Drug Delivery via the Mucous Membranes of the Oral Cavity

JOURNAL OF PHARMACEUTICAL SCIENCES

January 1992 Volume 81, Number 1

A publication of the American Phermeceutkal Association

REVIEW ARTICLE

Drug Delivery via the Mucous Membranes of the Oral Cavity DAVIDHARRIS** AND JOSEPH R. ROBIN SON'^ Received December 3,1990,from the ‘sdrool of Pharma University of Wisconsin-Madison,425 North Charter Street, Madison, WI 53706. Accepted for publication March 16, 1991. * Present ad%&: Schering-PloughResearch, 2000 Galloping Hill Road, Kenilworth, NJ 07033.

Abstract 0 The delivery of drugs via the mucous membranes lining the oral cavity (i.e., sublingual and buccal), with consideration of both systemic delivery and local therapy, is reviewed in this paper. The structure and composition of the mucosae at different sites in the oral cavity, factors affecting mucosal permeability, penetration enhancement, selection of appropriate experimental systems for studying mucosal permeability,and formulationfactors relevant to the design of systems for oral mucosal delivery are discussed. Sublingual delivery gives rapid absorption and good bioavailability for some small permeants, although this site is not well suited to sustained-delivery systems. The buccal mucosa, by comparison, is considerably less permeable, but is probably better suited to the development of sustained-delivery systems. For these reasons, the buccal mucosa may have potential for delivering some of the growing number of peptide drugs, particularly those of low molecular weight, high potency, andlor long biological half-life. Development of safe and effective penetration enhancers will further expand the utility of this route. Local delivery is a relatively poorly studied area; in general, it is governed by many of the same considerations that apply to systemic delivery.

Absorption of drug via the mucous membranes of the oral cavity was noted a s early as 1847 by Sobrero, the discoverer of nitroglycerin,l and systematic studies of oral cavity absorption were first reported by Walton in 1935 and 1944.293 Since then, reviews of the subject have been provided by Katz and Barr in 1955,4 Gibaldi and Kanig in 1965,b and Squier and Johnson in 1975.6 Drug delivery via the membranes of the oral cavity can be subdivided as follows: (1)sublingual delivery, which is the administration of drug via the sublingual mucosa (the mem-

brane of the ventral surface of the tongue and the floor of the mouth) to the systemic circulation; (2) buccal delivery, which is the administration of drug via the buccal mucosa (the lining ofthe cheek) to the systemic circulation; and (3) local delivery, for the treatment of conditions of the oral cavity, principally aphthous ulcers, fungal conditions, and periodontal disease. These oral mucosal sites differ greatly from one another, in terms of anatomy, permeability to an applied drug, and their ability to retain a delivery system for a desired length of time. It is not surprising, therefore, that these sites have received different treatments from researchers investigating oral mucosal drug delivery. The sublingual route is by far the most widely studied of these routes. The sublingual mucosa is relatively permeable, giving rapid absorption and acceptable bioavailabilities of many drugs, and is convenient, accessible, and generally well accepted. In fact, Paulson7 suggested as much almost 75 years ago when, in the context of a survey of alternate delivery routes, he wrote: “May I ask anyone reading this, to be good enough to refresh his memory, by looking in the nearest mirror, at the sublingual space? Note the thin membrane, the considerable area, and the large veins, denoting a free flow and return of blood. The space is always smooth, never firred like the tongue, never shielded by mucus, as the stomach always is when in active rebellion, or, may be, masked by half-digested food.” This route has been investigated clinically for the delivery of a substantial number of drugs; it is the route of choice for the administration of nitroglycerin and is also used clinically for the delivery of buprenorphine and nifedipine (see Table I).

Table CCllnical Appllcatlons of Systemlc Delivery via the Oral Mucosae

Mucosa

~~

Sublingual Buccal a

Brand Narne/Manufacturer

Drug ~

Nitroglycerin Buprenorphine Nifedipine Methyltestosterone ProchlorDerazine

Comment

~

Nitrostat, Parke-Davis; etc. Temgesic,’ Reckitt 8 Coiman Adalat, Bayer; Procardia, Pfizer Metandren, Ciba; Oreton Methyl, Schenng Buccastem: Reckitt 8 Colman

Delivery route of choice Alternative to injectable form Gives more rapid onset of action than enteral form Avoids first-pass hepatic metabolism Alternative to enteral tablet

Not available in U.S.A.

0022-3549/92/0 100-OOO1$02.50/0 6 1992, American Pharmaceutical Association

Journal of Pharmaceutical SciencesI 1 Vol. 81, No. 1, January 1992

The buccal mucosa is considerably less permeable than the sublingual area, and is generally not able to provide the rapid absorption and good bioavailabilities seen with sublingual administration. Twenty-five years ago, the buccal route was quite widely used for the administration of steroid analogues,6 since it avoided first-pass hepatic metabolism which limited the efficacy of these drugs when given via the gastrointestinal route. Most of these products have since been superseded by the development of more resistant analogues, although a few remain on the market (see Table I). The current resurgence of interest in the buccal mucosa as a route for systemic drug delivery stems from very different considerations. It has been suggested that, since drug absorbed directly from the oral cavity avoids enzyme- or acid-mediated degradation in the gastrointestinal tract and first-pass metabolism by the liver, buccal administration may be of value in the delivery of some of the growing number of peptide drugs. Local delivery t o tissues of the oral cavity has a number of applications, including the treatment of toothache, periodontal disease, bacterial and fungal infections, and aphthous and dental stomatitis, and in facilitating tooth movement with prostaglandins.&12 Conventional formulations for local oral delivery are principally lozenges and mouthwashes that give high drug levels in the oral cavity, but for only a short time. Although it is feasible to design delivery systems that improve localization or retention of the dose, few of these have seen commercial success. The principles underlying the design of improved delivery systems such as these are broadly similar to those pertaining to oral mucosal systems for systemic delivery.

Composition of the Oral Mucosae For a fuller account of the structures and functions of the human oral mucosae, the reader is referred to Meyer et al.13 The following is a summary of the most important points, as they pertain to oral mucosal drug delivery. The oral mucosa consists of an outermost layer of stratified squamous epithelium, below which lies a basement membrane, and below this, in turn, a lamina propria and submucosa (see Figure 1). Epithelium-Oral epithelium is broadly similar to stratified squamous epithelia found elsewhere in the body in that it consists of a mitotically active basal cell layer, progressing through a number of differentiating intermediate layers to the superficial layers, where cells are shed from the surface of the epithelium. The buccal epithelium is composed of -40-50 cell layers, while the sublingual epithelia contain somewhat fewer. As cells mature and migrate from the basal layer towards the epithelial surface, they increase in size and become progressively flattened, while showing increasing levels of protein tonofilaments and declining levels of most other cytoplasmic organelles. The turnover time for the buccal epithelium has been estimated at &6 days, and this is probably representative of the oral mucosa as a whole. The thickness of the oral epithelium varies considerably between sites: in humans, dog, and rabbit, the buccal mucosa measures 500-800 pm in thickness, while the hard and soft palates, floor of the mouth, ventral tongue, and gingivae measure -100-200 pm.”16 The composition of the epithelium also varies with its location in the oral cavity. Thus, mucosae of the gingiva and hard palate (areas subject to mechanical stress) are keratinized in a similar manner to the epidermis. Keratin is laid down in the superficial cells of the epithelium, and these cells become flattened in shape and virtually devoid of organelles. The mucosae of the soft palate and the sublingual buccal regions, on the other hand, are generally not keratinized. The superficial cells in these regions become less flattened and retain their nuclei and some 2 I Journal of Pharmaceutical Sciences Vol. 81, No. 1, January 1992

Lamina propria

Submucosa

t

Figure l-Generalized structure of oral mumsae.

cytoplasmic function. Biochemical Composition-A reasonable account of the biochemistry of the oral mucosae is given by Gerson and Harris.17 A notable feature of the oral mucosae is the large amount of protein present in the form of tonofilaments in the cells of all layers, in both keratinized and nonkeratinized epithelia. These tonofilaments are composed of a t least seven component proteins, termed keratins, with M, ranging from 40 to 70 kDa. Cells of “nonkeratinized” mucosae contain mostly the lower M, keratins, while in those of “keratinized mucosae, the higher M, keratins predominate.18 It thus appears that keratinization is not an “all-or-nothing” process, but that these terms simply represent extremes of a spectrum of keratinization. Comparatively little is known about the lipid composition of the oral mucosae. Furthermore, the studies that have been reported have determined lipid profiles that differ markedly from one another in certain respects.’@-21 Wertzlo suggested that the lipid compositions of the various oral epithelia and the epidermis correspond with the water permeabilities of these tissues. In particular, the keratinized oral mucosae (gingiva and palate) and the epidermis contain sizeable amounts of acylceramides and ceramides, which have been associated with barrier function, while the nonkeratinized oral mucosae (buccal and floor of the mouth), which are also more permeable, contain smaller amounts of these components. The cells of the oral epithelia are surrounded by an intercellular ground substance, the principal components of which are carbohydrate:protein complexes, some of which may be intimately associated with particular sites on the cell surfaces. It is thought that this matrix may play a role in cell-cell adhesion, as well a s acting as a lubricant, allowing cells to move relative to one another. Another aspect of the biochemical composition of the oral mucosae is the so-called “membrane-coating granules”, and their role in the biochemical changes which occur during the

maturation of the epithelium; this subject is dealt with in some detail under the heading Membrane-Coating Granules. Intercellular Junction-A number of different intercellular junctions are found in epithelia, principally gap junctions, tight junctions, and desmosomes and/or hemidesmosomes. In a gap junction, the plasma membranes of adjacent epithelial cells are separated by a gap of 2-5 nm, which is believed to be continuous with the intercellular space: a gap of this dimension would probably permit the passage of permeants of M, up to several thousand Da. These junctions are generally no >1pm in diameter and are probably confined to the basal and intermediate layers of the epithelium. In a tight junction, by comparison (macula occludens or zonula occludens),the membranes of adjacent cells appear to be fused together. These junctions are uncommon in oral epithelia, however, and appear to be confined to the more superficial cell layers, though they are not found in the very outermost layers. The remaining junctional complexes are desmosomes and hemidesmosomes, which are mechanical attachments between adjacent epithelial cells and between basal cells and the basement membrane, respectively. Basement Membrane and Connective Tissues-The basement membrane is a continuous layer of extracellular material, forming the boundary between the basal layer of the epithelium and the connective tissues of the lamina propria and submucosa. It is a trilaminar structure, consisting of an upper amorphous layer (40-80 nm in thickness; the lamina lucida), a central dense layer of similar thickness (the lamina densa), and a broader region of fibrous material below. The lamina densa is the primary structural component of the basement membrane: it is composed of a form of collagen, thought to be arranged as a highly ordered network, which would impart considerable strength to the structure. Hemidesmosomesin the membranes of the basal cells connect across the lamina lucida with the lamina densa. The functions of the basement membrane are probably twofold: (1) to provide adherence between the epithelium and the connective tissues beneath and to provide mechanical support for the epithelium, and (2) to form a barrier to the passage of cells and some large molecules across the mucosa. Below the basement membrane lies the lamina propria: this is a continuous sheet of connective tissue containing collagen and elastic fibers and cellular components in a hydrated ground substance. It also carries blood capillaries and nerve fibers that serve the mucosa. The uppermost elements of the lamina propria mesh with the anchoring fibrils of the basement membrane, and the lower layers border onto the submucosa, if present, or onto periosteum or muscle. The junction between the epithelium and the lamina propria is not flat, but is thrown into ridges and folds and drawn out into connective tissue papillae that project into the epithelium (see Figure 1). Consequently, the area of the basement membrane is greater than that of the epithelial surface: this provides a broader area for attachment of the epithelium and for metabolic exchange between the epithelium and the lamina propria.

Permeability of the Oral Mucosae The permeability characteristics of the oral mucosae have been reviewed by a number of authors, most notably by Squier and Johnson,G and Siege1.22 The permeability of the oral mucosae in general is probably intermediate between that of the epidermis and that of intestinal mucosa. Galey et al.23 estimated the permeability of the buccal mucosa to be 4-4000 times greater than that of the skin (seeTable I1 for some comparative permeabilities for different absorption sites). As may be expected from the diverse structures and functions of the various oral mucosae,

Table ICComparlson of Permeabillty Coetflclents for Various Absorption Sltes

Permeability Coefficients, cm/s

Water

20

Progesterone Glycerol Octanol Estradiol Ouabain Amphetamine

314 92 130 272 584 135

3.7x 2.6 x 8.9 x 6.0 x 2.2 x 6.6 x 6.5x 1.5 x

10-56 10-56

1.5x 10-4

-

1.8 x 4.5 X 10-56

lo-'

-

1.4 x 10-7c 4.4x 10-7e 4.2 x lo-"

-

1.4 x 10-5' 1.1 x lo-1.1 x 10-98 3.9 x 10-96

a Reference 27. bRabbit (ref 28). CReference29. dDog (ref 23). Reference 23.

there appear to be considerable differences in permeability between different regions of the oral cavity.24 In general, the permeabilities of the oral mucosae decrease in the order sublingual > buccal > palata1,25.26 although few studies have compared the permeabilities of these tissues quantitatively. This rank order, however, is in accordance with the physical characteristics of these tissues, with the sublingual mucosa being relatively thin, the buccal thicker, and the palatal intermediate in thickness, but keratinized. Experimental Systems-Squier and Johnson6 reviewed the experimental methods commonly used in the study of oral mucosal permeability and outlined the factors that should be considered in selecting an appropriate experimental system. Numerous in vitro and in vivo studies of mucosal drug absorption have been reported in the literature in recent years, utilizing a wide range of animal tissues, experimental designs, and probe molecules. To illustrate the range of species variation that exists, Table I11 summarizes the characteristics of buccal mucosa from several experimentally used species. As can be seen from Table 111, the oral epithelia of a number of experimental animals are entirely keratinized studies using these species are thus of dubious relevance from the point of view of human mucosal drug delivery. More useful models are the rabbit and dog buccal mucosae, both of which have been claimed to be broadly similar in structure and composition to human buccal mucosa.14J6.30 Some of the probe drugs used to assess permeability are nitrates,26,31.32 steroids,26,33 salicylic acid,9,34 temazepam,35 morphine and other opiates,sW8 several of the pblockers,39.40 a number of small organic molecules,41.42 and a small number of p e p tides.4-9 In broad terms, there are four experimental approaches that have been used in order to determine the permeability of the oral mucosae to drug molecules. One of the more widely used methods has been the so-called '%uccal absorption test" Table IiCCharacterlstIcs of Buccal Mucoaa from Various Species

Species Human Rabbit

Thickness, f l

Keratinizatione

500-600e

-

6006

-

Dog

770'

Pig

770d

Rat Hamster

-

+ +

-

Investigations Using this Species (References) 25,34,50. 51 28,52-55 30,45,56 46,57 56,58-60 28,39,61

a Reference 14. Reference 16.'References 15 and 24. Reference 15. Keratinizationis indicated by "+", and no keratinization is indicated by "-".

Journal of Pharmaceutical Sciences I 3 Vol. 81, No. 1, January 1992

least 3 h. devised by Beckett and co-workers. In this method, a solution of the permeant is swirled around the mouth by a human A further complication in studying the permeability of subject for a defined length of time, after which the solution excised sublingual mucosa, in particular, is that numerous is expelled and assayed, and disappearance of the permeant is ducts from the submandibular and sublingual salivary glands calculated. Beckett and Hossie61 reviewed studies utilizing open to the mucosal surface in this region; it may therefore be this method. Tucker62 recently reported a more sophisticated difficult to obtain a sufficiently large expanse of mucosa which variant of this method. The extent of absorption may also be is not perforated by these ducts. determined by blood sampling or by quantification of a Recently, it has been proposed that cell or tissue cultures be pharmacologic response, although these variations have not used in the study of buccal permeability.73 The chief limitabeen widely used. While these experiments are relatively tion of this approach is that these cultures generally do not straightforward to perform, they are limited by the sensitivity show the patterns of cell differentiation that are observed in or accuracy of the methods available to estimate absorption. intact tissue. However, as tissue culture techniques develop, Measurement of disappearance is accurate only where the it seems likely that such approaches will prove useful in percent absorbed is substantial, and, obviously,plasma levels tackling some of the observed transport phenomena a t the are frequently only measurable for well-absorbed compounds. cellular and subcellular levels. Estimation of absorption from pharmacologic data requires Permeability Barrier-The permeability barrier of the detailed knowledge of the pharmacodynamics of the drug. The oral mucosae is thought to reside with the superficial layers principal drawback of this method, from a mechanistic un(approximately the outermost quarter) of the epithelium. derstanding point of view, however, is that the observed Permeation studies have been performed using a number of absorption is a function of the relative areas and permeabiltracers, such as horseradish peroxidase (an enzyme of M, ities of all the oral mucosae that were in contact with the 40 000 and 5-6 nm in si~e)16~24~74.75 and lanthanum nitrate (an solution. electron-dense marker, 2 nm in sizel.76.77 When applied to the A second approach, which overcomes the above-mentioned outer surface of the epithelium, these tracers are seen to limitation, involves the use of an in situ perfusion cell, or penetrate only through the outermost layer or two of cells. similar device, that is clamped or otherwise attached to the When applied to the submucosal surface, they appear to intact mucosa.50B56,63,64 In this way, absorption can be limited permeate up to, but not into, the uppermost cell layers (approximately the top 2625%) of the epithelium.75.76.78This to a particular region of the oral mucosae, the surface area of which can be precisely determined. Once again, absorption is is consistent with the idea that flattened surface cell layers assessed by measuring disappearance from the perfusate, present the major barrier to permeation, while the more appearance in the bloodstream, or a pharmacologic response; isodiametric lower cell layers are relatively permeable. These thus, this method is subject to some of the same limitations as patterns of penetration are seen in keratinized and nonkerathe preceding method. tinized epithelia alike, and it is thought unlikely that keraThe third approach, which really encompasses a number of tinization per se plays a major role in barrier function.16 experimental variations, involves the use of a range of Some workers have suggested that the basement memdelivery systems, including ointments and gels,68 tabbrane is the functional permeation barrier of the oral mucoleta,31-65.% adhesive patChes,34.44.66 and drug-impregnated sae79.m or that it presents a t least a degree of resistance to pads.43 Absorption is assessed by the same methods as used in permeants.81 Studies with horseradish peroxidase and lanthe preceding approaches. The main limitation of these thanum, however, have shown it to be easily traversed by systems is that certain experimental parameters, such as these tracers, while the outer layers of the epithelium appear contact area and drug concentration at the mucosa, are often impermeable to both. These results indicate that, although unknown. It is thus difficult to draw meaningful conclusions the basement membrane may present some resistance to from these studies, beyond simple estimates of percentage of permeation, the outer epithelium is the rate-limiting step of dose absorbed. mucosal penetration.la.75.76.82Other attempts to locate the The fourth main approach is to study the permeability of permeability barrier have made use of a number of stripping excised tissue in an in vitro diffusion apparat~s.28~69.677-71 This techniques: these involve the removal of successive cell layers method has been used extensively in the study of transdermal using adhesive tape69974 or by dermatoming.80 The main permeation, but less widely for studies of oral mucosal criticism of these methods is that it is not known whether the permeation. This method, too, suffers from a number of observed changes in permeability are due solely to the drawbacks. While it is assumed that permeation from the removal of these cell layers or whether the effects of trauma donor (mucosal) to the receiver (serosal) compartment of a or of the treatment also contribute. diffusion cell is representative of permeation from the mucoThe lamina propria is not generally thought to present a sal surface to the vascular bed in vivo, this assumption has barrier to permeation. Its structure is insufficiently dense to not, to the knowledge of the authors, been fully tested. exclude even relatively large molecules, and its hydrated Another issue is that of tissue viability during such studies; matrix should facilitate the passage of hydrophilic peneone must always bear in mind the possibility of changes trants. occurring in the tissue postmortem which might affect perMembrane-Coating Granules-The most plausible explameability. Some studies carried out in our laboratorie~~~72nation at present for the origin of the permeability barrier of addressed this issue in rabbit buccal mucosa, using ATP oral epithelia involves the so-called “membrane-coating granlevels, microscopic methods, and linearity of transport data as ules” (MCG),which are spherical or oval organelles, 10&300 indicators of tissue viability. The ATP levels were found to nm in diameter, and are found in the intermediate cell layers fall to -70% of control values after 3 h of incubation in a of many stratified epithelia.as-86 These organelles were first diffusion cell, and to -60% of control after 6 h. Light and described in detail by Odland in epidermis86 (hence, their electron microscopy studies showed no apparent morphologalternative designation “Odland bodies”),and they have since ical changes after 3 h, although cell separation and widening been widely reported in both keratinized and nonkeratinized of intercellular spaces began to appear after 6 h. Finally, epithelia8749 and epidermis.f’0391The term MCG was first transport data for thyrotropin releasing hormone (TRH), a coined in 1965 by Matoltsy and Parakka1,gO who suggested tripeptide, appeared to be linear to -6 h. On the basis of this that they were involved in the thickening of the plasma evidence, it was concluded that excised rabbit buccal mucosa membrane, a process which occurs a t about the same level. It could remain viable and provide reliable transport data for at is now generally accepted that this process (probably due to 4 I Journal of Pharmaceutical Sciences Vol. 81, No. 1, Januaty 1992

de sition of protein on the cytoplasmic aspect of the distal cep" 1 membrane) is independent of the MCG, but the name has persisted. The MCG first appear at around the midpoint of the epithelium, concentrated close to the distal cell membrane: in the third quarter of the epithelium they appear to fuse with the cell membrane, the membranes of the MCG becoming incorporated into the cell membrane, and their contents being discharged into the intercellular space.92 In keratinized epithelia, MCG are generally lamellated in appearance: they contain polar lipids (glycolipids and phospholipids),glycoprotein, and a considerable number of hydrolytic enzymes.03~~ It is thought that these polar lipids are precursors of the nonpolar lipids which constitute the permeability barrier of keratinized epithelia. The enzymes present within the MCG may be involved in these conversion processes; alternatively, or perhaps additionally, they may be responsible for breaking down the desmosomal junctions between the cells of the outermost layers, leading to desquamation.92~06~~ Considerably less is known about the MCG of nonkeratinized epithelia, however. The MCG in these tissues are almost exclusively nonlamellated in appearance,84 and they generally do not show the staining reactions for glycoproteins and/or glycolipids that are seen in keratinized epithelia.87.89 Otherwise, they show similar patterns of morphology and distribution to the MCG of keratinized epithelia, and they likewise possess a number of hydrolytic enzymes.95There is thus good circumstantial evidence for a role of MCG in the formation of the permeability barrier of the nonkeratinized oral mucosa. Since the MCG are thought to be involved in the development of the intercellular matrix, it follows that interference with the MCG themselves or with their released contents may influence the permeability of the mucosa, and of the paracellular route in particular. This strategy was adopted by Squier and others47.97who attempted to increase mucosal permeability by the use of glycoprotein- or glycolipid-specificenzymes (see Penetration Enhancement). However, the scope for such strategies is likely to remain limited without a reasonable understanding of the role of MCG in nonkeratinized epithelial permeability; since these organelles lack any characteristic staining reactions or microscopic features, this level of understanding remains elusive. Nature of Permeant-Most published studies of oral mucosal permeability have focused on relatively small numbers of permeant molecules, and few have attempted to systematically evaluate the relative contributions of factors such as M,,lipid solubility, and ionization on permeability. From the systematic studies that have been reported, however, it appears that the ability of a molecule to permeate through the mucosa can be related to molecular size, lipid solubility, and ionization. Molecular Siz-For hydrophilic substances, the rate of absorption is a function of the molecular size. Small molecules (<75-100 Da) appear to cross the mucosa rapidly, but permeability falls off rapidly as molecular size increases.67,98,* Since permeability has been observed to decrease sharply as molar volume is increased beyond 80 mUmol, investigators have proposed two distinct polar routes, one of which has an upper size limit. The evidence for this is tenuous, however, and it may be more logical to expect a more continuous loss of permeability a s molecular size increases. This relationship between size and permeability has not been demonstrated for lipophilic substances, although common sense suggests that such a relationship must exist. Lipid Solubility-For any series of un-ionizable compounds, their relative permeabilities are functions of their oil-water partition coefficients, with the more lipid-soluble compounds having higher permeabilities.61,1~102

Ionization-The degree of ionization of a permeant is a function ofboth its pK, and the pH at the mucosal surface. For many weak acids and weak bases, only the un-ionized form possesses appreciable lipid solubility. The absorption of many compounds has been shown to be maximal a t the pH at which they are mostly un-ionized, tailing off as the degree of ionization increases.36.40.~0.~1 Other studies, however, have failed to show this pattern.= It is important that the possible mechanisms by which this pH effect operates are fully appreciated. When applied to a simple membrane, this principle suggests that a change in pH alters the ionization of the probe, and thus its hydrophilicity and/or lipophilicity, as it traverses the membrane. In a complex tissue such as the oral epithelium, however, it is unlikely that a pH a t the mucosal surface will be the same throughout the length of the permeation pathway (i.e., the thickness of the mucosa). It is more likely that altering pH so as to favor formation of the un-ionized form in the oral cavity simply causes an enhanced partitioning of un-ionized drug into the surface layers of the mucosa. (This is borne out by work of Henry et al.103 who have shown that, while a pH favoring the un-ionized form causes increased drug uptake from the oral cavity, much of this "absorbed" drug can be rapidly recovered by bathing the oral mucosae with a solution of pH favoring the ionized form.) It is expected that this high local concentration of drug then presents an elevated concentration gradient, thus enhancing diffusion of the (largely ionized) drug across the mucosa. In common with drug transport across other epithelia, there are a number of possible permeation pathways across the oral mucosae.99 The classical distinction is between transcellular and paracellular permeation, referring to passage across the individual cells of the epithelium and passage between these cells, respectively. For transcellular permeation, the permeant must be capable of passing through pores in the cell membranes, or diffusing through the lipid bilayers of these membranes. Passage through membrane pores would probably be limited to small molecules, while diffusion across cell membranes would require appreciable aqueous and lipid solubilities. Paracellular permeation requires the epithelium to have a sufficiently open matrix and requires the permeant to have an appreciable difisivity in the intercellular milieu. It seems likely that large and/or highly polar permeants may be unable to pass through the epithelial cell membranes and might, therefore, follow the paracellular route. An alternative classification is into polar and nonpolar routes, the former involving passage of water-soluble substances through aqueous channels in the mucosa (either aqueous pores in the plasma membranes of the individual epithelial cells or hydrophilic channels in the intercellular spaces of the epithelium), and the latter involving partitioning of the drug into the lipid bilayer of the plasma membrane or into the lipid of the intercellular matrix and diffusion through these lipid elements. Almost all studies have shown that, for most permeants, passage across the oral mucosae appears to be a first-order simple diffusion process. It has also been suggested, however, that the oral mucosae contain active or carrier-mediated systems for small molecules such as monosaccharides~0~J~~ and amino acids.62 However, these processes have not been fully characterized in terms of location, transport capacity, or specificity. The kinetics of oral mucosal absorption have been studied by a number of workers.106,107 Some investigations have shown a slow onset of appearance of permeant in the systemic circulation and a depot-like behavior of the oral mucosae, which have been attributed to some form of binding within the mucosae.1o8Jo9 To date, however, this area has not been systematically investigated and remains for the most part Journal of Pharmaceutical Sciences I 5 Vol. 81, No. 1, January 7992

poorly understood. Permeability Coefficient-When surveying the large number of oral mucosal permeability studies reported in the literature, it is apparent that, while many of these studies are of interest, they are generally so different in terms of experimental design that meaningful comparisons between studies are virtually impossible. It is therefore difficult to draw any definitive conclusions from these studies as to the influence of such factors as molecular size, hydrophilicity and/or lipophilicity, and ionization on mucosal permeability. This problem is compounded by the lack of a single, generally accepted measure of permeability, one that allows comparisons between experimentally diverse situations. One such parameter is the permeability coefficient,one form of which is shown in eq 1:

P=

%permeated x

A-t

X

vd

100

where P is the permeability coefficient(cds),Vdis the volume of donor compartment, A is the surface area for permeation, and t is time. Equation 1 assumes that the concentration gradient of permeant across the mucosa remains constant with time. This assumption is valid as long as the percent absorbed remains small (-<&lo%). Only a handful of investigators have determined this parameter. Some examples of the calculated permeabilities are presented in Table IV. It is not possible to derive permeability coefficients from many published studies of oral mucosal permeability since, in general, the experimental designs did not allow variation of the necessary parameters to be quantified. Peptide Delivery-A recent survey has shown that there are currently >150 recombinant protein drugs undergoing clinical trial,112 and it seems likely that such drugs will become an increasingly important part of our clinical armaTable IV-Permeablllty Coefflclents (9In Oral Mucosae

Probe

Mucosa

Tritiated water Glycerol Octanol Progesterone Glutamic acid Lysine Serine Glycine Leucine Benzylamine Salicylic acid

Rabbit buccal Rabbit buccal Rabbit buccal Rabbit buccal Rabbit buccal Rabbit buccal Rabbit buccal Rabbit buccal Rabbit buccal Rabbit buccal Rabbit buccal TRH Rabbit buccal Dextran 4000 Rabbit buccal DGAVP Pig buccal lsoproterenol Dog buccal Tritiated water Dog buccal Tritiated water Dog buccal Estradiol Dog buccal Amphetamine Dog buccal Ouabain Dog buccal Tritiated water Hamster cheek pouch Butanol Hamster cheek pouch Benzoic acid Hamster cheek pouch Rat sublingual Urea Glucose Rat sublingual Glycerol Rat sublingual Rat sublingual Butanol Urea Rabbit sublingual Rabbit sublingual Sucrose Dextran 20000 Rabbit sublingual 6 ,! Journal of Pharmaceutical Sciences Vol. 81, No. 1, January 1992

P, cm/s

3.7x 6.0x 2.2 x 8.9 x 4.0x 2.3 x 1.3 X 8.3x 1.9x 1.5 x 9.3x 2 x 2.2x 1.1 x 6 X 5.0 x 2.6 x 6.6 X 1.5x 6.5 x 2.9 x 4.3 x 4.6x 2.2x 4.8x 1.1 x 2.9 x 8.4x 8 x 4 x

10-5 10-7 10-5

lo-'

10-7 10-7 10-7 10-7 10-5 10-7 10-7 10-9 10-8 10-5 10-5 10-5

lo-'

10-5 10-5 10-5

10-6 10-7 10-6 10-5 10--7

Reference 28 28 28 28 52 52 52 52 52 52 52 54

53 71 63 28 23 23 23 23 110 56 110 98 41 98 98 111 111 111

mentarium. It is probable that the mucosal routes in general will play significant roles in the delivery of these peptide drugs, since mucosal absorption avoids the problems of acidand enzyme-mediated degradation in the GI tract and hepatic first-pass metabolism. To date, however, comparatively few studies have looked specifically a t peptide permeation across the oral mucosae. The work that has been reported is summarized in Table V. These studies have shown bioavailabilities of only a few percent or less when compared with parenteral routes, and other studies have shown no measurable penetration at all. These results are not altogether surprising, given that these substances have low lipid solubilities, and their permeation via the polar route is likely to be limited by their relatively large molecular weights. It is worth pointing out that the bioavailability determined for buccal delivery of TRH in humans, presented in Table V, is not inconsistent with the permeability coefficients estimated for this drug by other workers in rabbit (Table IV). Not only is membrane permeability a serious barrier to peptide absorption, it has also been reported that the oral mucosae, in common with most other mucosae, show substantial levels of esterase and peptidase activity.116118Depending on the animal species and substrates used, buccal homogenates have shown enzyme activities between a few and several hundred percent of the activities of intestinal homogenates. These results raise the possibility of significant enzymatic degradation of the drug before it reaches the systemic circulation. Studies in these laboratories (unpublished data) have shown that peptide degradation during in vitro buccal permeation studies can be substantial. It should be noted, however, that determination of enzyme activity in tissue homogenates does not necessarily reflect enzyme activities in vivo, as many of these enzymes may be contained within lysosomes or otherwise sequestered within the intact cell. Penetration Enhancement-While the sublingual mucosa is sufliciently permeable to allow the therapeutic delivery of a number of small drug molecules, low mucosal permeability is perceived to be a significant obstacle to buccal delivery, particularly in the context of buccal delivery of peptide drugs. In many instances, the buccal absorption rates achieved for such permeanta have been low: absorption of a t most a few percent of an applied dose over a realistic period of time is typical. Attention is thus focused on some of the strategies that have been proposed for enhancing the permeability of the oral mucosae. A considerable number of agents have been proposed as penetration enhancers, mostly in the context of transdermal drug delivery (the reader is directed to reviews by Gibaldi and Feldmanllg and by Walters120). The agents used have mostly been small hydrophilic molecules [e.g., dimethyl sulfoxide (DMSO), dimethylformamide, ethanol, propylene glycol, and the 2-pyrrolidonesl, long-chain amphipathic molecules (decylmethyl sulfoxide, Azone, sodium laurylsulfate, oleic acid, and the bile salts), and nonionic surfactants (polysorbates). Table V-Estlmated Buccal Avallabllltles of Some Peptide Drugs Relatlve to Intravenous or Intramuscular Admlnlstratlon

Peptide

Bioavailability

TRH

1-5%'

<1o%8

Oxytocin Insulin

0.5-27Yob

Streptokinase and/or streptodornase

(no penetration without enhancer) Clinical effect observed: availability not determinede

~~

Human. Rat and dog.

Reference 43,44 113,114 10,45,46, 115 49

These have generally been employed empirically. Although some are effective, either alone or in combination, their modes of action are not fully understood. It has been proposed, again in the context of transdermal delivery, that these agents interact at either the polar head groups or the hydrophilic tail regions of the molecules comprising the lipid bilayer.121.122 Interaction at these sites will probably have the effect of disrupting the packing of the lipid molecules, increasing the fluidity of the bilayer, and facilitating drug diffusion. Interaction of an enhancer with the polar head groups may also cause or permit the hydrophilic regions of adjacent bilayers to take up more water and move apart, thus opening the paracellular pathway. In addition to these effects, certain enhancers may have direct effects on the bulk properties of the aqueous regions of the skin. Agents such as DMSO, polyethylene glycol, and ethanol can, if present in sufficiently high concentrations in the delivery vehicle, enter the aqueous phase of the stratum corneum and alter its solubilizing properties, thereby enhancing the partitioning of drug from the vehicle into the skin. With the growing interest in some of the alternate delivery routes (e.g., buccal, nasal, ocular), penetration enhancement is becoming more widely studied in these mucosal tissues. To a large extent, these studies have used the same agents as have been used in transdermal delivery, notwithstanding the substantial differences that exist between the skin and these mucosal tissues in terms of hydration, biochemical composition, and nature of the permeability barrier. Relatively few enhancement studies have been carried out in the buccal mucosa. The few that have been reported used the same agents as have been used transdermally and have shown similar results to the transdermal studies. Agents such as Azone, ionic and nonionic surfactants, and the bile salts have increased buccal permeability to such probes as insulin, dextrans, and small hydrophilic and lipophilic permeants.lo,4647.80.123.124These agents produce varying degrees of enhancement, depending on the characteristics of the permeant, the composition of the delivery vehicle, whether the tissue was pre-treated with enhancer, and other factors. As with transdermal enhancement, the reasons for these effects are frequently not fully understood. A different approach was adopted by Squier,97 and subsequently by Aungst and Rogers,47 who investigated the effects of phospholipase, hyaluronidase, neuraminidase, and chondroitinase enzymes on the buccal permeability to horseradish peroxidase in vitro. Squier showed that chondroitinase increased permeability without causing undue tissue damage, although the other three enzymes caused varying degrees of tissue damage, but showed little effect on permeability. It was suggested that chondroitinase was acting on glycoprotein or glycolipid constituents of the permeability barrier of the rnucosa. An important issue which must also be considered is the toxicity of penetration enhancers. The line between effective enhancement and tissue damage is necessarily fine, since enhancement of permeability implies, by definition, some alteration of the protective permeability barrier of the mucosa. It may even be argued that enhancement and tissue damage are not, in fact, distinct effects, but are simply degrees of the same phenomenon. This certainly appears to be the case with many of the enhancers reported in the literature. For a penetration enhancer to be of value, it is imperative that the change in permeability of a mucosal tissue be reversible within a time frame appropriate to the desired duration of drug delivery. Furthermore, there should be no cumulative toxicity nor permanent changes induced in the barrier properties of the mucosa with long-term use. A related issue is the use of enzyme inhibitors to reduce the enzymatic degradation of the drug within the mucosa, par-

ticularly in the case of a peptide drug. Hirai et a1.125 showed improved nasal absorption of insulin using bile salts, and proposed that the improvement was due, a t least in part, to the inhibitory effects of the bile salts on tissue peptidase activity. To date, however, these and other enzyme inhibitors have not been evaluated in oral mucosal delivery.

Formulation Factors Several factors related to the formulation or delivery system can influence the bioavailability or therapeutic efficacy of a drug delivered via the oral mucosae; obviously, the rate, duration, and kinetics of drug release are critical in this respect. However, there are additional factors, not all of which are encountered in conventional drug delivery, which must be considered in oral mucosal drug delivery. Sublingual Delivery-As outlined earlier, the sublingual mucosa appears to be relatively permeable, capable of giving rapid and appreciable absorption of low-molecular-weight drugs. To the knowledge of the authors, the only delivery systems evaluated clinically for sublingual delivery have been rapidly disintegrating tablets and liquid-fiIled soft gelatin capsules which a r e crushed open i n the mouth.2~Jb.~,l26 These systems are designed to give rapid drug release, leading to high local drug concentrations in the sublingual region. As a result of salivary flow, these concentrations are sustained for a relatively short period of time, probably in the order of only minutes. The sublingual area does not appear to have an expanse of smooth and relatively immobile mucosa that would be suitable for attachment of a retentive delivery system. For this reason, the main application of the sublingual route is likely to remain the delivery of small permeants for which short delivery times and infrequent delivery intervals are appropriate, or for which a rapid onset of action is desirable (e.g., nitroglycerin). Buccal Delivery-The buccal site differs from the sublingual in a number of important respects. First, the buccal mucosa is less permeable than the sublingual and does not give the rapid onset of absorption seen with sublingual delivery. Second, the buccal mucosa appears to be better suited to the use of retentive systems, such as a mucoadhesive tablet or patch system, in that it has a n expanse of smooth and relatively immobile surface for placement of such systems. These attributes make the buccal mucosa more suitable for sustained-delivery applications, delivery of less well permeating molecules, and perhaps peptide drugs. These applications place a number of additional constraints on the design of buccal delivery systems. First, it may be desirable to modify the local environment at the absorption site in order to optimize the buccal delivery of a drug. This may involve addition of a cosolvent or alteration of the pH a t the mucosal surface in order to increase the local solubility of the drug or enhance its partitioning into the mucosal tissues. Alternatively, it may involve use of a penetration enhancer or enzyme inhibitor, as has been described earlier (see Penetration Enhancement). Obviously, these approaches require incorporation of one or more additional agents into the delivery system, which can greatly increase the complexity of the formulation and perhaps also the manufacturing process. Second, it may be necessary to confine the applied dose of a drug to a proscribed region of the mucosa, in order to maximize the residence time of the drug a t the mucosal surface (i.e., the time available for absorption). Harris and Robinson127 demonstrated that the utility of mucosal delivery routes, such as the buccal route, can be resolved into two issues: (1)whether therapeutic drug levels can be attained in the systemic circulation via this route (this is a function of, among other factors, the permeability of the delivery site to Journal of Pharmaceutical Sciences I 7 Vol. 81, No. 1, January 1992

the drug), and (2) whether these therapeutic levels can be sustained (this is a function of the clearance half-life of the drug and the residence time of the dose at the delivery site). For example, this study calculated that, based on realistic delivery rates, therapeutic systemic levels could probably be achieved buccally for the synthetic vasopressin analogue DDAVP, and that this route could provide effective therapy if delivered contiduously (or at least at frequent intervals) a t this rate. Buccal delivery of insulin, on the other hand, could achieve only a few percent of therapeutic levels, apd effective buccal delivery of this drug would require both penetration enhancement and, due to its short biological half-life, virtually continuous delivery. For drugs with half-lives of several hours and wide therapeutic indices, maintaining therapeutic levels is not a major problem. For drugs with short half-lives, however (and many peptide drugs, for example, have half-lives of minutes only), this issue becomes highly significant. In order to maintain therapeutic levels of such short half-life drugs, it is necessary to deliver these drugs virtually continuously, or at least a t very frequent intervals. Obviously, retaining a delivery system a t the oral mucosae for a period of several hours or more can be a major problem, due to such challenges as salivary flow, ingestion of food and beverages, mastication, and speech. The same principles would also apply to the delivery of other constituents, such as penetration enhancers. In this case, it would be desirable to restrict the enhancer to the site of delivery not only to prevent a loss of enhancing effect at the absorption site, but also to prevent a generalized change in permeability of the membranes of the oral cavity. A third constraint of buccal delivery systems is that, since we are dealing with a n unconventional delivery route, patient acceptance is likely to be an obstacle in many cases, particularly for retentive delivery systems. Success will therefore depend on the patient's motivation, and on the delivery system being convenient to use and unobtrusive once in place. When all of these factors are considered, it appears likely that resilient adhesive systems, such as patch systems30.34.44.66 or adhesive controlled-release tablets,1os3l366 have a greater potential for sustained delivery than, for example, gels or disintegrating tablets. As has been outlined, however, the critical factor will probably be the development of delivery systems that are able to provide delivery of active constituent(s) for periods of 6 to 12 h or thereabouts, while still being acceptable to patients. In addition to these sustained-delivery applications, the buccal route has also found clinical use in delivery over somewhat shorter time-spans, using more conventional tablet formulations as opposed to the retentive systems described above. Examples here include delivery of methyltestosterone (see Table I), prochlorperazine,lZs rnorphine,129 and oxytocin.113.114 Local Delivery-The simplest and probably the most widely used delivery systems for local delivery to the oral mucosae are conventional mouthwashes, oral suspensions, and lozenges.llJ30 These give high drug levels in the oral cavity as a whole, but only for a short time.131 Indeed, salivary flow is probably even more important an issue in the context of local delivery than in sublingual or buccal delivery. For these types of therapy to be effective, therefore, it is necessary either to select drugs which are rapidly absorbed and effective under conditions of discontinuous delivery, or to make use of frequent dosing intervals. The duration of action of a drug can be improved somewhat by the use of ointments or creams which can be applied to the oral mucosae.11 More recently, a number of strategies have been proposed to give sustained drug levels in the oral cavity as a whole, by means of mucoadhesive lozenges131 or denture8 I Journal of Pharmaceutical Sciences Vol. 81, No. 1, January 1992

based delivery systems.12 Alternatively, delivery systems may be designed which provide sustained drug levels within a particular tissue or region of the oral cavity:10,13oJ32 such systems have applications in the treatment of periodontal disease and aphthae in particular.

Conclusions It is evident from the foregoing that the various oral mucosae, in particular the buccal and sublingual mucosae, vary greatly in terms of both their composition and their utility from the point of view of drug delivery. The sublingual mucosa is relatively permeable and is able to provide rapid absorption and onset of action for small drug molecules: indeed, the existing commercial sublingual formulations take advantage of these rapid onsets of action. However, this site appears to be poorly suited to the use of retentive systems; thus, development of sustained-delivery sublingual formulations may well be problematic. Future developments in sublingual delivery are perhaps likely to focus on small drug molecules that permeate rapidly and that can benefit from short delivery times and infrequent delivery intervals. Sublingual delivery is less likely to be of value in the case of peptide drugs, given the requirement for more-or-less continuous delivery of these agents. The buccal mucosa is considerably less permeable; therefore, longer delivery times are generally needed to achieve therapeutic levels, and onset of action is also slower. Owing to the nature of the site, however, it is probably feasible to design retentive systems which are able to provide sustained delivery for a number of hours. Thus, subject to considerations of permeability, this route may have potential for delivering peptides and other large drug molecules, where sustained delivery is frequently necessary. The chief limitation of the buccal mucosa as a drug delivery route is the relatively low permeability of this site. However, the fact that the permeability barrier is apparently located in the outermost cell layers of the epithelium is significant, as it should in principle allow the modification of this barrier by the surface application of penetration enhancers. In order to achieve this, it is necessary that we understand the nature and origin of the permeability barrier and the routes and mechanisms of permeation in rather more detail than we do at present. Obviously, the issue of toxicity, both acute and chronic, will be critical to the success of such penetrationenhancing strategies. Local delivery to tissues of the oral cavity is indicated in a relatively small number of conditions and does not have the broad range of applications common to systemic buccal and sublingual delivery. In many respects, however, the principles of local delivery, in terms of tissue permeability and formulation design, are broadly similar to those of systemic delivery. For these reasons, local delivery of drugs to the tissues of the oral cavity is likely to remain a relatively poorly studied area, at least for the foreseeable future.

References and Notes 1. Sobrero, A. Com tes Rendus Hebdomadaires &s Sciences de 1'Acadumie des diences 1847,24,247-248. 2. Walton, R. P.; Lacey, C. F. J . Pharmncol. Exp. Ther. 1935, 54, 61-76. 3. Walton, R. P. JAMA 1944,124, 138-143. 4. Katz, M.; Barr, M. J . Am. Pharm. ASSOC. 1955,44,419-423. 5. Gibaldi, M.; Kanig, J. L. J . Om1 Ther. Pharmacol. 1965, 1 , 440450. 6. Squier, C. A.; Johnson, N. W. Br. Med. Bull. 1975,31,169-175. 7. Paulson, W. Practitioner 1916,97,389. 8. Nagai, T.; Konishi, R. J . Controlled Release 1987, 6, 353-360. 9. Ishida, M.; Nambu, N.; Nagai, T. Chem. Phurm. Bull. 1983,31, 45614564. 10. fiagai, T. J . Controlled Release 1985,2, 121-134.

11. Gallagher, F.J.; Taybos, G. M.; Terezhalmy, G. T. J. Indiana Dent. Assoc. 1985,64,26-28. 12. Lamb, D. J.; Douglas, C. W. I. J.Dent. 1988,16,219-221. 13. Meyer, J.; Squier, C. A.; Gerson, S. J. The Structure and Function of Om1 Mucosa;Pergamon: Oxford, 1984. 14. Chen, S.-Y.; Squier, C. A. In The Structure and Function of Om1 Mucosa;Meyer, J.; Squier, C. A.; Gerson, S. J., Eds.; Pergamon: Oxford, 1984;pp 7-30. 15. Squier, C. A.; Hall, B. K. Arch. OmZ Bwl. 1984,29,45-50. 16. Chen, S.-Y. M.S. Thesis, University of Illinois, Chicago, 1970. 17. G e m n , S.J.; Harris, R. R. In The Structure and Function of Om1 Mucosa; Meyer, J.; uier, C. A.; Gerson, S. J., Eds.; Pergamon: Oxford, 1984;pp 1 4 1 . 18. Teng S.C. G.; Jarvinen, M. J.; Nelson, W. G.; Huang, J.-W.; Woodcack-Mitchell, J.; Sun, T.-T. Cell 1982,30,361-372. 19. W e d , P. W.; Cox, P. S.; uier, C. A.; Downing, D. T. Comp. Biochem. Physiol. B 1986, 3,529-531. 20. Rabinowitz, J. L.;Tavares, C.; Marsh, J. B. J. Periodont. Res. 1972,7,230-235. 21. Lekholm, U.; Svennerholm, L. J.Dent. Res. 1977,85,279-290. 22. Siegel, I. A. In The Structure and Function of Om1 Mucosa; Meyer, J.; Squier, C. A.; Gerson, S. J., Eds.; Pergamon: Oxford, 1984;pp 95-108. 23. GaleyiW. R.; Lonsdale, H. K.; Nacht, S. J. Invest. Dermatol. 1976,67,713-717. 24. Squier, C . A.; Hall, B. K. J.Invest. Dermutol. 1985,84,176179. 25. Pimlott, S.J.;Addy, M. Om1Surg. Om1 Med. OralPathol. 1985, 60.35-37. 26. Pimlott, S.J.; Addy, M. Om1 Surg.Om1 Med. Oral Pathol. 1985, 59, 145-148. 27. Graas, G. M.; Robinson, J. R. J. Pharm. Sci. 1988,77,3-14. 28. Lon er, M. A. Ph.D. Thesis, University of Wisconsin, Madison,

s9

2

19d

29. Scheuplein, R. J.; Blank, I. H. Physwl. Rev. 1971,51,702-745. 30. Ebert, C. D.; John, V. A.; Beall, P. T.; Roeenzweig, K. A. In Controlled-Release Technology: Pharmaceutical Applications; Lee, P. I.; Good,W. R., Eds.; American Chemical Society: Washington, 1987;pp 310-321. 31. Schor, J. M.; Davis, S. S.; Nigalaye, A.; Bolton, S.DrugDev.Znd. Pharm. 1983,9,1359-1377. 32. Abrams, D. Am. Heart J. 1983,105,8484354. 33. . Pitha, - ~ -..-J.; Harman, S. M.; Michel, M. k. J.Pharm. Sci. 1986,75, ICibltiI.

34. Anders, R.; Merkle, H. P. Znt. J. Pharm. 1989,49,231-240. 35. Russell, W. J.; Badcock, N. R.; Frewin, D. B.; Sansom, L. N. Eur. J. Clin. Pharmacol. 1988,35,437-439. Johnston, A.; Turner, P. J. Pharm. Phar36. Al-Sa ed-Omar, 0.; macof 1987.39.934-935. 37. Hussain, M.'A.;Aun st, B. J.; Kearney, A.; Shefter, E. Znt. J. Pharm. 1987,36,12f-130. 38. Brewster, D.;Hum hrey, M. J.; McLeavy, M. A. J. Pharm. Pharmacol. 1981,3~500-506. 39. Le Brun, P. P. H.; Fox, P. L. A.; De Vries, M. E.; Bodde, H. E. Znt. J . Pharm. 1989,49,141-145. 40. Hicks. D. C.Br. J.Phnrmucol. 1973.47.680P41P. 41. Siegel, I. A,; Gordon, H. P. Arch. Om1 Bwl.1985,30,4347. 42. Wright, E. M.; Pietras, R. J. J. Membr. Biol. 1974,17,293-312. 43. Anders, R.; Merkle, H. P.; Schurr, W.; Ziegler, R. J.Pharm. Sci. 1983,72,1481-1483. 44. &hum, W.; Knoll, B.; Zeigler, R.; Anders, R.; Merkle, H. P. J. Endocrinol. Invest. 1985,8,41-44. 45. Iahida. M.: Machida. Y.: Nambu. N.: Nawi. - . T. Chem. Pharm. Bull. 1981;29,810&16: 46. Aungst, B. J.; Rogers, N. J.;Shefter, E. J.Pharmacol. Exp. Ther. 1988,244,23-27. 47. Aunast, B. J.; Roaers, N. J. Znt. J.Pharm. 1989,53,227-235. 48. Daw-d, M. Y.; YEkorkala, 0.;Fuchs, F. Am. J.Obstet. Gynecol. 1980.138.20-24. 49. Morrison,-M. C. T.; Williams, J. P. G. Postgmd. Med. J. 1961, 37,9697. 50. Barsuhn, C. L: Olanoff, L. S : Gleason, D. D.; Adkins, E. L.; Ho, N. F. H. Clin. hharmacol. T L r . 1988,44,225-231. 51. Beckett, A. H.; Hoasie, R. D. Handbook Erp. Pharmacol. 1971, 28,25-46. 52. Gandhi, R. B. Ph.D. Thesis, University of Wisconsin, Madison, 1990. 53. Harris, D.; Robinson, J. R. Proceed. Intern. Symp. Controlled Release Bioact. Mater. 1990,17,228-229. 54. Dowty, M. E.; Irons, B. K.; Knuth, K. E.; Robinson, J. R. Proceed. Intern. Symp. Controlled Release Bioact. Mater. 1990,17, I

I

226-227. 55. Adams, D. J. Dent. Res. 1975,54,B19-B26. 56. Veillard, M. M.; Longer, M. A.; Martens, T. W.; Robinson, J. R. J. Controlled Release 1987,6,123-131. 57. Merkle, H. P.; Anders, R.; Sandow, J.; Schurr, W. In Delivery Systems forpeptide Drugs; Davis, S. S.; Illum, L.; Tomlinson, E., Eds.; Plenum: New York, 1986;pp 159-175. 58. Tanaka, M.; Yanagibashi, N.; Fukuda, H.; Nagai, T. Chem. Pharm. Bull. 1980,28,10561061. 59. Reid, C. 0.; Hardcastle, J.; Smith, C. J. J.Dent. Res. 1986,65, 673-678. 60. Kurosaki, Y.;Hisaichi, S.; Nakayama, T.; Kimura, T. Znt. J. Pharm. 1989,51,47-54. 61. Squier, C.A. J. Appl. Toxicol. 1986.6, 123-128. 62. Tucker, I. G. J.Pharm. P h a r m o l . 1988,40,679-683. 63. Zhang, J.; Ebert, C.; Mdames, S.; Gijsman, H. J.; Stanley, T. H. Pharm. Res. 1989,6,S135. 64. Yamahqra, H.; Suzuki, T.; Mizobe, M.; Noda, K.; Samejima, M. J . Pharm. Sci. 1990,79, 963-967. 65. Bell, M. D. D.; Murra ,G. R.; Mishra, P.; Calvey, T. N.; Weldon, B. D.; Williams, N. $Lancet 1985,1,71-73. 66. Bardgett, D.; Howard, C.; Murray, G. R.; Calvey, T. N.; Williams, N. E. Br. J. Clin. Pharmacol. 1984.17,198P-199P. 67. Siegel, I. A.; Izutsu, K. T.; Watson, E. Arch. Om1 Bwl. 1981,26, 357-361. 68. Tolo, K.; Jonsen, J. Arch. Oral Bwl. 1975,20,41-22, 69. Siegel, 1. A.; Izutsu, K. T.; Burkhart, J. J.Pharm. Sci. 1976,65, 129-131. 70. Siegel, I. A.; Izutsu, K. T. J.Dent. Res. 1980,59,1604-1605. 71. De Vries, M. E.; Bodde, H. E.; Verhoef, J. C.; Junginger, H. E. J . Controlled Release 1990,13,316. 72. Dowty, M. E. M.S. Thesis, University of Wisconsin, Madison, 1988. 73. Tavakoli-Saberi, M. R.; Audus, K. L. Pharm. Res. 1989, 6, 160-166. 74. Squier, C. A.; Hall, B. K. Arch. Om1 Biol. 1985,30,485-491. 75. Squier, C.A. J. Ultrastruct.Res. 1973,43,160-177. 76. Squier, C.A.; Rooney, L.J. Ultmstruct. Res. 1976,54,286-295. 77. Hill, M. W.; Squier, C. A. J.Anat. 1979,128,169-178. 78. Elias, P. M.; Friend, D. S. J. Cell Bwl. 1975,65,180-191. 79. Alfano, M. C.; Drummond, J. F.; Miller, S. A. J.Dent. Res. 1975, 54. 1143-1148. ~~~- ~ ~ 80. De Vries, M. E.; Bodde, H. E.; J u n 'nger, H. E. Proceed. Intern. Symp. Controlled Release Biwct. 8;ater. 1990,17,295-296. 81. Tolo, K. J. Arch. Om1 Biol. 1971,16,881-888. 82. Squier, C.A.; Hopps, R. M.Br. J.Dermatol. 1976,95,123-129. 83. Matoltsy, A. G. J.Invest. Dermatol. 1976,67,20-25. 84. Squier, C. A. J. Ultrastruct. Res. 1977,60,212-220. 85. Hayward, A. F. Znt. Rev. Cytol. 1979,59,97-127. 86. Odland, G. F. J.Invest. Dermatol. 1960,34,11-15. 87. Ha ard, A. F.; Hackemann, M. J. Ultrastruct. Res. 1973,43, 20E19 88. Hayward, A. F. Arch. Oral Bwl. 1973,18,67-75. 89. Squier, C. A. Arch. Om1 Biol. 1982,27,377382. 90. Matoltsy, A. G.; Parakkal, P. F. J. Cell Bwl. 1965,24,297-307. 91. Hayward, A. F. J. A n d . 1976,121,313-321. 92. Lavker, R. M. J. Ultmstruct. Res. 1976,55, 79-86. 93. Grayson, S.;Johnson-Winegar, A. G.; Wintroub, B. U.; Isseroff, R. R.; Epstein, E. H.; Elias, P. M. J.Invest. Dermatol. 1985,85, 289-294. 94. Freinkel, R. K.; Traczyk, T. N. J. Invest. Dermatol. 1985,85, 295-298. 95. Silverman, S.;Kearns, G. Arch. Om1 Biol. 1970,15,169-177. 96. Chapman, S.J.; Walsh, A. J. Invest. Dermatol. 1989,93,466470. 97. Squier, C. A. Br. J.Dermatol. 1984,111,253-264. 98. Siegel, I. A. Arch. Om1 Biol. 1984,29,13-16. 99. Sieael. I. A.: Gordon. H. P. J. Dent. Res. 1979.58.109. 100. Beciett, A. H.; Moffat, A. C. J. Pharm. Phar'm&ol. 1969,21, 1445-150s. 101. Kurosaki, Y.; Aya, N.; Okada, Y.; Nakayama, T.; Kimura, T. J. Pharmucobwdyn. 1986,9,287-296. 102. Sieeel. I. A. J. Invest. Dermutol. 1981. 76. 137-140. 103. Heiry; J. A.; Ohashi, K.; Wadsworth, J.; Turner, P. Br. J. Clin. Pharmacol. 1980,10,61-65. 104. Manning, A. S.;Evered, D. F. Clin. Sci. Mol. Med. 1976,51, 127-132. 105. Evered, D. F.; Vadgama, J . V. Biochem. Soc. Tmns. 1981, 9, 132-133. - - I

Journal of Pharmaceutical Sciences 1 9 Vol. 81, No. 1, January 1992

106. Schurmann, W.; Turner, P. J. Pharm. Phurmacol. 1978, 30, 137-147. 107. Kates, R. E. J. Med. 1977,8,393-402. 108. Duchateau, G. S.M. J. E.; Zuidema, J.; Merkus, F. W. H. M. Phurm. Res. 1986,3,108-111. 109. Dearden, J. C.; Tomlinson, E. J. Phurm. Pharmacol. 1971,23, 68S-728. 110. Kurosaki, Y.; Hisaichi, S.;Hamada, C.; Nakayama, T.; Kimura, T. J. Phurmacobwdyn. 1987,10,180-187. 111. Siegel, I. A.; Gordon, H. P. Exp. Mol. Path. 1986,44, 132-137. 112. Manning, M. C.; Patel, K.; Borchardt, R. T. Pharm. Res. 1989, 6,903-918. 113. Dillon, T. F.Surg. Forum 1959,9,679-682. 114. Dillon; T. F.; D& las, R. G.; du Vigneaud, V.; Barber, M. L. Obstet. Gynecol. d60,15,587692. 115. Nagai, T..; Machida, Y. Pharm. Int. 1985,6 , 196-200. 116. Stratford, R. E.; Lee, V. H. L. Znt. J . Pharm. 1986,30,73-82. 117. Garren, K. W.; Repta, A. J. Znt. J . Phurm. 1988,48,189-194. 118. Lee, V. H.L.; Yamamoto, A. Adv. Drug Del. Rev. 1990, 4, 171-207. 119. Gibaldi, M.; Feldman, S.J. Phurm. Sci. 1970,59,579-589. 120. Walters, K. A. In Tmnsdermal Drug Delivery: Developmental

10 I Journal of Pharmaceutical Sciences Vol. 81. No. 1, January 1992

Issues and Research Initiatives; Hadgraft, J.; Guy, R. H., Eds.; Marcel Dekker: New York, 1989;pp 197-246. 121. Barry, B. W. In Pharmacology of the Skin, Volume 1;Shroot, B.; Schaefer, H., Eds.; Karger: Basel, 1987;pp 121-137. 122. Barry, B. W. J. Controlled Release 1987,6,85-97. 123. Kuroaaki, Y.; Hisaichi, S.; Hamada, C.; Nakayama, T.; Kimura, T. Znt. J. Pharm. 1988,47,13-19. 124. Kurosaki, Y.; Hisaichi, S.; Hon , L -Z.; Nakayama, T.; Kimura, T. znt. J. Phrrn. 1989,49,47-%5.. 125. Hirai, S.;Yashiki, T.; Mima, H. Int. J.Pharm. 1981,9,173-184. 126. Nyberg, G. Eur. Heart J. 1986,7, 673-678. 127. Harris, D.;Robinson, J. R. Biomateriuls in press. 128. Hessell, P. G.; Llo dJones, J. G.; Muir, N. C.; Parr, G. D.; Sugden, K. Znt. J. $harm. 1989,52,159-164. 129. Fisher, A. P.; Fung, C.; Hanna, M. Br. J. Clin. Pharmacol. 1987, 24.685-687. , --130. Goodson, J. M. In Medical Applications of Controlled Release, Volume 2;Lan er, R.S.; Wise, D. L., Eds.; CRC: Boca Raton, FL, 1984;pp llk138. 131. Collins, A.E.; Deasy, P. B. J. Phurm. Sci. 1990,79,116119. 132. Brook, I. M.;Tucker, G. T.; Tuckle , E. C.; Boyes, R. N. J. Controlled Release 1989,10,183-18s: