovan, 1998a; Chow et al., 1999).
`
`0928-0987/ 00/ $ — see front matter © 2000 Elsevier Science B.V. All rights reserved.
`
`AQUESTIVE EXHIBIT 1015 Page 0001
`
`

`

`2
`
`L. Illum / European Journal of Pharmaceutical Sciences I I (2000) 1—18
`
`Studies have also been directed to investigate the mecha-
`nism of transport of drugs across the olfactory epithelial
`membrane and the salient physicochemical characteristics
`of the drug affecting CNS uptake (Sakane et al., 1994,
`1995; Chou and Donovan, 1998a). Most studies have been
`carried out
`in animal models, such as the rat and the
`monkey, and only few studies have been published where
`evidence of transport of drugs from nose to brain in
`humans has been given (Pietrowsky et al., l996a,b; Kern
`et al., 1997, 1999; Okuyama, 1997; Derad et al., 1998).
`The present review sets out
`to discuss the barriers
`involved in the passage of drugs from the nose to the brain
`and the likely transport pathways that drugs might take.
`Furthermore, studies in the literature describing the trans-
`port of drugs to the CNS in different animal models and
`man will be reviewed together with a discussion of the
`feasibility of deep brain penetration following nasal ad-
`ministration.
`
`2. The human nose
`
`The structure and the function of the human nose have
`
`been comprehensively described by various authors includ—
`ing Mygind (1978) and Hilger (1989). Therefore, only
`details that are relevant
`for
`the understanding of the
`
`morphological and physiological factors affecting the nasal
`absorption of drugs, and in particular the transport of drugs
`from the nasal cavity to the central nervous system, will be
`provided in this review.
`
`2.1. The human nasal structure and function
`
`The nasal cavity is subdivided along the centre into two
`halves by the nasal septum. The two cavities open to the
`facial side through the anterior nasal apertures and to the
`rhinopharynx via the posterior nasal apertures. The total
`surface area of the nasal cavity in man is about 150 cm2
`and the total volume about 15 ml. Each of the two nasal
`
`cavities can be subdivided into three regions; namely the
`nasal vestibule, the olfactory region and the respiratory
`region. The olfactory region in man covers an area of
`about 10 cm2 and is positioned on the superior turbinate
`and opposite the septum. The respiratory region is domi-
`nated by the presence of the large inferior turbinate, the
`middle turbinate, which has similarities to a polyp and
`further back in the nose,
`the superior turbinate. Fig.
`1
`shows an outline of the human nasal cavity with a clear
`indication of the position of the olfactory region.
`
`2.1.1. The respiratory region
`In the respiratory region, which is considered the major
`
`
`
`
`
`
`
`Olleclory
`
`Poelerlor
`
`ewerlnr hterel
`neeel Drenchee
`
`Mellon nerve
`
`
`
`Poeterlor
`lnlerlor leIerel
`neeel brenchee
`
`Fig. 1. Outline of human nasal cavity indicating the nasal turbinates, the olfactory region and the olfactory bulb. With permission from W.B. Saunders,
`Philadelphia, PA, USA.
`
`AQUESTIVE EXHIBIT 1015 Page 0002
`
`
`

`

`L. Illum / European Journal of Pharmaceutical Sciences I I (2000) 1—18
`
`3
`
`site for drug absorption into the systemic circulation, the
`mucosa consists of an epithelium resting on a basement
`membrane and a lamina propria (Fig. 2). The anterior part
`of the respiratory region is covered with squamous epi-
`thelium, which changes to a transitional epithelium and
`converts in the posterior part of the cavity to a pseudo—
`stratified columnar epithelium. The pseudostratified epi-
`thelium, also named the respiratory epithelium, consists of
`four dominant cell
`types; ciliated columnar cells, non-
`ciliated columnar cells, goblet cells and basal cells. Of
`these the basal cells are situated on the basal membrane
`
`and do not extend to the apical epithelial surface, as do the
`other three cell types. A total of 15—20% of the respiratory
`cells is covered by a layer of long cilia of size 2—4 p.rn.
`The cilia move in a coordinated way to propel mucus
`across the epithelial surface towards the pharynx. The
`respiratory cells are also covered by about 300 microvilli
`per cell. Microvilli increase the surface area of the cell
`considerably, which in turn promotes the transport of
`substances and water between the cells. Goblet cells are
`
`interspersed between the columnar cells and are the main
`entities responsible for the secretion of the mucus covering
`the epithelial cell layer. The mucus layer consists of a low
`viscosity sol
`layer that surrounds the cilia and a more
`viscous gel layer forming a layer on top of the sol layer
`and covering the tips of the cilia.
`The epithelial cells are closely connected on the apical
`surface,
`surrounded by intercellular
`junctions whose
`
`specialised sites and structural components are commonly
`known as the junctional complex. Each complex is com-
`posed of three regions; the zonola occludens closest to the
`apical surface, further down the zonola adherens and last
`the macula adherens. The zonola occludens (20) forms a
`tight band around the upper part of the cell and is also
`known as the tight junction. The Z0 contains the integral
`protein 20-1 and controls the diffusion of ions and neutral
`molecules through the intercellular spaces. Tight junctions
`and selective paracellular permeability has recently been
`reviewed by Balda and Matter (1998).
`It has been shown by electrophysiological analysis that
`the size of the largest molecule that can permeate the 20
`varies between epithelial tissues in the body. Generally, the
`20 limits the degree of permeability of molecules with a
`hydrodynamic radii larger than 3.6 A. Tight junctions are
`impermeable to molecules with a radius larger than 15 A
`(Madara and Dharmsathaphom, 1985; Madara et al.,
`1986).
`It has also been shown that
`the extracellular
`environment has an effect on the permeability of the
`epithelial membrane in that
`the integrity of the tight
`junction is dependent on the concentration of extracellular
`calcium ions (Stevenson et al., 1988). Compounds, such as
`calcium chelators, are known to be able to open up tight
`junctions by chelating the calcium ions
`(Citi, 1992).
`However,
`it should be noted that
`this effect might be
`associated with changes in other junctional elements.
`Chelators have been shown to induce global changes in
`
`
`
`CT”
`
`Fig. 2. Schematic illustration of the various cell types in the nasal respiratory epithelium. (I) Non—ciliated columnar epithelial cell with microvilli; (II)
`goblet cell with mucus granules and Golgi apparatus; (III) basal cell; (IV) ciliated columnar cell with mitochondria. DM, double membrane; CTM,
`connective tissue membrane. With permission from Blackwell Science Publishers, Oxford, UK.
`
`AQUESTIVE EXHIBIT 1015 Page 0003
`
`

`

`4
`
`L. Illum / European Journal of Pharmaceutical Sciences I I (2000) 1—18
`
`cells, such as the disruption of adherent junctions, di-
`minished cell adhesion and disruption of actin filaments
`(Citi, 1992).
`The blood supply to the respiratory region of the nasal
`cavity comes from the external and the internal carotid
`arteries through a dense network of capillaries in the
`lamina propria. The blood from the main part of the nasal
`cavity is drained via the sphenopalatine foramen into the
`pterygoid plexus or via the superior ophthalmic vein,
`whereas blood from the anterior part of the nose is drained
`via the facial vein.
`
`2.1.2. The olfactory region
`The science of olfaction to include the olfactory region
`in vertebrates has been very adequately described in a
`recent publication by Serby and Chobor
`(1992). The
`olfactory epithelium is dedicated to the detection of smells
`and both the cellular composition and organisation of the
`epithelial
`layers maximise the accessibility of air
`to
`neuronal structures bearing odorant detectors. In man the
`epithelium is restricted to a small area in the roof of the
`nasal cavity of about 10 cm2 as compared to an area of
`about 150 cm2 in dogs, which reflects the major impor-
`tance of the sense of smell to the dog. The olfactory region
`is situated between the nasal septum and the lateral wall of
`each of the two nasal cavities and just below the cribriform
`plate of the ethmoid bone separating the cranial cavity
`from the nasal cavity (Fig. 1). Since the olfactory mucosa
`
`is above the normal path of the airflow, odorants normally
`reach the sensitive receptors by diffusion. The process of
`sniffing enhances the diffusion process by drawing air
`currents upward within the nasal cavity so that a greater
`percentage of the molecules comes into contact with the
`receptor neurones. The olfactory epithelium is a modified
`form of respiratory epithelium in that
`it
`is a pseudo-
`stratified epithelium that consists of three cell types; the
`olfactory receptor cells, supporting epithelial (sustentacu-
`lar) cells and basal cells. Beneath the basement membrane
`is
`the lamina propria, which contains blood vessels,
`olfactory axon bundles, trigeminal and autonomic nerve
`fibres and Bowman’s glands, which secrete the mucus, that
`cover, the epithelial surface. Fig. 3 shows an outline of the
`olfactory epithelium indicating the various types of cells
`present and the connection of neurones to the olfactory
`bulb.
`
`The olfactory receptor cells are bipolar neurones which
`are located in the middle stratum of the olfactory epi-
`thelium and interspersed among the sustentacular cells. 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 a dozen extremely long
`modified non-motile cilia. Chemical detectors, presumably
`receptor protein molecules, are located in the plasma
`membrane of these specialised cilia that float at the tissue/
`mucus/ air interface (Krieger and Breer, 1999). At
`the
`basal aspect of the neuron cell body, each receptor gives
`
`Olfactory
`
`Olluctory
`tract
`
`
`
`Fig. 3. Schematic illustration of the various cell types in the olfactory region in the vault of the human nose. With permission from W.B. Saunders,
`Philadelphia, PA, USA.
`
`AQUESTIVE EXHIBIT 1015 Page 0004
`
`
`

`

`L. Illum / European Journal of Pharmaceutical Sciences I I (2000) 1—18
`
`5
`
`rise to a single fine non-myelinated 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 (Schwann cells) and cross into the
`cranial cavity through small holes in the cribriforin plate of
`the ethrnoid bone. The bundles of nerve fibres travel along
`the surface of the olfactory bulb, eventually synapsing with
`mitral cell dendrites and other neuronal
`targets in the
`glomeruli of the bulb (Fig. 3).
`The sensory neurones, due to their anatomical organisa—
`tion, are in direct contact with volatile odorants and also
`exposed to detrimental airborne substances, which include
`chemicals and viral and bacterial pathogens. This results in
`neuronal death being a feature of normal olfactory epi-
`thelium. In mice, the lifetime of olfactory receptor neuro-
`nes has been estimated to be of the order of 30—90 days
`(Mackay-Sirn and Kittel, 1991). The dying neurones are
`replaced by new sensory neurones. This neurogenesis is a
`specific feature of the olfactory epithelium and not found
`elsewhere in the adult nervous system.
`The sustentacular cells are elongated columnar cells
`with their tapered bases resting on the basement membrane
`and many long microvilli extending from their luminal
`surface to form an entanglement with the cilia of the
`receptor cells. Most sustentacular cells have an unbranched
`stem but some have been shown to have finger-like
`processes
`that wrap themselves around the adjoining
`receptor cells (Moran et al., 1982). The function of the
`sustentacular cells is poorly understood but the cells most
`likely provide mechanical support for the receptor cells
`and have been suggested to be a source of xenobiotic
`metabolising enzymes (Harkema, 1991). The basal cells
`are small, conical cells similar in position to those in the
`respiratory epithelium. The cells are able to differentiate
`into neuronal receptor cells that replace the dying cells as
`described above.
`
`the plasma membranes of
`surface,
`the luminal
`At
`adjoining receptor cells and sustentacular cells are con-
`nected by typical junctional complexes in line with those
`described for the respiratory epithelium (Engstrom et al.,
`1989). The olfactory region is supplied with blood from
`principally the anterior and posterior ethmoidal branches of
`the ophthalmic artery supply. Venous drainage is via the
`same principal veins as for the rest of the nasal cavity.
`
`3. The central nervous system
`
`The central nervous system (CNS) is protected from
`trauma by the skull and vertebra. The brain is surrounded
`by the subarachnoid space in which runs the cerebrospinal
`fluid (CSF). This
`space is again surrounded by the
`meninges which consists of three membranes;
`the dura
`mater, which lies directly beneath the skull, the pia mater
`that
`lies directly over
`the brain, and in between the
`
`Dura mater
`
`
`Arachnoid
`villus
`Arachnoi -
`mater
`
`Dural sinus
`
`Pia mater
`
`Subarachnoid
`.
`Venous sinus
`
`space of brain
`
`
`Fig. 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.
`
`arachnoid. Between the pia mater and the arachnoid is the
`subarachnoid space (Fig. 4).
`The CSF is produced almost entirely by secretion at the
`four choroid plexi, especially at
`the fourth and lateral
`ventricles. CSF is not a plasma filtrate but a secretory fluid
`produced by the choroid plexi. Each choroid plexus
`comprises a secretory epithelium that is perfused by blood
`at a local high perfusion rate. The epithelium is polarised,
`with the apical and basolateral membranes facing the
`ventricular and vascular surfaces, respectively. The CSF is
`secreted by the epithelium across the apical membrane.
`The cells have tight junctions, which constitute the blood—
`CSF barrier. The CSF flows from the choroid plexi and
`circulates over the surfaces and convexities of the brain in
`a rostral to caudal direction and leaves the subarachnoid
`
`space at the arachnoid villi to be absorbed into the blood.
`The volume of CSF present is dependent on age and varies
`from 40 ml in infants up to 160 ml in adults (Allison and
`Stach, 1978). In comparison, the mouse brain and the rat
`brain contains only 35 ul and 150 ul of CSF fluid,
`respectively (Rieselbach et al., 1962; Davson et al., 1987).
`The rate of CSF fluid production, which equals the rate of
`CSF absorption into the peripheral bloodstream at
`the
`arachnoid villi, varies from 21 ml/ h in humans to 0.18
`ml/h in rats and 0.018 ml/h in mice (Table 1). It can
`hence be calculated that for a rat the entire CSF volume
`
`would be totally replaced every hour (i.e. 24 times a day)
`
`Table 1
`Production rates and volumes of CSF in different species3
`
`Species
`Production rate
`Volume in brain
`
`(ml/ h)
`(ml)
`0.018
`0.035
`Mouse
`0.18
`0.15
`Rat
`0.6
`2.3
`Rabbit
`2.5
`—
`Monkey
`7.1
`14.2
`Sheep
`
`
`Man 100.0 21.0
`
`" From Davson et a1. (1987).
`
`AQUESTIVE EXHIBIT 1015 Page 0005
`
`

`

`6
`
`L. Illum / European Journal of Pharmaceutical Sciences I I (2000) 1—18
`
`ARACI-INOID VILLI
`
`
`
`
`_, BRAIN ISF
`
`CHOROID
`
`
`PLEXUS
`BRAIN CAPILLARY
`
`
`Fig. 5. The relationship between the cerebrospinal fluid and the intersti-
`tial fluid/brain tissue and their functional
`interaction with the blood-
`stream. Adopted from Pardridge (1991).
`
`whereas in humans the CSF is turned over every 5 h, 4—5
`times a day. Thus, the CSF fluid is constantly formed at
`the choroid plexi and subsequently drained into the
`peripheral bloodstream at the arachnoid villi. These differ—
`ences may have significant
`impact on nose to brain
`delivery studies, especially in terms of the possibility of
`the diffusion of drugs from the CSF deeper into the brain
`tissue, and should be considered when choosing an appro-
`priate animal model.
`The diffusion of drugs from the surface of the brain into
`the brain tissue is slow with the rate of diffusion, expressed
`as diffusion coefficient
`(D),
`inversely related to the
`molecular weight of the drug. The time it
`takes for a
`molecule to diffuse a given distance is related to the square
`of the distance (Pardridge, 1991). The time (for even a
`small molecule such as glucose (Mw 180 Da, 3.5 A size))
`to diffuse 5 mm is about 11.7 h with a D of 6X10_6
`cmz/s. For a protein such as albumin (MW 68 kDa, 50 A
`size) the time is 4.2 days with a D of 0.7x 10_6 cmZ/s.
`According to Pardridge (1991), the distinct difference in
`CSF bulk flow properties and the diffusional flow rates of
`drugs in brain tissue (and ISF) creates a functional barrier
`between the CSF and the cells of the brain tissue to include
`
`the ISF. This prevents complete equilibration between the
`two fluid compartments and a significant drug concen—
`tration difference exists between CSF and brain ISF. A
`
`cartoon of these two central extracellular compartments of
`the brain and their functional interaction with the blood-
`
`stream is given in Fig. 5. Hence, although no anatomical
`barrier exists between the CSF and the brain it can be
`
`concluded that a drug administered nasally which success-
`fully reaches the CSF (and available drug receptors at this
`site) cannot automatically be considered to distribute
`further into the brain parenchyma.
`
`4. Transport pathways from nose to brain
`
`The different routes by which a drug delivered nasally
`can reach the CSF and the brain are shown schematically
`
`
`
`
`
`I
`/ Brain Tis
`
`Olfactory Reg
`
`
`
`
`/
`
`Clearance 1:
`Elimination
`
`CSF
`
`Fig, 6. The nose to brain transport routes,
`
`in Fig. 6, where the thickness of the arrows indicates the
`likelihood of drugs exploiting the route in question. When
`drugs are administered nasally the drug will normally be
`rapidly cleared by the mucociliary clearance system (Illum
`et al., 1994). Some of the drug (for lipophilic drugs up to
`100% but normally much less) will be absorbed into the
`bloodstream from where it reaches the systemic circulation
`directly and subsequently is eliminated from the blood-
`stream via normal clearance mechanisms (Hussain et al.,
`1980). The drug can reach the brain from the blood by
`crossing the blood—brain barrier (the so-called systemic
`pathway to the brain) but can also be eliminated from the
`CSF into the blood. Of particular interest to this review is
`the fact that the drug can also be absorbed from the nose
`via the olfactory region into the CSF and possibly further
`into the brain. The amount of drug absorbed or lost via the
`different pathways has been shown to be highly dependent
`upon the characteristics of the drug, especially lipophilicity
`and molecular weight, but also the drug formulation
`(Sakane et al., 1991b, 1995).
`In order for a drug to travel from the olfactory region in
`the nasal cavity to the CSF or the brain parenchyma, it has
`to transverse the nasal olfactory epithelium and, depending
`on the pathway followed, also the arachnoid membrane
`surrounding the subarachnoid space. In principle, one can
`envisage three different pathways across the olfactory
`epithelium; (i) transcellularly especially across the susten—
`tacular cells, most likely by receptor mediated endocytosis,
`fluid phase endocytosis or by passive diffusion, the latter
`pathway most likely for more lipophilic drugs, (ii) paracel-
`lularly through tight junctions between sustentacular cells
`or the so—called clefts between sustentacular cells and
`
`olfactory neurones, (iii) by the olfactory nerve pathway
`where the drug is
`taken up into the neuron cell by
`endocytosic or pinocytotic mechanisms and transported by
`intracellular axonal transport to the olfactory bulb.
`One of the first
`indications of the existence of the
`
`olfactory nerve pathway into the nasal cavity for non-
`microbial or non-viral agents administered was published
`by Faber (1937) who administered various materials such
`as potassium ferrocyanide and iron ammonium citrate
`nasally to rabbits. However, the mechanism of transport
`
`AQUESTIVE EXHIBIT 1015 Page 0006
`
`

`

`L. Illum / European Journal of Pharmaceutical Sciences I I (2000) 1—18
`
`7
`
`was not fully elucidated. De Lorenzo (1970) placed ‘98Au—
`labelled gold particles intranasally in squirrel monkeys.
`The particles were traced with electron microscopy and
`after 15 min were found to have travelled to the tips of the
`olfactory receptors. They reached the olfactory bulb in
`30—60 min. The rate of progression in the olfactory neuron
`was estimated as 2.5 m/ h. This was confirmed by
`Czerniawska (1970) who injected rabbits with colloidal
`198Au under the mucosa of the olfactory region. The gold
`was found to reach the CSF within 1 to 2 h. Gopinath et a1.
`(1978) later found in rhesus monkeys, that similar gold
`particles not only entered the olfactory rods but also the
`sustentacular cells. In the sustentacular cells the particles
`were found either as discrete particles or aggregated into
`electron-dense masses
`in the mitochondria. The gold
`particles were also found in the endothelial cells of the
`blood vessels in the lamina propria. Such particles were
`thought
`to originate from the base of the sustentacular
`cells. Gold particles were not seen in the intercellular
`spaces (tight junctions or clefts) of the olfactory mucosa.
`The Lancet reported an experiment performed by Perl and
`Good (1987) where rabbits were exposed nasally to
`aluminium lactate in solution. These researchers found that
`
`the aluminium lactate travelled to the olfactory bulb and
`the cerebral cortex where it was found inside granulomas
`of macrophages, lymphocytes and occasional plasma cells.
`A possible association between the occurrence of Alzheim—
`er’s disease and the inhalation of aluminosilicates present
`in polluted air was suggested.
`Shipley (1985) confirmed the existence of a neuronal
`pathway from nose to brain by administering gelfoam
`implants soaked in wheat germ agglutinin—horseradish
`peroxidase (WGA—HRP) solution to the nose of rats. The
`WGA—HRP was found to have travelled a substantial
`
`distance within the brain, with labelled neurons being
`visualised by light microscopy in the raphe nuclei in the
`midbrain and pons and throughout the entire expanse of
`the olfactory cortex to the caudal pole of the cerebral
`hemisphere. Broadwell and Balin (1985) performed similar
`experiments in mice and squirrel monkeys with HRP and
`WGA—HRP and observed the HRP reaction product in the
`superficial fiber layer of the main olfactory bulb bilaterally
`within 45 to 60 min. The coloration product was evident
`throughout
`the intercellular clefts in the olfactory epi-
`thelium. At a much later time the olfactory fiber and
`glomerular layers of the olfactory bulbs were also labelled
`densely with the reaction product. However,
`this was
`attributed to the slower anterograde axoplasmic transport
`in the olfactory sensory neurones. An identical axonal
`transport was seen for the WGA—HRP. Similar results
`were shown by Itaya (1987) who observed secondary order
`neuron labelling of the accessory olfactory bulb and by
`Balin et a1. (1986).
`Apart from a study by Hastings and Evans (1991), who
`evaluated the transport of cadmium from the nasal cavity
`to the brain by the olfactory neurones, Thome et a1. (1995)
`
`were the first to determine quantitatively the intraneural
`transport of HRP and WGA—HRP after nasal administra-
`tion. These authors measured the quantity of protein
`accumulated in the olfactory bulb 48 h after nasal or
`intravenous administration in the rat model. They found
`that WGA—HRP given nasally as a 1.0% solution resulted
`in a concentration of 140 nM in the olfactory bulb tissue
`whereas HRP given in the same concentration by the same
`route and WGA—HRP given in the same concentration by
`an intravenous injection only resulted in background levels
`in the olfactory bulb. This showed that WGA—HRP was
`not able to penetrate significantly the blood—brain barrier
`to enter the olfactory bulb parenchyma. It also indicated
`that the transport pathway of HRP may be different to that
`of WGA—HRP. The authors suggested that the WGA—HRP
`was taken up into the neuron by adsorptive or receptor-
`mediated endocytosis and transferred in organelles associ-
`ated with the Golgi system within the cell to the axon
`terminal
`for
`release at
`the first order synapse in the
`glomeruli of the olfactory bulb. It was also suggested that
`HRP could have exploited two potential pathways due to
`the lack of binding sites on the plasmalemma; namely fluid
`phase endocytosis into the neuron followed by degradation
`in lysosomes (a pathway that was also suggested by
`Broadwell and Balin (1985)), or transport through the open
`intercellular clefts. However, the authors considered the
`availability of clefts large enough to let through HRP, to be
`limited in number.
`
`It has been shown that the transneuronal pathway is very
`slow and that agents reach the CNS as late as 24 h after
`administration in the nasal cavity (Kristensson and Olsson,
`1971). Hence, the transneuronal pathway cannot explain
`the rapid appearance of drug in the CSF and the brain seen
`for a range of small molecular weight drugs such as
`dihydroergotamine (Wang et al., 1998), cocaine (Chow et
`al., 1999), lignocaine (Chou and Donovan, 1998b), antihis-
`tamines
`(Chou and Donovan, 1997) and cephalexin
`(Sakane et al., 1991a).
`Faber (1938) very early observed that if Prussian blue
`was placed in the nasal cavity of animals it could later be
`observed in the perineural spaces of the olfactory nerve
`and in the subarachnoid space (in the CSF) of the brain,
`apart from also appearing in the nasal lymph vessels and
`the cervical lymph nodes. This extracellular transport of
`agents would involve transport via patent intercellular tight
`junctions or clefts in the epithelium into the CSF and
`would rely on a direct anatomic connection between the
`submucosa and the subarachnoid extensions, the perineural
`space, surrounding the olfactory nerves as they penetrate
`the cribriform plate. The agent
`is thought
`to enter the
`perineural space, either due to the perineural epithelium
`surrounding the olfactory axon being loosely adherent
`(‘open-cuff’ model), or to enter through the epithelium cell
`junction if the perineural epithelium is closely adhered to
`the axon (‘closed—cuff’ model) (Jackson et al., 1979). The
`extracellular transport pathway is considered to be very
`
`AQUESTIVE EXHIBIT 1015 Page 0007
`
`

`

`8
`
`L. Ilium / European Journal of Pharmaceutical Sciences I I (2000) 1—18
`
`fast and possibly accounts for most of the rapid transport
`of small molecular weight drugs from the nasal cavity into
`the CSF (Frey et al., 1997).
`Frey et a1. (1997) and Chen et a1. (1998) also showed
`that
`larger molecular weight drugs, such as the protein
`nerve growth factor (MW 37 kDa), could be transported
`rapidly to the CNS in a rat model. Hence, after intranasal
`administration of 125I labelled nerve growth factor ([251-
`NGF), the radiolabeled drug (less than 20% free label)
`appeared within 20 min in the olfactory bulb and to a
`lesser extent in other parts of the brain. The concentration
`of radiolabel in the blood was similar for both intranasal
`and intravenous administration studies but the radiolabel in
`
`the CNS, after intravenous administration, was only 0.001
`nM as compared to up to 2.0 nM after nasal administra-
`tion. The authors suggested that the rapid appearance of
`the drug in the olfactory bulb, the cerebrum and the brain
`stem was consistent with the drug being transported via
`intercellular clefts in the olfactory epithelium or extracellu-
`larly along the neurones to reach the CSF and the brain.
`The fact that no receptors for NGF are found on primary
`olfactory neurons supports the suggestion that the transport
`is not intracellular via the axonal pathway. The efficiency
`of the NGF delivery to the brain was found to be much
`less than that found by Sakane et al. (1991a) for the low
`molecular weight drug, cephalexin (Mw 401 Da) of 13 uM
`in the CSF. This could most likely be due to limitations in
`the diameter of the open clefts or tight junctions.
`It can be concluded that
`the pathway employed for
`delivery of a pa