Potential and Problems of Developing

Advanced Drug Delivery Reviews 50 (2001) 175–203www.elsevier.com/ locate /drugdeliv

Potential and problems of developing transdermal patches forveterinary applications

a , b*Jim E. Riviere , Mark G. PapichaCenter for Cutaneous Toxicology and Residue Pharmacology, Department of Farm Animal Health and Resource Management,

College of Veterinary Medicine, North Carolina State University, Raleigh, NC 27613, USAbDepartment of Anatomy, Physiological Sciences and Radiology, College of Veterinary Medicine, North Carolina State University,

Raleigh, NC 27613, USA

Abstract

A new frontier in the administration of therapeutic drugs to veterinary species is transdermal drug delivery. The primarychallenge in developing these systems is rooted in the wide differences in skin structure and function seen in species rangingfrom cats to cows. The efficacy of a transdermal system is primarily dependent upon the barrier properties of the targetedspecies skin, as well as the ratio of the area of the transdermal patch to the species total body mass needed to achieveeffective systemic drug concentrations. A drug must have sufficient lipid solubility to traverse the epidermal barrier to beconsidered for delivery for this route. A number of insecticides have been developed in liquid ‘pour-on’ formulations thatillustrate the efficacy of this route of administration for veterinary species. The human transdermal fentanyl patch has beensuccessfully used in cats and dogs for post-operative analgesia. The future development of transdermal drug delivery systemsfor veterinary species will be drug and species specific. With efficient experimental designs and available transdermal patchtechnology, there are no obvious hurdles to the development of effective systems in many veterinary species. 2001Elsevier Science B.V. All rights reserved.

Keywords: Veterinary drug delivery; Transdermal drug delivery; Fentanyl; Iontophoresis

Contents

1. Introduction ............................................................................................................................................................................ 1762. Overview of comparative cutaneous anatomy and physiology..................................................................................................... 176

2.1. Structure and function of skin ........................................................................................................................................... 1762.2. Species and regional differences........................................................................................................................................ 1802.3. Quantitating percutaneous absorption ................................................................................................................................ 181

2.3.1. Fick’s law of diffusion............................................................................................................................................ 1812.3.2. In vivo methods ..................................................................................................................................................... 1832.3.3. In vitro methods..................................................................................................................................................... 184

2.4. Transdermal pharmacokinetics .......................................................................................................................................... 1863. Transdermal drug delivery ....................................................................................................................................................... 187

3.1. Topical formulations versus patches .................................................................................................................................. 187

*Corresponding author. Tel.: 11-919-513-6305; fax: 11-919-513-6358.E-mail address: Jim [email protected] (J.E. Riviere).

]

0169-409X/01/$ – see front matter 2001 Elsevier Science B.V. All rights reserved.PI I : S0169-409X( 01 )00157-0

176 J.E. Riviere, M.G. Papich / Advanced Drug Delivery Reviews 50 (2001) 175 –203

3.2. Systems developed for human use ..................................................................................................................................... 1893.3. Percutaneous absorption in veterinary species .................................................................................................................... 190

3.3.1. Topical pesticides................................................................................................................................................... 1913.3.2. Transdermal patches............................................................................................................................................... 193

3.3.2.1. Fentanyl.................................................................................................................................................... 1933.3.2.2. Testosterone.............................................................................................................................................. 196

3.3.3. Iontophoresis ......................................................................................................................................................... 1964. Transdermal patches in veterinary species ................................................................................................................................. 198

4.1. Important considerations unique to veterinary medicine ...................................................................................................... 1984.2. Potentials ........................................................................................................................................................................ 2004.3. Limitations ...................................................................................................................................................................... 200

5. Conclusions ............................................................................................................................................................................ 201References .................................................................................................................................................................................. 201

1. Introduction information on chemical and drug absorption insome veterinary species, often because select species

The primary routes of drug administration utilized (pigs, dogs) serve as models for human drug deliv-in veterinary medicine include oral, intramuscular, ery. However, these studies and those conducted insubcutaneous and intravenous dosing. Over the last rodents mainly for toxicological risk assessmentdecade, great strides have been made in using topical purposes, have taught that significant differences‘pour-on’ and ‘spot-on’ applications of pesticides exist in the barrier functions of different mammalianand antiparasitics (e.g., fenthion, ivermectin, species. In order to interpret drug delivery data inlevamisole, fipronil) for transdermal delivery in veterinary species, as well as to consider the designveterinary species. In this article, the term transder- constraints in developing patches for animal health, amal implies use of a topical drug application to thorough knowledge of the mechanism of chemicalachieve systemic pharmacological effects. In human absorption across the skin barrier and criteria formedicine, there has been an increase in marketing selecting appropriate model systems is needed.and in acceptance of transdermal patches (e.g.,scopolamine, nitroglycerin, estradiol, testosterone,nicotine, clonidine, fentanyl) as an alternate method 2. Overview of comparative cutaneous anatomyfor systemic drug delivery. ‘Patches’ differ from and physiologyother topical formulations developed for systemicdelivery by virtue of the fact that the patch controls One of the most obvious factors, which wouldthe rate of drug delivery from these systems, rather impact the design of transdermal patches for vet-than the drug’s permeability through the skin that erinary species, is the anatomical difference in theoccurs with topical formulations. Finally, electrically skin structure among species. Even from a cursorydriven transdermal iontophoretic delivery systems visual inspection, it is obvious that pig, dog andare currently being developed to provide a more sheep skin have very different appearances whichcontrolled systemic delivery of peptide drugs not would impact the design of transdermal patches.easily delivered by other routes. The literature de- Therefore, prior to a discussion on patch design andscribing the development of transdermal patch sys- function, the comparative anatomy and function oftems for humans is extensive, but there is little skin must be briefly reviewed.development and even less literature in the veterinaryfield. 2.1. Structure and function of skin

In order to probe the application of transdermalpatch technology for veterinary medicine, it is im- Skin is one of the largest organs of the bodyperative to understand relevant dermal anatomy and having a complex structure that reflects its myriad ofphysiology that interacts with drug transport through biological functions. Recent reviews of this topicthe skin. Surprisingly, there is a great deal of [1–3] will serve as the basis for this synopsis. The

J.E. Riviere, M.G. Papich / Advanced Drug Delivery Reviews 50 (2001) 175 –203 177

function most pertinent to its role in transdermal and a wide range of blood perfusion for modulatingdelivery is that of being a barrier to the outside heat transfer to the environment. All of these func-environment. The skin is considered to be an effi- tions may affect the rate and extent of absorption ofcient barrier to prevent penetration, and thus sys- topical chemicals across the skin. Additional func-temic absorption, of most hydrophilic and ionic tions of skin include mechanical support, neuro-compounds. It may be relatively permeable to com- sensory reception, endocrinology (e.g., activation andpounds of moderate lipophilicity as will be discussed metabolism of Vitamin D), immunology, andbelow. A second function of the skin which impacts glandular secretion.our topic, is the skin’s role in thermoregulation The histological structure of skin is depicted inmanifested by the presence of hair and fur for Fig. 1. This heterogeneous organ is composed ofinsulation, sweat glands for evaporative heat loss, three layers and a number of distinct appendages.

Fig. 1. Microstructure of mammalian skin (with permission from Monteiro-Riviere [1]).


The outermost layer, and the one responsible for the removed by successive stripping with adhesive tape,skin’s barrier properties, is the avascular epidermis all but a superficial barrier to chemical penetrationwhich is structurally and functionally divided into remains. For hydrophilic compounds, the stratumthe stratum corneum, stratum lucidum, stratum corneum has been estimated to provide 1000 timesgranulosum, stratum spinosum, and the innermost the diffusional resistance to penetration as the cel-stratum basale which rests on the basement mem- lular layers beneath it. Stratum corneum’s barrierbrane or epidermal–dermal junction. The most function can be assessed by measuring how easilymetabolically active layer is the stratum basale, dermal water can traverse to the outside of the bodyprimarily composed of keratinocytes which actively in the process of normal insensible water loss. Thisdivide, undergo differentiation and migrate outward process termed transepidermal water loss (TEWL)to populate the other epidermal layers. Interspersed can be noninvasively measured on the skin surfaceamong these basal keratinocytes are other non- and is often employed as an estimate of epidermalkeratinocytes which include the pigmented barrier function. Tape-stripping dramatically in-melanocytes, the immunologically important Langer- creases TEWL. However, for very lipophilic com-hans cells, and the sensory Merkel cells; all three of pounds with lipid /water partition coefficients greaterwhich have minimal involvement in the skin’s than 400, the aqueous dermis may provide a signifi-physical barrier properties. cant barrier to systemic absorption. After lipophilic

As basal keratinocytes divide and move upward, drugs are absorbed across the intact stratum cor-they lose their cuboidal shape and prominent nu- neum, further absorption of exceedingly lipophiliccleus. Their metabolism changes to synthesize the drugs is limited by their tendency to partition into thestructural protein keratin and the enzymes necessary stratum corneum and not diffuse further into theto produce a variety of lipids. The stratum epidermis, forming a stratum corneum reservoir.granulosum layer is named for the presence of Ultimately, drugs can be eliminated in this fashion askeratohyalin granules in histological sections. They corneum cells are sloughed off into the environment.are depicted as flattened with an almost nonexistent There are a number of other factors that cannucleus and cytoplasm being primarily occupied by modulate stratum corneum permeability. Hydrationkeratin and membrane coating granules. The stratum of the stratum corneum increases its permeability tocorneum, the outermost layer, is where all of the many compounds. Thus, at high relative humidityskin’s chemical barrier properties reside. The cells (e.g., greater than 80%), compound penetration mayare extremely flat and consist entirely of keratin drastically increase as is seen in Fig. 2 with para-‘cemented’ to adjacent cell layers by an intercellular thion absorption dosed under different degrees oflipid matrix derived from the contents of the ex- relative humidity [5]. Complete hydration of theteriorized membrane coating granules. This forms a barrier, obtained by using an occlusive dressing‘brick and mortar’ structure [4] that has been used to which prevents insensible water loss, drasticallyconceptualize the barrier property of skin. The increases penetration. Most transdermal deliveryintercellular lipid ‘mortar’ component of the struc- devices, by their very nature, function as occlusiveture is the primary route of penetration of all topical systems.chemicals studied to date. The stratum corneum is The dermis is a highly vascular tissue arising fromcomposed of numerous layers of stratified, highly its role in thermoregulation via modulation of bloodorganized keratinized cells embedded in a lipid flow. The dermal vasculature provides oxygen andmatrix primarily composed of ceramides, sterols, and nutrients to the avascular epidermis above it. Thisother neutral lipids. The stratum corneum layers are blood supply is under complex neural and localcontinually sloughed off to the external environment humoral control. Cutaneous blood flow may beby friction and are replaced by the migration of shunted away from perfusing capillary beds throughdifferentiated cells from layers below. This cycle non-exchanging shunts when the mammal is in atakes approximately 4 weeks in humans. mode of heat retention. In contrast, when heat

The stratum corneum is the primary barrier to exchange with the cooler skin is desired, shunts arechemical penetration. When the stratum corneum is closed and exchanging capillaries closer to the


Fig. 2. Effect of relative humidity on the percutaneous absorption of parathion using in vitro porcine skin (mean6S.E.M.).

epidermis are perfused. Dermal perfusion has beenestimated to change by a factor of 100 depending onthermoregulatory status. For certain compounds,modulation of dermal perfusion may alter transder-mal drug delivery. Vasoconstriction decreases dermalperfusion and traps drugs which have penetrated theepidermal barrier, thereby decreasing systemic ab-sorption but enhancing the local activity. In contrast,vasodilatation would promote systemic delivery butminimize local accumulation. Fig. 3 depicts this laterscenario where lidocaine absorption into the systemiccirculation was increased when cutaneous vascula-ture was dilated with co-administered tolazoline, an

Fig. 3. Effect of cutaneous vasodilation on systemic absorption ofa-adrenergic receptor antagonist [6]. The subcuta-lidocaine administered by transdermal iontophoresis in pigs

neous tissue is composed of adipose cells which (mean6S.E.M.). Solid line is lidocaine alone and dotted line isprovide shock-absorbing functions, acts as ther- lidocaine coadministered with the vasodilator tolazoline.moinsulation, and provides a reserve depot ofenergy. However, relative to transdermal delivery, transport from the skin surface into the dermis,this lipoidal layer may also serve as a depot for bypassing the stratum corneum barrier. Most recentlipophilic drugs that have penetrated to the dermis. studies suggest that this is not an important route of

The appendages including hair, sweat and sebace- entry for topically applied drugs since the surfaceous glands provide another potential route for drug area of their orifices are very small compared to that


of the interfollicular epidermis. However, drug par- palmar and plantar [8]. The palmar and plantarticles that lodge in the openings of hair follicles or regions are highly cornified due to their functionsglandular ducts, may slowly release drug into these which involve constant abrasion. In addition toappendages. A similar mechanism occurs with some thickness, hair follicle density, structure and arrange-topical flea products in dogs and cats. The drug ment varies greatly across different body regions [3].fipronil for dogs and cats has its advantage over Because of this factor, in humans the scalp should beother topical flea products in that animals can be viewed differently than other body regions. Finally,bathed after topical administration. The drug collects cutaneous blood flow varies among the differentin the sebum and oils of the follicles and is slowly body regions [9,10].released. This results in a 30-day effect after a single These differences are exacerbated when diversifi-treatment. cation among species is considered. Table 1 lists the

Hair follicle density and arrangement is a major comparative skin thickness and blood flow deter-variable when comparing drug absorption across mined using laser Doppler velocimetry, across threespecies. Species with high hair density, tend to have different body sites in various veterinary speciesattenuated interfollicular epidermis, which may pro- [9,10]. As one can appreciate, there are significantvide a reduced barrier to drug penetration. Addition- differences in all parameters across body sites andally, some drugs may physically bind to hair further species. The ear and abdomen generally have thepreventing absorption. The secretions of sweat and highest rate of blood flow while the buttocks has thesebaceous glands also provides another variable, thickest epidermis and stratum corneum. The hairboth by providing a mechanism to hydrate skin via and sweat / sebaceous gland density, as well assweat or decreased water loss due to a layer of anatomical structures, between these sites andsebum on the skin’s surface. Similarly, sebum may species is also significantly different. Another ana-serve as a vehicle for drugs that are soluble in it. tomical variable which has been reported to sig-

The final factor which must be considered is the nificantly affect percutaneous absorption is the sizeability of the epidermis to metabolize drugs before of an individual corneocyte, although this parameterbeing absorbed by the systemic circulation [7]. has not been evaluated across veterinary species. TheStudies have demonstrated that the skin possesses primary site used for transdermal patch delivery inboth phase I and II biotransformation pathways, veterinary medicine is the back, because this is oneallowing first pass cutaneous metabolism of com- of the few locations that they cannot lick, chew, orpounds. In addition to cellular enzymes, soluble scratch. However, in sick or debilitated animals,esterases are also present which have been used in other sites such as the lateral thorax or inguinal areatransdermal drug delivery studies. In these cases, are used. It is important if drug delivery frompro-drugs are formulated as esters which promote patches applied to different sites is compared, thesestratum corneum transport. As these pro-drugs reach anatomical and physiological factors may impact thethe epidermis, soluble esterases cleave off the ester drug flux observed.groups and release active parent drug into the There have been a number of studies comparingsystemic circulation. the percutaneous absorption of chemicals across

laboratory animal species, pigs, primates and humans2.2. Species and regional differences focused on evaluating animal models for human drug

development and chemical risk assessment. TheseAs can be appreciated from the above discussion, are extensively reviewed elsewhere [11–13]. For

there are significant species differences which could most compounds, the rank order of the extent ofimpact on the design of transdermal delivery patches absorption is mouse 4 rats / rabbits . humans /pigs /for veterinary species. In fact, significant differences primates. In general, pigs and monkeys have similarexist among different body regions within a species. absorption characteristics as humans. This is due toFor example, in humans it is accepted that the rate of similar anatomy, relatively sparse hair coats (ab-penetration for most non-ionized compounds is: domen of monkeys), and biochemically similar com-scrotal . forehead . axilla / scalp . back/abdomen . positions of stratum corneum lipids. Unfortunately,


Table 1Stratum corneum thickness, epidermal thickness, and number of epidermal cell layers at various sites on various species as determined in

afrozen sections

Species Area Number of Epidermis Corneum Blood flowcell layers (mm) (mm) (ml /min per 100 g)

Cat Abdomen 2.0760.12 11.1261.76 8.1561.29 6.1960.94Back 1.6060.19 11.5862.33 9.7361.01 2.3960.35Buttocks 1.6060.12 10.0161.77 11.4962.73 1.8260.59Ear 2.1360.17 10.0161.53 8.9060.91 6.4662.30

Dog Abdomen 2.1760.14 13.7560.94 12.2062.12 8.7861.40Back 2.0060.19 14.2162.61 11.5861.83 1.9460.27Buttocks 2.1160.36 17.3067.52 9.7361.80 2.2160.67Ear 2.7860.39 18.5363.68 15.0961.83 5.2161.53

Pig Abdomen 3.1160.19 30.5861.62 15.2961.14 10.6862.14Back 4.2260.34 41.5561.94 15.0660.63 2.9760.56Buttocks 4.1760.36 46.0263.80 18.0761.82 3.0860.48Ear 4.9460.48 44.6364.26 13.1361.80 11.7063.02

Cow Abdomen 2.0660.23 16.6862.66 10.0461.40 10.4962.13Back 2.6760.27 24.4061.47 11.5861.53 5.4961.49Buttocks 2.7160.27 24.0962.71 25.68610.71 6.0361.84Ear 3.0060.31 31.5161.69 13.5962.09 6.9862.19

Horse Abdomen 2.2260.24 21.3163.84 8.1961.38 8.9061.46Back 2.9460.23 20.5461.59 8.3461.10 2.9960.86Buttocks 2.0060.23 22.8661.89 7.8861.04 3.1661.22Ear N/A N/A N/A N/A

a All data are mean6S.D. Adapted from Monteiro-Riviere et al. [10].

similar studies have not been conducted in veterinary 2.3.1. Fick’s law of diffusionspecies except for one excellent review on topical The movement of chemicals across the stratumdelivery to cattle and sheep [14]. In fact, this corneum barrier into the epidermis occurs primarilyliterature is heavily skewed toward studies conducted by passive diffusion driven by the applied con-in pigs due to their use as an accepted animal model centration of drug on the surface of the skin. This isfor humans. Studies are also available in dogs due to best expressed using Fick’s Law of Diffusion whichtheir occasional use in biomedical research as well as states that the steady state of drug flux across astudies of human transdermal patches utilized in the membrane can be expressed as:veterinary clinic.

Flux (J)

DP]5 (Concentration Gradient) (Surface Area)2.3. Quantitating percutaneous absorption h

In order to quantitate the delivery of drug across where D is the diffusion coefficient or diffusivity ofthe stratum corneum and epidermis, a framework the drug in the intercellular lipids of the stratummust be developed to relate transdermal flux to corneum, P is the partition coefficient for the drugapplied surface concentration. In most cases, transfer between the stratum corneum and the dosing mediumacross the skin is similar to transfer across any on the skin surface, and h is the skin thickness ormembrane and thus these techniques have been actual path-length through which the drug diffusesextensively applied to dermal transport [15–18]. across the diffusion barrier. The driving force for thisThese approaches will be reviewed briefly. process is the concentration gradient that exists


between the applied dose and the blood-perfused phenomenon which would favor drug movement outdermal environment. The term DP/h is often called of the formulation into the more favorable lipidthe permeability coefficient. Kinetically, this is a environment of the stratum corneum. The reversefirst-order rate constant that is the basis for the scenario would be operative for a hydrophilic drug.absorption rate constant (K ) obtained in phar- Independent of these formulation effects, the penet-a

macokinetic analyses of transdermal drug delivery rating drug must have some propensity to partitionstudies. Transdermal flux is expressed in terms of into the intercellular lipids of the stratum corneum. Itskin surface area, making the two important prop- is generally accepted that the optimal log octanol /erties of dosage in a transdermal system the con- water partition coefficient for a drug to penetrate thecentration of drug applied and the surface area of stratum corneum is approximately two. In otherapplication. Finally, Fick’s law expresses the steady- words, the drug is partitioned in the lipid phasestate flux of drug that occurs when this rate becomes approximately 100-fold. For hydrophilic drugs withconstant. In skin diffusion studies, this occurs after low partition coefficients, pro-drugs can be formu-passage of a lag time which is a function of the drug lated (e.g., by esterification as discussed above)‘loading’ the stratum corneum and dermis, diffusivi- which increases the drug’s permeability across thety, and thickness of the skin. Because of the differ- stratum corneum. The lipophilic moiety (e.g., ester)ences in skin thickness seen between body sites and is then cleaved in the epidermis, dermis or evenspecies, this is a major variable in comparative plasma and active parent drug is then distributedtransdermal delivery studies. throughout the systemic circulation. If the partition

The diffusivity of a drug is a function of the coefficient is too great, drug may have a tendency tomolecular weight, molecular size or more specifical- sequester into the stratum corneum and not enter thely volume, molecular interactions with skin con- more aqueous dermis, thereby decreasing systemicstituents (e.g., hydrogen bonding, hydrophobic inter- delivery. If it does penetrate into the dermis, the highactions, etc.), the drug’s solubility in the membrane lipid partition coefficient may favor formation of amilieu, and the degree of ionization. Large molecular dermal depot. It must be noted, that the drug mustweight drugs (approximately greater than 400 Da, also have partitioning properties which are favorablee.g., proteins) have extremely low diffusivities, thus for entering into solution into the aqueous plasma, oreffectively preventing them from being absorbed be able to bind to plasma proteins for systemicacross the skin barrier. These properties are often absorption to occur. These characteristics whichcorrelated to a drug’s melting point. For compounds promote stratum corneum permeability and plasmathat are partially ionized, diffusivity is decreased, uptake are confounding and impact on the ability ofindicating that only the non-ionized fraction of a using pure physicochemical properties to predict inweak acid or base is available for diffusion across vivo drug delivery.the stratum corneum. This is a function both of the Vehicle or formulation effects in themselves maypH of the dosing medium as well as the pH of the become very complex. In addition to their primaryskin that normally ranges from 4.2 to 7.3 depending role in determining the functional partition coeffi-on species and environmental conditions. cient for the system, they themselves may also be

The partition coefficient determines the ability of absorbed into the stratum corneum and alter thethe drug to gain access to the diffusion pathway. permeability properties of the lipids. This is thePartition coefficient is usually determined in ex- mechanism of many absorption enhancer moleculesperimental systems by measuring octanol /water or that are intentionally incorporated into a formulationlipid /water partitioning. The higher the ratio, the to promote transdermal delivery. These includegreater the lipophilicity. Dermal formulations may substances such as ethanol, oleic acid, terpenes, andsignificantly modify percutaneous absorption. For a a host of other lipophilic chemicals. As they diffuselipid-soluble drug, a lipid base formulation would into the lipid barrier, the physical chemical prop-tend to decrease absorption by retaining applied drug erties of the lipid are altered which in turn changesat the skin surface. In contrast, an aqueous base the permeability coefficient of the lipid relative to thewould promote absorption solely by this partitioning penetrating drug. Occlusion and hydration of the


stratum corneum, as was discussed in the context of transdermal preparation is expressed in terms ofFig. 2 above, operates by this mechanism. There are concentration per unit of applied surface area. Ana number of excellent technical reviews describing examination of Fick’s law also illustrates how thethese mechanisms which are beyond the scope of this comparative anatomical differences among speciesreview [19,20]. The vehicle may also extract the may alter drug flux. These include skin thickness,intercellular lipids, thereby reducing barrier function hair density and thus interfollicular epidermal thick-and enhancing absorption. ness, as well as differential lipid composition which

Finally, for compounds that are absorbed across would change the permeability coefficient. Furtherthe stratum corneum by passive diffusion, other modulating factors include different rates of cuta-properties of the skin may modulate the rate and neous blood flow and capacity or mechanisms ofamount of compound which is absorbed into the first-pass cutaneous biotransformation. Similarly,systemic circulation. We have already shown in Fig. environmental or clinical factors such as occlusion,3 how the extent of blood flow may directly effect high relative humidity, temperature (which couldthe amount of drug which has crossed the stratum increase permeability as well as increase dermalcorneum that is ultimately absorbed into the systemic blood flow), disease-induced changes in skin struc-circulation. In general, blood flow is not rate limiting ture or function, or abrasion (which would removefor topically applied drugs unless the normal rate of the stratum corneum barrier) may alter the transder-dermal perfusion is not sufficient to absorb the mal flux.amount of drug entering the dermis [21]. This occurswhen the drug is not soluble in the blood, or when

2.3.2. In vivo methodsthe rate of transdermal flux exceeds the capacity ofThe target of most transdermal delivery systems isthe blood to clear it from the dermis. The latter is

the intact animal, making in vivo methods theonly seen in transdermal systems which deliver apreferred approach for assessing feasibility and patchrelatively high flux of drug. A second factor whichdesign. In many ways, these methods are no differentmay modulate the amount of drug absorbed into thethan those employed for evaluating the absorption ofsystemic circulation is the extent of first-pass cuta-drugs from any route of administration. The goalsneous biotransformation. We have discussed this inare to determine the fraction of topically appliedterms of ester pro-drugs, however if any drug is adose that leaves the delivery device and enters thesubstrate for epidermal enzyme systems, a fraction ofsystemic circulation, as well as the concentrations ofdrug, which has diffused through the stratum cor-delivered drug relative to efficacy and safety in theneum, may be metabolized prior to absorption intotarget animal.the systemic circulation. Our laboratory has demon-

The primary method used to assess transdermalstrated this phenomenon with topically applied pes-absorption in the intact animal, is to compare theticides [22–24]. Finally, environmental and skinarea under the concentration versus time (AUC)temperature may alter drug flux across the skin.curve after topical dosing to that obtained afterThere are two components to this effect. The first,intravenous administration (bioavailability) or viaeven documented evidence using avascular in vitroanother route (bioequivalence). These techniquessystems, relates to an increased diffusivity of thehave been extensively reviewed elsewhere for vet-penetrating drug in the stratum corneum arising fromerinary scenarios [26–28] and transdermal / topicalthermodynamic effects [25]. The second mechanismpreparations in humans [29–31].is due to an increase in dermal blood flow secondary

The basic approach for assessing bioavailabilityto systemic temperature regulation.(compare to intravenous route) or bioequivalenceIn conclusion, transdermal drug flux occurs pri-(compare to existing product route) is to calculate themarily by passive diffusion through the intercellularfollowing ratio:lipid domain of the stratum corneum. Its rate is

dependent upon the permeability coefficient of the(AUC ) (Dose )transdermal routedrug, applied concentration and surface area. Unlike ]]]]]]]](AUC ) (Dose )other routes of drug administration, dosimetry for a route transdermal


If bioavailability (F ) is being determined, then the added to the amount of drug recovered from urine,feces and carcass, is less than the applied dose.value cannot exceed 1.0 since by definition an

The final in vivo strategy employed is to assessinjected dose is totally available systemically. Ifabsorption by measuring how much drug is absorbedbioequivalence is being determined, then the ratiointo the skin under the application site. This iscan vary from almost zero to any number since thisusually done simultaneously with assessing absorp-approach is essentially comparing the bioa-tion in plasma or excreta in order to assess thevailabilities of two formulations. If the goal is tocutaneous distribution of the applied dose. Afterdevelop a transdermal system which is bioequivalentremoval of the formulation or device, the skinto another already approved formulation, then thesurface is swabbed to remove non-penetrated doseabove ratio must be within the range of 0.8–1.25and a core biopsy is then taken and assayed for drug.relative to the pioneer formulation, estimated from aAlternatively, the biopsy may be quick-frozen inspecific study design with sufficient statistical powerliquid nitrogen and cryostat sections collected toand level of confidence (a). Other metrics of bio-determine the penetration profile in the skin. Otherequivalence, such as C and partial AUC may alsomaxworkers have used cellophane tape to strip off thebe appropriate. The use of partial AUC (e.g.,stratum corneum and measure drug profile in thisAUC ) is often suggested to insure that ‘dose0–Tmax tissue. Rougiere et al. [32,33] have suggested that thedumping’ from the patch does not occur.amount of drug remaining in the stratum corneum 30A simple approach often employed is to assay themin after removal of a topical formulation correlatesdrug remaining on the skin, or in the delivery deviceto its systemic absorption. This is based on the factfor patch studies, and compare this to the amountthat this measurement relates to the driving force fororiginally applied. The difference is the amount ofdiffusion of drug as presented earlier in Fick’s Law.

drug that was delivered by the formulation. TheThe utility of this approach is in comparing formula-

limitation to this approach is that no information istions using a non-invasive technique, although ve-

available as to the biological fate of the deliveredhicle effects that alter the stratum corneum per-

drug. meability, may confound the correlation. To com-Another approach that has been used to assess pletely describe the absorption of a topically applied

absorption of topical formulations, relative to the transdermal dose, a mass balance study should beactual percent of drug dose which has been absorbed conducted which assesses systemic delivery byinto the systemic circulation and distributed to the measuring blood concentrations, mass balance andanimal, is to use the above equation but substitute systemic distribution from total recovery in urine,recovery in urine and feces for AUC. This is feces, and internal organs; as well as skin biopsiespreferred when the blood or plasma concentrations and stratum corneum samples to measure local skinobtained after topical dosing are too low to adequate- distribution under the application site. Such a studyly quantitate with existing analytical methodologies. defines the total disposition of the drug and greatlySuch studies are often used to quantitate systemic facilitates the design of subsequent studies aimed toabsorption of topically applied pesticides. If the drug assess bioequivalence or clinical efficacy. The draw-undergoes tissue binding or sequestration in the back to in vivo studies is that the pattern of percuta-animal’s body, a mass balance study is often per- neous absorption is confounded with the variabilityformed where drug recovered in the carcass is also inherent to systemic distribution and elimination.factored into this equation. This impacts on the precision of in vivo data analysis

A problem that plagues all of these approaches is to identify formulation-specific factors that alter drugwhen drug is absorbed across the stratum corneum delivery. This problem has led to increased reliancebut does not reach the dermal circulation. This on and use of in vitro methods early in the develop-occurs when drug forms a depot in the stratum ment cycle.corneum or subcutaneous tissue. These drug con-centrations could be related to local irritation of the 2.3.3. In vitro methodsapplied drug. Mass balance studies will detect these A significant amount of studies are conductedwhen the total dose recovered from the dosing site, using various in vitro techniques. Before these are


presented, a distinction must be made between in lipophilic penetrants in the perfusate. If viability ofvitro studies used to assess dissolution or release the skin membranes is desired for cutaneous metabo-from a transdermal device, and those used to assess lism studies, an energy source such as dextrose andabsorption across the skin. The former are very O /CO gas is fed into the perfusate to maintain2 2

similar to techniques used in assessing dissolution of some degree of metabolism. These systems have aoral drug tablets and measuring the rate of drug regulated temperature at 32–378C to mimic skinrelease from the formulation. Unlike compendial oral surface temperature or core-body temperature.dissolution methods, there are standards developed In both static and flow-through systems, drug fluxfor transdermal systems such as those described in is monitored in the receptor cell. The data plotted inthe US Pharmacopeia [34]. However, it must be Fig. 2 above was obtained using split-thicknessstressed that they do not assess the absorption of porcine skin mounted in a finite-dose flow-throughdrug across the stratum corneum barrier [29]. diffusion cell system. In vitro systems are used to

The primary connotation of the term in vitro in obtain parameters for solving Fick’s Law. Thetransdermal studies is the use of skin sections surface concentration is known and receptor con-removed from an animal to assess transport in centrations are monitored until steady-state fluxes arediffusion cell systems. These have been extensively obtained. Knowing the partition coefficient anddescribed elsewhere [30,35]. In these systems, a thickness of the skin, and by measuring the steady-section of skin is obtained from an animal donor state flux, the equation can now be solved for theeither using a dermatome, which is a slicing ap- diffusion coefficient.paratus that cuts sections of skin parallel to the As can be appreciated, in vitro systems aresurface, to produce a split-thickness preparation or economical to conduct and generate ‘clean’ databy simply dissection to obtain a full-thickness prepa- amenable to quantitative analysis since the confound-ration. This membrane is then mounted in a two ing variability inherent to the in vivo setting ischambered diffusion cell apparatus. The donor half absent. If absorption through the stratum corneum isfaces the epidermal surface of the membrane while the rate-limiting step in percutaneous absorption,the receiver half of the cell faces the dermis. In many these methods are appropriate. However, if thecases where only penetration through the rate limit- drug’s pattern of absorption is dependent upon othering stratum corneum is being studied, frozen skin biological functions such as blood flow, dermalmay be thawed and mounted in the cells. Freezing distribution, etc., then these systems may generatedoes not appear to effect the integrity of the stratum misleading results. Another concern is that in full-corneum barrier. thickness and split-thickness preparations, significant

The receiver solution may be contained in one of dermis remains that may be a barrier to lipophilictwo configurations: static where the cell is filled with drugs and even serve as an artificial reservoir. Toa fixed volume solution and samples are taken from a overcome this, some investigators have used tech-port; or flow-through where the receptor solution is niques that split the epidermal–dermal junction (heatconstantly perfused under the dermis and samples separation, enzyme digestion). However, if the skincollected after they exit the system. The donor being studied has significant hair density, then thecompartment may be left open to the environment resulting epidermal membrane may have holes fromand drug deposited onto the surface ( finite dose), or where the hair once exited to the surface. Anotheralternatively the donor half of the cell may be also limitation is that prolonged perfusion may result infilled with liquid containing dissolved drug (infinite membrane degradation due to cellular death. Thisdose). In the latter case, the skin is fully hydrated generally does not occur until after 24 h, howeversince it is being bathed in donor solution. Receptor histological examination of in vitro skin sectionssolutions generally consist of saline in pharmaceu- demonstrates epidermal changes as early as 8 h [36].tical studies assessing hydrophilic drug absorption, Recent developments in this field have been theor buffered albumin or cell culture media in tox- use of artificial skin membranes to assess absorptionicological studies where lipophilic drugs are being across human skin. Such systems are organ culturesstudied. Some investigators have added other sol- which possess intact stratum corneum but no appen-vents to the receptor phase to promote solubility of dages [37]. Studies conducted to date suggest that


permeability through these systems compared to IPPSF is well suited to assess formulation develop-split-thickness human skin may be ten or more fold ment since transdermal flux can be measured withoutgreater. However, they may be useful to assess the confounding systemic interference. This increasesability of human cells to metabolize applied drugs. the power of detecting differences between drug

In vitro techniques are amenable to study absorp- delivery systems.tion from transdermal patches. In these configura-tions, a finite dose flow-through diffusion cell system 2.4. Transdermal pharmacokineticsis used and the patch is sandwiched against theepidermis before mounting in a diffusion cell. Drug Many workers in transdermal delivery use tech-flux is then monitored as described above. niques that have been developed for other routes of

Our laboratory has developed an alternate in vitro / administration to quantitate drug flux as describedex vivo technique, the isolated perfused porcine skin above when AUC is monitored. However, when theflap (IPPSF), which more closely reflects the ana- concentration versus time profile is to be predictedtomical and physiological state of skin in the intact from in vitro flux data, or the rate of drug absorptionanimal [38]. The IPPSF is a fully vascularized, must be obtained from plasma concentration versusviable model which has been shown to closely time profiles, the techniques of pharmacokineticspredict the absorptive flux profile seen in vivo [39– becomes useful.41], and because of the similarity of pig and human There is no fundamental difference between theskin, to predict in vivo human percutaneous absorp- application of pharmacokinetics to transdermal deliv-tion [42,43]. A single pedicle, axial pattern tubed ery as compared to other routes and a generalskin flap is created on the abdomen of a pig [44]. pharmacokinetics text should be consulted for ap-Two days later, the flap is harvested and the perfus- proaches [28,45,46]. If the absorption of drug is firsting superficial epigastric artery is cannulated and the order and described by Fick’s Law, standard model-preparation transferred to a specially designed iso- ing techniques may be used to estimate K usinga

lated organ perfusion chamber. The flap is main- classical compartment-based ‘sum-of-exponential’tained in a temperature and humidity controlled curve stripping or feathering techniques; or the meanenvironment, and the flap perfused with a Krebs– absorption time (MAT) using statistical momentRinger-buffered solution containing physiological analysis. In many topical preparations, applied doselevels of albumin and glucose. The drug or transder- saturates the absorptive capacity of the skin and amal patch is placed on the surface of this tubed flap significant fraction of dose is not available forand venous effluent collected over the course of an absorption. When expressed as a fraction of doseexperiment. At the termination of a study, the surface absorbed, fluxes will be smaller with higher dosesis wiped, stratum corneum tape strips collected, and since all of the topically applied dose is not availablea core biopsy taken to allow for a complete mass for absorption. However, when expressed as per-balance analysis. meability coefficients, rate constants such as K , ora

The advantage of this model over other in vitro MATs, these will be constant and independent ofapproaches is that the anatomical structure of the dose. Many workers estimate flux from the in vitroskin is maintained and drug-induced alterations in systems described above and calculate K from thesea

blood flow or epidermal metabolism will affect drug data. With a knowledge of the systemic phar-absorptive flux as it does in vivo. Two flaps can be macokinetic model from an intravenous study, plas-raised from a single pig. This model also allows ma concentration profiles can then be estimated. Ansimultaneous monitoring of the venous effluent for excellent reference describing the myriad of ap-production of inflammatory cytokines (e.g., TNFa, proaches possible is the text by Gosh et al. [29].interleukins, prostaglandins) to be used as early These approaches are applicable for topical drugbiomarkers of drug-induced irritation. Relative to preparations, however some modifications must bepredicting systemic absorption, the venous effluent is made when transdermal delivery patches are em-considered as input into the circulation making the ployed. As will be discussed in the next section,IPPSF conceptually a biological infusion pump. The most transdermal systems control the rate of drug


release and result in constant flux expressed as mass robust tool in designing transdermal patches as wellreleased per unit time. In this case, absorption is now as in utilizing them in the clinic. By rearranging thisa zero-order rate and not a first-order fractional rate relation, one can calculate the C obtained with ass

constant. These data should be treated as if delivered specific delivery rate as:from a constant-rate intravenous infusion. For exam-

C 5 Delivery rate /Clss Bple, when statistical moment techniques are used, therate of input is now not a MAT defined as 1/K ; but Similar equations can be developed if stochastica

is rather one-half the total patch release rate. Similar- methods are used to analyze the observed concen-ly, compartmental-based curve feathering techniques tration versus time profiles. These methods calculatemay not be appropriate to estimate a K . Instead, pharmacokinetic parameters using area under thea

methods such as nonlinear deconvolution, Wagner– curve (AUC) and area under the moment curveNelson or Loo–Riegelman methods may be more (AUMC). Since Cl 5D/AUC , one can obtain thisB IV

appropriate. parameter after intravenous dosing and use the aboveIt must be stressed that the sources of variability formulae. The observed mean residence time

due to flux across the skin and systemic disposition (MRT5AUMC/AUC) after application of a trans-are independent. We have explored the sources of dermal patch system is:variability in transdermal systems using the IPPSF MRT 5 MRT 1 (Patch Duration /2)Transdermal Patch IVmodel discussed earlier. Accurate estimates of themean plus variance of plasma concentration versus As can be appreciated, these techniques are powerfultime profiles observed in vivo in pigs may be tools to assess the function of transdermal patches inobtained using so-called boot-strap or full-space veterinary species. To be properly applied, intraven-methods of combining IPPSF and in vivo intravenous studies should be conducted to independentlyous pharmacokinetic parameters [47]. In this ap- obtain estimates of Cl , AUC and MRT so thatB IV IV

proach, all possible combinations of IPPSF efflux the actual patch delivery rate and systemic bioavail-profiles (X) are combined with all observed intraven- ability can be unambiguously determined.ous disposition profiles (Y) to generate XY possibletransdermal concentration versus time profiles. Thestatistical properties of these XY profiles closely 3. Transdermal drug deliverymatches observed experimental studies conductedusing the same transdermal systems on in vivo pigs Use of the skin provides an alternative approach to

Since transdermal patches deliver a constant rate systemic delivery of drugs in veterinary species. Theof drug release once steady-state (C ) concentrations problems which arise are due to the large differencesss

are achieved, it is relatively straightforward to de- in skin structure and function among the animalstermine the actual rate of patch delivery in an in vivo which must be targeted. However, these differencesexperiment if one knows the total body clearance could actually be considered as a prime motivation(Cl ) of the drug after intravenous administration: for developing transdermal patch systems.B

Delivery rate (mass / time) 5 C (mass /volume) 3.1. Topical formulations versus patchesss

3 Cl (volume/ time)BThe primary difference between a topical liquid or

This rate of delivery from the patch can then be gel formulation and a transdermal patch is that thecompared to its manufacturer’s stated rate. However, rate of drug absorption in a formulation is controlledthis requires that Cl be obtained from an intraven- by diffusion through the stratum corneum, while in aB

ous study, ideally at the same dose that the patch patch the rate is controlled by release from theshould have delivered. Systemic effects on Cl due system itself. As discussed above, there are manyB

to metabolism or unique distribution or recycling sources of variability inherent to passive absorptionphenomenon would be detected in the intravenous by Fickian diffusion. In a transdermal preparation,study, making this simple relationship a relatively such variability in absorptive flux will translate to


variability in the systemic concentration versus time (Revolution ). They are all three distinct products.profile obtained. However, this variability can be For example, selamectin is formulated in an alcoholeliminated if the rate of release from the topical base and is absorbed systemically to kill fleas,delivery system is controlled to be less than the heartworms and intestinal parasites for 30 days.diffusive flux across the stratum corneum. In this Fipronil is contained in an oily vehicle which formsscenario, which is operative in transdermal patches, a depot within the hair-follicle allowing continueddelivery across the skin is now truly a zero-order release even in the face of washing. These products,constant rate process controlled by the specific and the other topical insecticides discussed below,design of the patch. It must be stated that controlled have gained market acceptance and clearly demon-release from gel formulations is also possible, how- strate the potential of topical administration forever this approach has not been extensively utilized delivering systemically active drugs.due to the other attributes of patch delivery described Transdermal patch technology would not offer anybelow. benefits for these compounds. A similar situation

The rate of control required for systemic effects is exists with many hormone delivery schemes wherean important variable when selecting whether a efficacy is related to achieving a minimal effectivecontrolled patch or normal formulation is required concentration and adverse effects are not seen with[29]. This is directly related to the nature of the intermittent high fluxes of drugs. Testosterone gelpharmacodynamics of the process being targeted. For delivery systems are presently being developed inexample, if the targeted process requires that serum humans based on this premise.concentrations be maintained within a relatively well There are some very specific limitations to thedefined ‘therapeutic window’ for either safety or types of drugs which can be formulated in transder-efficacy endpoints, patch technology may be im- mal patch systems. The first is that the normal flux ofportant. However, if the therapeutic process being drug across the stratum corneum must be greattargeted only requires that a minimum effective rate enough that patch-control to a lower rate of deliveryof delivery occurs, and intermittent high peak blood still results in sufficient blood concentrations forconcentrations do not result in adverse effects, then efficacy. Poorly absorbed drugs are thus not amen-liquid formulations which are capable of delivering able to incorporation into patches. However, com-this minimal flux may be satisfactory. ponents can be built into the patch which enhance

This scenario is encountered in veterinary medi- percutaneous absorption to the point that delivery iscine where topical application of pesticides (pour- feasible. For example, by their very nature, allons) for systemic control of heartworm, fleas, ticks, patches are occlusive systems that hydrate theand other parasites only requires achieving a minimal stratum corneum and enhance drug absorption bythreshold for efficacy. In fact, in these cases, the this means (recall Fig. 2 above with parathion). Thistarget is often the skin of the remainder of the occurs solely because the patch itself, being essen-animal. tially a plastic sheet adhered to the skin, prevents

Pour-ons provide an easy method of dosing where- water loss and thereby hydrates the underlyingby absorbed drug forms a depot in the skin at the site stratum corneum. This simple fact results in anof application which delivers a low flux of drug to increased rate of delivery which would not bethe systemic circulation, which in turn distributes possible with liquid formulations. Secondly, pharma-back to the skin for a prolonged duration of action. cological enhancers may be incorporated into theIntermittent bursts of chemical flux do not lead to system (e.g., ethanol) that increases diffusive flux.adverse effects and temporary breaks in delivery do Finally, recalling that the rate of topical delivery of anot affect steady-state concentrations. These formu- drug is also related to the applied surface area,lations are thus very effective and have been exten- increased systemic concentrations can be obtainedsively utilized in the animal health market. Three of simply by making larger patches. These three ap-the most popular and commercially successful pour- proaches often increase the rate of percutaneous

on products are fipronil (TopSpot , Frontline ), absorption to a degree sufficient that patch technolo-imadacloprid (Advantage ), and selamectin gy can then control release of the active ingredient.


It is generally accepted that under current opti- clonidine) and a function of the enhancers used.mized techniques in humans, approximately 1 mg of Some approaches to counteract this adverse effecta favorable drug (molecular weight of ,400 Da, are under development and include incorporation ofmelting point .1508C) can be delivered from a anti-irritant compounds into the patch system itself.

21-cm area of skin over a 24-h period [29]. Un- Third, the patch must have sufficient adhesion to thefortunately, data do not exist to translate this work- application site that continuous delivery of druging limit to veterinary species outside of the pig, occurs. Also, in animals compared to people, an areawhich is similar to humans. In fact, these species of the skin to which the patch is applied must bemay have higher inherent drug permeability which shaved or clipped free of hair in order for the patchwould allow increased rates of daily dosages. to adhere. Significant advances in adhesive technolo-

Transdermal patches offer a number of other gy over the past few years have largely eliminatedbenefits, in addition to maintaining control when the this problem for humans.therapeutic index is narrow, over other routes ofdelivery. The most obvious one is that they can be 3.2. Systems developed for human useremoved from the animal when one desires termina-tion of delivery. This cannot be easily accomplished Since there have been no transdermal patcheswith outer routes, including topical formulations. specifically developed for veterinary species, it isSecondly, by virtue of entering the systemic circula- prudent to briefly review human patch developmenttion through the cutaneous vasculature, they com- for insights in to design and potential candidates forpletely circumvent the first-pass hepatic metabolism veterinary applications [29]. Table 2 lists the drugsseen with many oral drugs. Although the skin has the that have been developed for human transdermalability to metabolize drugs, as was seen by our group drug delivery systems. A number of additional drugswith isosorbide dinitrate in porcine skin [48], it is are in various stages of pre-clinical and clinicallimited and easily saturated compared to the ability development, and include albuterol, alprazolam,of the liver to metabolize drugs presented it via the atenolol, buprenorphine, cytarabine, deprenylportal vein. Transdermal delivery, both as a formula- (selegiline), dehydroepiandrosterone, dronabinol,tion or a patch, opens up a non-invasive route of enalapril, eptazocine, ethinylestradiol, isosorbide di-administration that is ideally suited for many drugs nitrate, ketorolac tromethamine, ketotifen, norethin-(e.g., nitroglycerin, scopolamine, testosterone). drone, prazosin, and terfenadine. Systems developedThird, for drugs with very short systemic half-lives for electrically assisted transdermal delivery (e.g.,that require sustained concentrations for therapeutic iontophoresis) for compounds such as peptideseffects, transdermal patches provide the sustained- (LHRH, insulin, etc.), lidocaine and other smallrelease delivery needed for longer periods of time. organic molecules will be discussed below.This results in a major advantage because it is more There are a wide number of approaches to trans-convenient for veterinary hospitals compared to dermal patch design, all of which share the attributesintravenous infusion systems and client compliance depicted in Fig. 4 [29]. All systems consist of ais improved due to the reduced frequency of dosing method to contain the drug in a patch, either directlyrequired.

There are a number of disadvantages of transder-Table 2mal delivery [29], the most obvious being that if aDrugs which have been incorporated into transdermal delivery

drug does not permeate skin, it is not amenable to systems for humanstransdermal delivery. This limitation is absolute

Clonidineunless novel enhancer technologies are developedEstradiol

which increase permeability to poorly absorbed Fentanyldrugs. Secondly, by virtue of the occlusive nature of Nicotine

Nitroglycerinepatch systems, cutaneous irritation is often encoun-Scopolaminetered which limits the duration that a patch can beTestosteroneworn at a single site. This is both drug specific (e.g.,


drug is contained in a reservoir if it is in a gel or aliquid form. However, if the drug is incorporated intoa laminate polymer, then a reservoir is not requiredand the system is a poly-laminated plastic sheet.

All systems require an adhesive layer to adhere themembrane to the skin. It is imperative in selectingthe adhesive that the drug’s release from the systemis not affected by passage through this adhesiveotherwise the rate of delivery may be adverselyaffected. An advance in patch design that has furthersimplified these systems is the development ofpolymer techniques which allow the drug to bedirectly incorporated into the adhesive. Release fromthe polymer controls the rate of delivery. In such asystem depicted in Fig. 4A, once the release liner isremoved, the patch then consists only of the drugcontaining adhesive adhered to a plastic backing.

These differences in patch design have a directimplication to the use of human products in vet-

Fig. 4. Structure of transdermal delivery systems.erinary species. Since the rate of topical drug deliv-ery is related to the area of skin application, vet-

in an adhesive (A), in a reservoir (B), or in a erinarians have used human patches in smallerlaminated matrix (C). All systems have an adhesive species by cutting patches to the required size.component for adhering the patch to the skin, and an Obviously, this can only be accomplished in patchesadhesive release liner which upon removal and where the drug is incorporated either in a matrix orplacement on the skin activates the system. Finally, the adhesive, as reservoir systems would rupture andthe entire system is covered by an impermeable lose their drug reservoir. Some veterinarians alsobacking which is responsible for retaining the drug in remove only a portion of the protective adhesivethe patch, as well as contributing to the occlusive backing ostensibly to expose a smaller surface areanature of these systems discussed above. There are a to the skin. However, at least in cats, this does notwide number of variations upon these themes, and an always seem to affect absorption.even wider number of issued patents to protect There is an extensive literature base and patentproprietary strategies. Space does not allow for a history on the use of transdermal delivery systems incomplete description of all developed systems. How- humans, which is adequately reviewed in severalever, this is not needed for the purposes of the texts [29,49]. Many of the principles outlined in thepresent review since they can be conceptually under- above discussion are based on this work. Thesestood using the three approaches depicted. references should be consulted for details on specific

As discussed above, the defining characteristic of systems and their applications in clinical humana transdermal patch compared to a topical formula- medicine.tion, is the controlled rate of release of drug from thesystem. This is achieved in many systems through 3.3. Percutaneous absorption in veterinary speciesthe use of a rate-controlled membrane separating thedrug from the skin surface. Diffusion through this The reader should now have a reasonable under-membrane, described by Fick’s Law of Diffusion standing of the mechanism of drug absorption acrosspresented above, provides the controlled rate of drug the skin and the factors that may modulate transder-release which is then absorbed across the skin which mal flux. The important aspects of comparativehas a greater permeability for the drug than the dermal anatomy and physiology, which could impactmembrane. In membrane-controlled systems, the on transdermal drug flux, have been reviewed.


Finally, the characteristics of available transdermal targets. Many of these products are formulated aspatch technology, as well as the factors which make ‘pour-on’ products which contain dilute solutions ofpatches different from liquid topical formulations, the active ingredient; or as ‘spot-on’ low volumehas been presented. formulations containing concentrated active ingredi-

The remainder of this review will focus on aspects ents to promote systemic absorption. Three effectiveof transdermal delivery, which highlight the unique topical pour-on products that were discussed earlierproblems caused by anatomical and physiological include fipronil, imidacloprid, and selamectin. So-differences seen in veterinary species. There are two called ‘inert’ ingredients include glycols, ethers,sources of information available to address these alcohols, hydrocarbon oils, amides and variousissues. The first is a reasonable body of literature on spreading agent that facilitate chemical delivery tothe percutaneous absorption of topically applied the skin surface, and may act as penetration en-pesticides in veterinary species which shed light on hancers [52,53].comparative aspects of dermal absorption, and clear- What is clear from an analysis of availablely demonstrate that this route of administration is literature, is that the solvent system used in topicalefficacious for systemic targets. The second are formulations may have a controlling influence overreports on the use of human transdermal patches in the rate and extent of active ingredient absorption; inveterinary species, which offer direct evidence of some cases acting as controlled-release systems. Intheir efficacy as well as contrast differences between the formulation of these products, it is hypothesizedhuman use for which they were designed, and that systemic concentrations are required for effica-veterinary species. cy, although there is very little information on what

actually constitutes an effective blood concentration,3.3.1. Topical pesticides a situation very different from biopharmaceutics.

There has been work done relative to the percuta- However, it is clear from a wealth of field experienceneous absorption of topically applied pesticides in with many products in many species, dermal ad-veterinary species, both from an animal health ministration is an efficacious route of administrationproduct perspective as well as environmental expo- for these products.sure. This is the only body of literature which clearly Organophosphates are amongst the most widelyindicates that there are tissue residue concerns with used topical insecticides in domestic animals. Der-topical applications that may impact food safety [50]. mal exposure may result in inhibition of bloodRecently, we have reviewed this area [51] and will cholinesterase activity and produce neurologicalhighlight important concepts relative to the present signs of toxicosis, two clear indications that systemictopic. A large literature base exists on the compara- absorption has occurred. It is interesting to note thattive absorption of chemicals in laboratory animal in most species, there is a rapid onset of clinicalspecies. However, due to the very small mass of signs after topical application of lipophilic or-these animals, in vivo data may be misleading when ganophosphates such as chlorpyrifos, except in thethey are metabolically scaled up to larger species. As cat where onset is delayed [54]. Fenthion is approveddiscussed earlier, chemical flux through mouse, rat in dogs, beef cattle, and non-lactating dairy cattle inand rabbit skin is much greater than that of humans concentrations ranging from 3 to 20%. In cattle, theand other species. tissue withdrawal time is 45 days which clearly

The insecticides which have been or currently are proves dermal absorption and systemic distribution.approved for topical application in veterinary species Phosmet is approved for use in dogs, swine andinclude chlorinated hydrocarbons (e.g., methox- beef cattle. In pigs, dermal bioavailability was lessychlor, lindane), organophosphates (e.g., fenthion, than 3% of the applied dose administered in threefamphur), carbamates (e.g., carbaryl), pyrethrins different vehicle systems [55]. This type of data is(e.g., permethrim, fenvalerate) and antihelmenthics conspicuously absent from other studies makingsuch as ivermectin and levamisole. Most other interspecies comparisons difficult. Famphur is ap-topicals used in veterinary medicine are applied for a proved in cattle with a meat tissue withdrawal timelocal dermatological effect, and not for systemic of 35 days. In a study of subcutaneous fat biopsies


following dermal application of 25, 50 or 150 mg/kg with greatest absorption and metabolism occurring infamphur in an oily vehicle, peak residues occurred at the back compared to abdomen. Absorption can be1 day of exposure and were insignificant by 11 days modulated by a number of coadministered com-for all doses. Coumaphos at a dose of 14 mg/kg pounds, including fenvalerate, para-nitrophenol,given as a pour-on to goats was slowly absorbed at a paraoxon, sodium lauryl sulfate, methyl nicotinate,constant rate for 7 days, with 45% of the dose stannous chloride and different solvent vehicles.remaining unabsorbed [56]. Chlorpyrifos is used in When combinations of these are dosed, complex butflea collars for cats and dogs, dips for dogs, and in reproducible patterns of absorption are seen. In someear tags for cattle. Studies tend to indicate that cases, systemic absorption does not parallel changesabsorbed chemical has the highest concentration in in dermal depot concentrations. Such studies clearlyfat which in swine results in detectable residues for 3 indicate that parathion absorption, as well as in someweeks. As discussed above, absorption in cats seems cases cutaneous biotransformation, can be modulatedto be different from other species. by co-administration of other compounds and

It is instructive at this point to reiterate the changes in site of application as well as environmen-difference between dermal penetration into the skin tal conditions.and absorption into the systemic circulation. This Carbaryl is the most widely used carbamatedifference is clearly illustrated from in vivo and in insecticide in veterinary medicine. Unlike many ofvitro studies of the disposition of topical piroxicam the earlier compounds where percutaneous absorp-in pigs [57]. A piroxicam (anti-inflammatory drug) tion was not specifically measured, studies do existtopical gel was applied to cranial or caudal ventral for carbaryl in rodents and domestic species. Car-abdominal body sites, differing only by the nature of baryl is extensively (50–95%) absorbed in rodents.the cutaneous vasculature perfusing the skin. The Studies conducted in the IPPSF in pigs demonstratedcranial site is perfused by penetrating musculocuta- 10% absorption in 8 h, with 23% of this absorbedneous vasculature while the caudal site is perfused dose being the metabolite 1-naphthol [23]. Phar-by direct cutaneous vasculature that does not perfuse macokinetic extrapolation of the total bioavailabilitydeeper tissue beds before reaching the skin. In vitro was 33% after 6 days, supporting the existence of aflux across skin from both sites was identical, dermal reservoir that continually releases drug afterhowever cutaneous deposition at the cranial sites was absorption. Baynes and Riviere [52] studied themuch greater than the caudal sites. Thus, dermal effects of ‘inert’ ingredients and metabolites on thedepots would be expected to be formed after topical absorption of carbaryl and demonstrated clear cutapplication to cranial sites perfused by musculocuta- enhancing effects of solvents, piperonyl butoxide,neous vessels, despite the fact that trans-epidermal and 1-naphthol. These types of studies illustrate whyflux was equivalent. Similar phenomena occurring this literature is difficult to evaluate, since all pes-with other drugs could confound interpretation of ticides studied are in commercial formulations ofdata when different body sites are used. various composition, making differences in absorp-

Extensive work has been done with parathion in tion difficult to correlate to drug, species or formula-swine as an animal model for assessing human tion.absorption [5,22–25,58–63]. In intact pigs and the Some interesting data pertinent to transdermalIPPSF, approximately 7–14% of a topically applied issues have been reported for the chlorinated hydro-dose is absorbed systemically. This is increased carbon lindane absorption in veterinary species.under occlusive conditions as well as with increased Dermal absorption of lindane is relatively slow andrelative humidity as depicted earlier in Fig. 2. occurs to a limited extent. In the IPPSF [23], 2% ofIncreased temperature and perfusate flow increases the dose of lindane was absorbed after 8 h whichabsorption in in vitro diffusion cell studies. Dermal extrapolated to a 6-day absorption of 7.6%. In abiotransformation to paraoxon and para-nitrophenol separate study, fenvalerate enhanced lindane absorp-has been demonstrated using both in vitro and in tion [58]. In sheep dipped in 0.025% lindane, 8–10%vivo models. Both degree of absorption and extent of of the applied dose was absorbed and fat residuesbiotransformation are body-site specific in the pig, were violative at 28 days but finally declined to safe


levels by 70 days. Absorption was bi-phasic, with the use of fentanyl patches in dogs and cats and the usemajority of the dosed absorbed by a slower process of the pig as a selective animal model for humanwith a half-life of 169–200 h, and the remainder by a delivery systems, typified by the development ofmore rapid process. A significant difference was transdermal testosterone systems. Since the authorsnoted in the absorption kinetics of shorn versus have extensive first-hands experience with both, thisun-shorn sheep. A deconvolution analysis was per- will be the focus of this section.formed on the time required to absorb 50% of theavailable dose. This was determined to be between 3.3.2.1. Fentanyl115 and 179 h and was hypothesized to be due to a Human fentanyl transdermal delivery systemsslow release of lindane from the stratum corneum were investigated for their use as a convenient post-reservoir in the skin [64,65]. surgery analgesic in dogs. Fentanyl (N-phenylethyl-

Levamisole is formulated as a 20% pour-on in N-(1,2-phenylethyl-4-piperidyl) propanamide) is adiethylene glycerol monobutyl ether. Several studies synthetic opioid with strong m and d opiate receptorsuggest that this formulation achieves effective blood agonist activity. In the dog, fentanyl elimination isconcentrations in cattle. The extent of absorption is rapid requiring intravenous infusion to maintaininfluenced by environmental temperatures, with in- constant plasma concentrations needed for effectivecreased absorption in summer over winter. In addi- analgesia. Kyles et al. [66] compared the dispositiontion to a direct temperature effect, seasonal changes of fentanyl in dogs after intravenous and transdermalin skin structure and hair coat length may be factors administration using a commercial human patchcausing this seasonal effect. (Duragesic-50 Fentanyl System; membrane-con-

These studies clearly indicate that systemic ab- trolled alcohol-based reservoir system) designed tosorption occurs after topical application of pesticides deliver 50 mg/h for 72 h. When applied to thein a wide range of veterinary species. Although clear clipped back (between shoulder blades) of Beaglescut quantitative studies specifically assessing absorp- weighing an average of 13.5 kg, the patch deliveredtion in multiple species are difficult to find, since the an average of 36 mg/h which was 72% of the statedpurpose of many of these published studies was food rate of delivery. This was sufficient to maintain ansafety and not quantification of absorption, systemic average steady-state concentration of 1.6 ng/ml fromexposure occurs as evidenced by efficacy in treat- approximately 24 until 72 h when the patch wasment and occurrence of tissue residues at slaughter. removed. The AUC in these dogs was 102 ng-hr /mlMany factors including temperature and body site per ml. After patch removal, plasma fentanyl con-modulate absorption and the formation of epidermal centration rapidly decayed at a half-life that corre-and dermal depots. These factors have impact on the lated to the initial distribution half-life seen withdesign of transdermal systems in veterinary species. intravenous administration. The plasma concentra-

tion versus time profile of fentanyl applied to six3.3.2. Transdermal patches beagle dogs is depicted in Fig. 5.

The above review indicates that there are signifi- When compared to pharmacokinetic studies con-cant differences in topical pesticide absorption be- ducted in humans with the same transdermal deliverytween species. Some of this difference is related to system, dogs achieve effective plasma concentrationsactual flux across the stratum corneum which can be relatively rapidly apparently due to the lack of aassessed by in vitro techniques. However, some dermal depot for fentanyl in dogs compared todifferences are related to dermal factors such as humans. This hypothesis is consistent with the moredifferential perfusion and the tendency to form rapid decline in plasma concentrations after patchdermal depots. As will be seen, these factors may removal in dogs compared to a prolongation inplay a role in transdermal patch design in veterinary half-life to greater than that seen with intravenousspecies. administration in humans. This is believed secondary

As indicated earlier, there is not a great deal of to a slow release of fentanyl from a dermal depot.literature on the use of transdermal patches in The efficacy of transdermal fentanyl patches inveterinary medicine. The clear cut exception is the dogs was subsequently evaluated for analgesia un-


Fig. 5. Transdermal delivery of fentanyl in dogs (mean6S.E.M.).

dergoing ovariohysterectomy [67]. The same 50-mg/ varied. As in the first post-surgery study, analgesiah patch was applied to 10 dogs of an average weight was effectively maintained and in this study, wasof 21 kg on the dorsal area of the neck 20 h before more effective then epidural morphine. In the threesurgery to assure that steady-state concentrations studies cited above [66–68], the assay was identicalwere achieved before surgery. The efficacy of fen- and performed in the same laboratory.tanyl was compared to a randomized treatment group A final study of transdermal fentanyl administra-of 10 dogs given oxymorphone intramuscularly (2.5 tion in dogs was conducted by a different group of

2mg/meter every six h starting immediately pre- investigators [69] and compared plasma concentra-surgery). In this trial, steady-state plasma fentanyl tions achieved after use of three different sizeconcentrations of 1.2 ng/ml were achieved between patches (50, 75 and 100 mg/h) in a cross-over19 and 46 h after patch application. This study experimental design. Patches were applied to theconcluded that transdermal fentanyl delivered as clipped lateral thorax of six dogs weighing aneffective post-operative analgesia as did the accepted average of 20 kg. Steady-state plasma concentrationsinjectable oxymorphone protocol, but was associated were achieved in 24 h and averaged 0.7, 1.4, and 1.2with less sedation. Similar to the previous phar- ng/ml with 50-, 75- and 100-mg/h patches, respec-macokinetic study, steady-state concentrations were tively. AUCs were 46, 101 and 80 ng-hr /ml per ml,on-average well maintained but significant inter-in- respectively. Some skin irritations were observed individual animal variation was reported. Another these dogs, however they were related to dose oranalgesia efficacy study was conducted in dogs individual animal. This study also demonstratedundergoing major orthopedic surgery [68] using 100- considerable inter-individual variability, as is alsomg/h patches in dogs weighing an average of 23 kg. observed in humans which is purported to be sec-These patches were applied to a clipped area of skin ondary to differences in skin and body temperatures,between the scapulae (dorsal shoulder). Again, hydration status, sweat gland function, ethnic group,steady state was reached in 24 h and were approxi- and the state and integrity of the stratum corneummately 0.95 ng/ml, but individual steady-state values [70]. It should be noted that the core body tempera-


ture of dogs is a few degrees greater than that of body sites of application and different breeds ofhumans. dogs. In the study evaluating delivery from three

These studies in dogs suggest that transdermal patch sizes, there was not a linear relation betweenfentanyl is an effective analgesic for post-surgical stated patch delivery rate and observed C or AUC,ss

pain. Although effective steady-state plasma con- suggesting either saturation of percutaneous absorp-centrations are maintained, there is considerable tion at the higher fluxes or altered systemic elimina-variability in their level. This can be seen in Table 3 tion.which compares the mean concentrations or AUCs in Lee et al. [71] compared fentanyl concentrationsthe above studies in dogs administered the same achieved after application of a 25-mg/h patch to the50-mg/h patch. The highest concentrations and clipped dorsal cervical region of six cats with anAUCs are achieved in the smallest dogs, which could average weight of 3.8 kg. Steady-state concentrationsbe partially predicted by normalizing patch delivery were achieved in 12–18 h and maintained at approxi-rate to body weight. If actual patch delivery rates mately 1.9 ng/ml for 100 h as seen in Fig. 6. Twowere available in all studies, further variability may cats in this study did not achieve measurable fentanylbe explained. Additional variables include different concentrations. However, when they were repeated,

Table 3Steady-state plasma fentanyl concentrations (C ) and AUC in dogs administered a single 50-mg/h transdermal patchss

Weight Rate /kg Body site C AUC Studyss

(kg) (mg/kg-h) (ng/ml) (ng-h /ml)

13.5 3.7 Thorax 1.6 102 6620 2.5 Dorsal neck 0.7 46 6921 2.4 Thorax 1.2 – 6723 2.2 Dorsal shoulder 0.95 – 68

Fig. 6. Transdermal delivery of fentanyl in cats (mean6S.E.M.).


delivery occurred and they were then included in the In conclusion, these studies on fentanyl adminis-calculation of mean values. Intravenous fentanyl tration in dogs and cats suggest that transdermalpharmacokinetic studies were also conducted which patches may be an effective route of drug administra-allowed the actual delivery rate from these patches to tion in veterinary species. Adequate plasma con-be estimated at only 8.5 mg/h using ratios of AUC to centrations are obtained and effective analgesiacalculate bioavailability as presented in Section 2.3.2 maintained for two types of surgical procedures. Theabove. In contrast to dogs, but similar to humans, patches are now widely used clinically in dogs andplasma fentanyl concentrations did not decline after cats for indications ranging from post-operative painpatch removal. This was not due to systemic phar- to cancer pain. The convenience compared to multi-macokinetic properties, since the observed intraven- ple injections is obvious. However, the studies alsoous half-life in cats of 2.4 h was shorter than the 6-h indicate significant degrees of inter- and intra-specieshalf-life reported in dogs. In contrast this suggests variability. These issues will be discussed below.the presence of a dermal depot in cats whichprolongs elimination due to slow absorption after 3.3.2.2. Testosteronepatch removal. In order to illustrate that transdermal patches may

In contrast to dogs, the lower rate of delivery be efficacious in species other than dogs and cats,compared to the stated patch release rate suggests Fig. 7 illustrates the blood concentration versus timethat feline stratum corneum is less permeable to profile seen after application of two human testo-fentanyl than that of canine or human tissue, or sterone transdermal delivery patches applied to thealternatively the interface between the patch adhesive abdomen of four pigs. (Androderm, combined 15-

2and skin surface is not uniform or another interaction cm surface area containing a total of 24.4 mgis present. There is not a straightforward correlation testosterone.) The concentrations achieved are in theto anatomical or physiological variables tabulated physiological range for efficacy. Compared to humanearlier (Table 1). As discussed earlier, it has been delivery, in vivo transdermal flux from these systemspostulated that the major anatomical variable relating is 2–3-fold less than humans. In contrast, withto stratum corneum absorption is the size of in- radiolabelled testosterone administered as a topicaldividual stratum corneum cells, a metric that has not solution to both pigs and humans, the fractionbeen determined for either cats or dogs. Alternative- absorbed was closer at 9 and 15% of applied dose,ly, the structure and arrangement of hair follicles in respectively. As discussed earlier, absorption in pigsfeline and canine skin are different, which may and humans is generally similar with no clear cutmodulate the effective delivery rates achievable with trend of over- or under-estimation evident when athese systems. Cats also have less active apocrine wide series of compounds are studied. Thus absorp-sweat glands than dogs which may modify the tion, as seen with all compounds discussed above, ispatch–skin interface [3]. In vitro studies are not compound, site, species and most likely patch /available to further probe these mechanisms. formulation specific.

Finally, transdermal fentanyl (50 mg/h) patcheswere applied to the clipped neck of eight goats 3.3.3. Iontophoresisweighing an average of 40 kg [72]. In contrast to There is another approach to transdermal deliverycats and dogs, steady-state concentrations of fentanyl which is receiving attention in the literature aswere not achieved in goats, however plasma half-life systems for human drug delivery are presently underafter patch removal was 5.3 h compared to the 1.2 h development. This relates to electrically assistedseen after intravenous administration. Similar to transdermal drug delivery systems typified by ion-other species, there was significant inter-individual tophoresis. As of the date of this review, no ion-variations. Peak concentrations were greater than tophoretic systems are approved for use in humans;those reported in other species. Curiously, total drug however, pigs have been extensively used in theirdelivery from the patches was calculated to be development. Some systems have been marketed ingreater than the stated delivery rate by the manufac- veterinary medicine; however, there is no publishedturer. The cause of this discrepancy is not clear. data either on their ability to deliver drugs transder-


Fig. 7. Transdermal delivery of testosterone in pigs (mean6S.E.M.). Used with permission of Cellegy Pharmaceuticals, Inc.

mally or on their clinical efficacy. The nature of the will promote the delivery of positively charged drugsformulations used in these systems would question (cations), while a negative current (cathodal) willboth endpoints. promote the delivery of negatively charged drugs

Electrically assisted transdermal drug delivery (anions). Flux is directly proportional to the totaldescribes systems where the drug flux across the charge a drug carries and inversely proportional to itsstratum corneum is driven by an electrical gradient molecular weight. Total drug flux is also dependentrather than the concentration gradient which drives upon all of the ionic compounds included in thediffusion. There are a number of mechanisms where- formulation, thus inclusion of a charged additive willby drug flux is potentiated using electrical energy, decrease delivery of the target drug. Formulation andincluding iontophoresis, electro-osmosis and elec- pH may also effect efficacy of transport by mecha-troporation. For the purposes of this review, ion- nisms very different than passive systems. If atophoresis will be presented for illustrative purposes. positive charge is placed on the skin, water willReviews of this area should be consulted for further move across the skin due to a process called electro-details [73,74] osmosis, which allows for the delivery of neutral

The two major advantages of iontophoretic sys- compounds by bulk-flow. Finally, iontophoresistems over passive diffusion-driven systems are the tends to increase permeability of the stratum cor-control achieved by linking drug flux to the mag- neum to large and/or charged drugs that cannot benitude of the applied electric current, and the ability delivered by passive diffusion. The delivery ofto deliver polar molecules (including small peptides) peptides has been a major focus of this research.which are not amenable to diffusion systems. Addi- Like the transdermal patch technology reviewedtionally, lag times do not occur making onset of drug above, the iontophoretic electrode controls the rate ofaction in the systemic circulation more rapid. Drug delivery across the skin, not the stratum corneum.flux is proportional to the applied current per unit of Lag times are eliminated leading to much shorter

2surface area (mA/cm ) as described by the Nernst– onset times than are seen with patch technology.Planck equation for membrane transport of ions Fig. 8 depicts the in vivo delivery of the peptideunder an electric field. A positive current (anodal) LHRH by anodal iontophoresis in pigs. The observed


Fig. 8. Iontophoretic transdermal delivery of LHRH in pigs (mean6S.E.M.).

profile is very similar to that seen in humans. 4. Transdermal patches in veterinary speciesSignificantly, these concentrations of deliveredLHRH caused an increase in LH in female pigs and The above review, although not comprehensive,FSH in male pigs, clearly demonstrating that ion- clearly indicates the potential for this route oftophoretic peptide delivery in swine is efficacious administration in veterinary medicine and identifies[75]. Similar efficacious delivery of lidocaine has several considerations which will now be explored.also been demonstrated [76], as was depicted in Fig.3 above. Significantly, there was an exact correlation 4.1. Important considerations unique to veterinarybetween IPPSF predicted and observed human plas- medicinema concentration profiles of iontophoretically de-livered arbutamine [42]. Combining iontophoresis There are a number of factors that must bewith electroporation, which is very short high-volt- considered in designing transdermal patches forage pulsing of skin, further increased LHRH deliv- veterinary applications. Most of these have beenery, decreased onset time, and significantly reduced alluded to in the discussions above but will now beinter-individual variability [77]. summarized in a conceptual context.

Transdermal iontophoretic drug delivery is another The rate of drug delivery in a transdermal patch ispotential administration strategy for use in veterinary designed to be controlled by the patch, and not themedicine. Drugs which cannot easily be delivered by stratum corneum. However, based on the data fromother routes have the potential to be delivered by this fentanyl above, it is obvious that human patchesstrategy. Unlike passive systems, drug flux only applied to dogs, cats and goats do not perform in thisoccurs when electrical current is present which opens fashion. A similar situation occurs with testosterone.the door to the very precise control of drugs with This suggests that there are a number of variablesvery narrow therapeutic windows. More importantly, that may affect transdermal patch design in vet-peptides can also be delivered by this route. erinary medicine which could modulate drug release


from the patch, penetration across the stratum cor- patch sizes for a cow or horse would be unrealistic,neum, and/or absorption into the systemic circula- even though larger surface areas of skin are avail-tion. These include, but are not limited to: able. However, if a very low effective plasma

concentration is required (e.g., pesticide, growth1. reduced permeability through the stratum cor- promotant), then patch technology may be appro-

neum resulting in a rate-limiting diffusion; priate.2. improper adhesion of the patch on animal skin; The potential flaw in this logic is the increased3. interaction of the patch adhesive with surface rate of drug clearance seen with animals of smaller

lipids changing the diffusional properties of the body masses, a phenomenon that underlies the use ofsystem; allometric relations to scale drug disposition across

4. different pH of skin surface; species [78]. That is, a greater dose on a mg/kg basis5. different surface density and structures of hair, may be required to achieve efficacy in a cat com-

sebaceous and sweat glands; pared to a cow. One advantage to patch technology6. differential depot formation in the stratum cor- in food producing animals would be clear demarca-

neum and/or dermis which affects the time to tion of the application site for tissue residue avoid-steady state as well as the time for blood con- ance.centrations to decay after patch removal; An additional variable is the systemic clearance of

7. different skin and body temperatures, and mecha- the compound, which as described above, willnisms for their regulation, which could affect both determine the steady-state concentrations that couldrelease from patches and well as transport through be achieved. For drugs with high clearances, highskin; patch delivery rates are required. However, for drugs

8. anatomical differences in skin, differing rates of with relatively low clearances, the patch delivery ratecutaneous blood flow and/or patterns of dermal could be low and still achieve effective concen-perfusion (shunts, etc.) that would control deliv- trations. This is the situation seen with the marketedery to the systemic circulation; pour-on products discussed above. These issues also

9. species-specific cutaneous biotransformation; must be viewed in terms of the pharmacodynamics10.formulation factors which are important for all of the drug. If a steady-state blood concentration

routes of delivery. with minimal fluctuation is desired, patches may bethe preferred mode of administration. In contrast, if

Another issue unique to veterinary medicine that wide fluctuations or immediately effective levels areconfounds the above considerations is the wide range needed, passive transdermal patches may not beof body sizes seen both within and among species. appropriate. In this situation, iontophoretic systemsUnlike humans, where body mass may only vary by may be ideal.a factor of 2–3-fold, domestic animals range from When a decision is made to explore the develop-the few-kilogram cat to the 500-kg cow. It is obvious ment of a transdermal patch for veterinary species,that a single patch could not be developed that would there are a number of experimental studies thatuniformly serve all species. Even with a species such should be conducted.as a dog, body sizes may range for a few to 50 kg.

The important variable in this regard is the ratio of 1. The ability of a drug candidate to be absorbedpatch area to total body mass, which sets limits on across the species of interest should be evaluatedthe amount of drug which can be delivered using in a validated in vitro model. This would assesstechnology available today. Assuming that the 1-mg/ the ability of the drug molecule to penetrate the

2cm per 24 h rule of thumb is accurate, it will be skin, a prerequisite for any transdermal deliveryeasier to develop transdermal patches for application system; as well as generate basic parametersin feline and canine medicine than in large animal governing diffusion.

2practice. A relatively small patch of 10 cm can 2. Initial formulation studies should be conducted indeliver almost 4 mg/kg per 24 h to a cat and 0.5 vitro to optimize drug flux in a more controlledmg/kg per 24 h to an average 20-kg dog. Equivalent environment.


3. Determine intravenous pharmacokinetic parame- be placed in locations where the animal cannot reachters to allow simulation of blood concentrations it; otherwise they will lick, chew, and scratch theachievable with this drug. Compare these to drug application site. Ease of dosing would be achieved ifconcentrations required for efficacy. adhesive-based, rather than reservoir transdermal

4. Conduct simple in vivo absorption study to patches were formulated for veterinary use (e.g., Fig.validate in vitro system with this specific drug. 4A or C). This would facilitate obtaining correct

5. Develop transdermal patch formulation, using the dosing in different species of animals simply byabove validated model systems, to deliver desired cutting the patch material to an appropriate area.delivery rate. In vitro drug release assays should Based on patch size to body mass considerations,also be used at this step to assess uniformity of it would appear that cats and dogs are ideally suiteddelivery. for development. Pigs, goats and sheep are similar in

6. When a prototype transdermal patch is developed, area to mass ratio as humans and may thus beassess inter-site and inter-species delivery and appropriate. Patch development in cattle and horsesdevelop an understanding as to the parameters may only be feasible for very potent drugs orwhich modulate drug delivery. compounds where minimal exposure (hormones) is

7. Assess patch performance under varied environ- efficacious.mental (temperature, activity level, etc.) andapplication (multiple patch placement on same 4.3. Limitationsand different sites, etc.) techniques.

Some drugs will never be appropriate for transder-The important concept to consider is that vet- mal patches simply because of their physicochemical

erinary species are a diverse group of animals that properties (too large, too charged, insufficient lipidmay show differences in drug delivery rates that are solubility, tendency to cause direct skin irritation) ornot immediately obvious. These factors should ideal- unfavorable pharmacokinetic or pharmacodynamicly be known before any drug development progresses behavior (too rapid of a clearance relative to achiev-to a ‘point of no return’. able rate of skin delivery, first-pass cutaneous bio-

transformation, requirement for intermittent high4.2. Potentials peak and low trough blood profiles, insufficient

potency). Some compounds, such as pour-on pes-There are a number of exciting potentials for ticides are optimally formulated with existing topical

development of transdermal drug administration in products which obviates the need to develop rate-veterinary species. The most obvious is ease of controlled patch delivery systems. As seen with thedosing small animals that may be difficult to ad- limited experience with a single drug fentanyl,minister drugs to by other routes. The best drug species-specific absorption occurs by mechanismscandidates would be those that otherwise require which at this point are not well understood.intravenous infusion, frequent administration, or that The acceptability of patches in veterinary practicehave poor oral systemic availability. Owner com- must overcome the potential behavioral patterns ofpliance is easier to maintain due to the decreased animals that remove objects form their skin thatfrequency of dosing and transdermal patches would irritate it. Thus patches must be designed which arebe beneficial for treating species that resist being not amenable to removal by scratching, biting ormedicated (e.g., cats). If specific veterinary patches licking. The use of patches in food-producing ani-are developed with geometries and adhesives tested mals would break new regulatory ground, as con-in the relatively hairy skin of animals, optimal cerns of dermal depot formation with slow drugdelivery may occur as compared to using a human release might cause concern for violative tissuepatch designed to be adhered to glabrous skin. residues. However, if a permanent dye were incorpo-Because of certain animal behaviors (especially cats rated in to a patch design, animals receiving patchesand dogs), patches must be formulated to be non- would be marked which would allow excision of thisirritating, capable of tight adhesion to skin, and/or depot at slaughter.


measurements at five cutaneous sites in nine species, J.5. ConclusionsInvest. Dermatol. 95 (1990) 582–586.

[11] M.J. Bartek, J.A. LaBudde, H.I. Maibach, Skin permeabilityThe development of transdermal patches for vet-in vivo, J. Invest. Dermatol. 58 (1972) 114–123.

erinary medicine appears to be a feasible scenario for [12] W.G. Reifenrath, E.M. Chellquist, E.A. Shipwash, W.W.the future. However, before extensive product de- Jederberg, Evaluation of animal models for predicting skin

penetration in man, Fundam. Appl. Toxicol. 4 (1984) S224–velopment occurs, considerable experimental inves-S230.tigations are required in the primary target species to

[13] R.C. Wester, H.I. Maibach, In vivo animal models fordefine some basic parameters of drug absorption.percutaneous absorption, in: R.I. Bronaugh, H.I. Maibach

The data reviewed to date leave many unanswered (Eds.), Percutaneous Absorption, 2nd Edition, Marcelquestion but very specific areas for study. Transder- Dekker, New York, 1989, pp. 221–238.mal delivery is feasible as demonstrated by the use [14] I.H. Pitman, S.J. Rostas, Topical drug delivery to cattle and

sheep, J. Pharm. Sci. 70 (1981) 1181–1194.of fentanyl patches in dogs and pour-on products in[15] B. Berner, Pharmacokinetics of skin penetration, in: A.F.all species. It is obvious that development of patches

Kydonieus, B. Berner (Eds.), Transdermal Delivery ofwill be drug and species specific. With the rapidDrugs, CRC Press, Boca Raton, FL, 1987, Chapter 9.

advancement of patch technology for humans, the [16] G.L. Flynn, Physico-chemical determinants of skin absorp-know how is presently available to develop effective tion, in: T.R. Gerity, C.J. Henry (Eds.), Principles of Route

to Route Extrapolation for Risk Assessment, Elsevier, Am-transdermal patches for at least canine and felinesterdam, 1990.therapeutics.

[17] R.H. Guy, J. Hadgraft, Mathematical models of percutaneousabsorption, in: R.L. Bronaugh, H.I. Maibach (Eds.), Percuta-neous Absorption: Mechanisms, Methodology and Drug

References Delivery, 2nd Edition, Marcel Dekker, New York, 1989, pp.13–26.

[18] K. Tojo, The prediction of transdermal permeation: mathe-[1] N.A. Monteiro-Riviere, Comparative anatomy, physiology,matical models, in: T.K. Ghosh, W.R. Pfister, S.I. Yumand biochemistry of mammalian skin, in: D.W. Hobson (Ed.),(Eds.), Transdermal and Topical Drug Delivery Systems,Dermal and Ocular Toxicology: Fundamentals and Methods,Interpharm Press, Buffalo Grove, IL, 1997, pp. 113–138.CRC Press, Boca Raton, FL, 1991, pp. 3–71.

[19] B.W. Barry, in: Dermatological Formulations: Percutaneous[2] N.A. Monteiro-Riviere, Anatomical factors affecting barrierAbsorption, Marcel Dekker, New York, 1993.function, in: F.N. Marzulli, H.I. Maibach (Eds.), Dermato-

[20] R.O. Potts, R.H. Guy, in: Mechanisms of Transdermal Drugtoxicology, Taylor & Francis, Washington, DC, 1996, pp.Delivery, Marcel Dekker, New York, 1997.3–17.

[21] J.E. Riviere, P.L. Williams, Pharmacokinetic implications of[3] N.A. Monteiro-Riviere, in: H.D. Dellmann, J. Eurell (Eds.),changing blood flow in skin, J. Pharm. Sci. 81 (1992)Textbook of Veterinary Histology, 5th Edition, Williams &601–602.Wilkins, Baltimore, MD, 1998, pp. 303–332, Chapter 16:

[22] M.P. Carver, P.E. Levi, J.E. Riviere, Parathion metabolismIntegument.during percutaneous absorption in perfused porcine skin,[4] P.M. Elias, Epidermal lipids, barrier function, and desquama-Pest. Biochem. Physiol. 38 (1990) 245–254.tion, J. Invest. Dermatol. 80 (1983) 44–49.

[23] S.K. Chang, P.L. Williams, W.C. Dauterman, J.E. Riviere,[5] S.K. Chang, J.E. Riviere, Effect of humidity and occlusionPercutaneous absorption, dermatopharmacokinetics, and re-on the percutaneous absorption of parathion in vitro, Pharm.lated biotransformation studies of carbaryl, lindane, malath-Res. 10 (1993) 152–155.ion and parathion in isolated perfused porcine skin, Toxicol-[6] J.E. Riviere, B.S. Sage, P.L. Williams, The effects ofogy 91 (1994) 269–280.vasoactive drugs on transdermal lidocaine iontophoresis, J.

[24] G.L. Qiao, P.L. Williams, J.E. Riviere, Percutaneous absorp-Pharm. Sci. 80 (1991) 615–620.tion, biotransformation and systemic disposition of parathion[7] H. Mukhtar, in: Pharmacology of the Skin, CRC Press, Bocain vivo in swine. I. Comprehensive pharmacokinetic model,Raton, FL, 1992.Drug Metab. Dispos. 22 (1994) 459–471.[8] R.J. Feldman, H.I. Maibach, Regional variations in percuta-

14 [25] S.K. Chang, C. Brownie, J.E. Riviere, Percutaneous absorp-neous penetration of C cortisol in man, J. Invest. Dermatol.tion of topical parathion through porcine skin: in vitro48 (1967) 181–183.studies on the effect of environmental perturbations, J. Vet.[9] T.O. Manning, N.A. Monteiro-Riviere, D.G. Bristol, J.E.Pharmacol. Ther. 17 (1994) 434–439.Riviere, Cutaneous laser Doppler velocimetry in nine animal

[26] G.E. Hardee, J.D. Baggot, in: Development and Formulationspecies, Am. J. Vet. Res. 52 (1991) 1960–1964.of Veterinary Dosage Forms, Marcel Dekker, New York,[10] N.A. Monteiro-Riviere, D.G. Bristol, T.O. Manning, J.E.1998.Riviere, Interspecies and interegional analysis of the com-

parative histological thickness and laser Doppler blood flow [27] M.N. Martinez, J.E. Riviere, Review of the 1993 Veterinary


Drug Bioequivalence Workshop, J. Vet. Pharmacol. Ther. 17 r-aminobenzoic acid in the isolated perfused porcine skinflap compared to man, Toxicol. Appl. Pharmacol. 151 (1998)(1994) 85–119.159–165.[28] J.E. Riviere, in: Comparative Pharmacokinetics: Principles,

[44] K.F. Bowman, N.A. Monteiro-Riviere, J.E. Riviere, De-Techniques and Applications, Iowa State University Press,velopment of surgical techniques for preparation of in vitroAmes, IA, 1999.isolated perfused porcine skin flaps for percutaneous absorp-[29] T.K. Ghosh, W.R. Pfister, S.I. Yum, in: Transdermal andtion studies, Am. J. Vet. Res. 52 (1991) 75–82.Topical Drug Delivery Systems, Interpharm Press, Buffalo

[45] M. Gibaldi, D. Perrier, in: Pharmacokinetics, 2nd Edition,Grove, IL, 1997.Marcel Dekker, New York, 1982.[30] B.W. Kemppainen, W.G. Reifenrath, in: Methods for Skin

[46] J. Gabrielsson, D. Weiner, in: Pharmacokinetic /Pharmaco-Absorption, CRC Press, Boca Raton, FL, 1990.dynamic Data Analysis: Concepts and Applications, 2nd

[31] M.S. Roberts, K.A. Walters, in: Dermal Absorption and RiskEdition, Apotekarsocieteten, Stockholm, 1997.

Assessment, Marcel Dekker, New York, 1998.[47] P.L. Williams, J.E. Riviere, A ‘full-space’ method for pre-

[32] A. Rougier, D. Dupuis, C. Lotte, R. Rouget, H. Schaefer, dicting in vivo transdermal plasma drug profiles reflectingCorrelation between stratum corneum reservoir function and both cutaneous and systemic variability, J. Pharm. Sci. 83percutaneous absorption, J. Invest. Dermatol. 81 (1983) 275– (1994) 1062–1064.278. [48] J.E. Riviere, J.D. Brooks, P.L. Williams, E. McGowan, M.L.

[33] A. Rougier, D. Dupuis, R. Rouget, C. Lotte, The measure- Francoeur, Cutaneous metabolism of isosorbide dinitratement of the stratum corneum reservoir. A predictive method after transdermal administration in isolated perfused porcinefor in vivo percutaneous absorption studies: influence of the skin, Int. J. Pharm. 127 (1996) 213–217.application time, J. Invest. Dermatol. 84 (1985) 66–68. [49] A.F. Kydonieus, B. Berner, in: Transdermal Delivery of

[34] United States Pharmacopeia 24 — The National Formulary Drugs, CRC Press, Boca Raton, FL, 1987.19 (1999), US Pharmacopeial Convention, Rockville, MD, [50] R.E. Baynes, A.L. Craigmill, J.E. Riviere, Residue avoidancepp. 1947–1951. after topical application of veterinary drugs and parasiticides,

[35] R. Bronaugh, H.I. Maibach, in: Percutaneous Absorption, 3rd J. Am. Vet. Med. Assoc. 210 (1997) 1288–1289.Edition, Marcel Dekker, New York, 1988. [51] J.E. Riviere, R.E. Baynes, Dermal absorption and toxicity

[36] N.A. Monteiro-Riviere, K.F. Bowman, V.J. Scheidt, J.E. assessment, in: M.S. Roberts, K. Walters (Eds.), DermalRiviere, The isolated perfused porcine skin flap (IPPSF): II. Absorption and Toxicity Assessment, Marcel Dekker, NewUltrastructural and histological characterization of epidermal York, 1998, pp. 625–645.viability, In Vitro Toxicol. 1 (1987) 241–252. [52] R.E. Baynes, J.E. Riviere, Inert ingredients in pesticide

[37] N.A. Monteiro-Riviere, A.O. Inman, T.H. Snider, J.A. Blank, formulations influence the dermal absorption of carbaryl,D.W. Hobson, Comparison of an in vitro skin model to Am. J. Vet. Res. 59 (1997) 159–169.normal human skin for dermatological research, Microsc. [53] K.A. Walters, M.S. Roberts, Veterinary applications of skinRes. Tech. 37 (1997) 172–179. penetration enhancers, in: K.A. Walters, J. Hadgraft (Eds.),

[38] J.E. Riviere, K.F. Bowman, N.A. Monteiro-Riviere, L.P. Dix, Pharmaceutical Skin Penetration Enhancement, MarcelM.P. Carver, The isolated perfused porcine skin flap (IPPSF). Dekker, New York, 1993.I. A novel in vitro model for percutaneous absorption and [54] J.D. Fikes, Chloropyrifos toxicosis in cats and cattle,cutaneous toxicology studies, Fundam. Appl. Toxicol. 7 NAPINet Report 1 (1988) 2.(1986) 444–453. [55] N. Gyrd-Hansen, L. Brimer, F. Rasmussen, Percutaneous

[39] J.E. Riviere, N.A. Monteiro-Riviere, The isolated perfused absorption of organophosphate insecticides in pigs — theporcine skin flap as an in vitro model for percutaneous influence of different vehicles, J. Vet. Pharmacol. Ther. 16absorption and cutaneous toxicology, Crit. Rev. Toxicol. 21 (1994) 174–180.

14(1991) 329–344. [56] A. Konar, G.W. Ivie, Fate of [ C]coumaphos after dermal[40] J.E. Riviere, N.A. Monteiro-Riviere, P.L. Williams, Isolated application to lactating goats as a pour-on formulation, Am.

perfused porcine skin flap as an in vitro model for predicting J. Vet. Res. 49 (1988) 488–492.transdermal pharmacokinetics, Eur. J. Biopharm. 41 (1995) [57] N.A. Monteiro-Riviere, A.O. Inman, J.E. Riviere, S.C.152–162. McNeill, M.L. Francoeur, Topical penetration of piroxicam

[41] P.L. Williams, M.P. Carver, J.E. Riviere, A physiologically is dependent on the distribution of the local cutaneousrelevant pharmacokinetic model of xenobiotic percutaneous vasculature, Pharm. Res. 10 (1993) 1326–1331.absorption utilizing the isolated perfused porcine skin flap [58] S.K. Chang, J.D. Brooks, N.A. Monteiro-Riviere, J.E.(IPPSF), J. Pharm. Sci. 79 (1990) 305–311. Riviere, Enhancing or blocking effect of fenvalerate on the

[42] J.E. Riviere, P.L. Williams, R. Hillman, L. Mishky, Quantita- subsequent percutaneous absorption of pesticides in vitro,tive prediction of transdermal iontophoretic delivery of Pest. Biochem. Physiol. 51 (1995) 214–219.arbutamine in humans using the in vitro isolated perfused [59] S.K. Chang, W.C. Dauterman, J.E. Riviere, Percutaneousporcine skin flap (IPPSF), J. Pharm. Sci. 81 (1992) 504–507. absorption of parathion and its metabolites paraoxon and

[43] R.C. Wester, J. Melendres, L. Sedik, H.I. Maibach, J.E. p-nitrophenol administered alone or in combination: in vitroRiviere, Percutaneous absorption of salicylic acid, theo- flow through diffusion cell system, Pest. Biochem. Physiol.phylline, 2,4-dimethylamine, diethyl hexylphthalic acid and 48 (1994) 56–62.


[60] G.L. Qiao, J.D. Brooks, R.L. Baynes, N.A. Monteiro-Riviere, Pluhar, K. Massa, D.E. Bjorling, A comparison of transder-P.L. Williams, The use of mechanistically defined chemical mal fentanyl versus epidural morphine for analgesia in dogsmixtures (MDCM) to assess component effects on the undergoing major orthopedic surgery, J. Am. Anim. Hosp.percutaneous absorption and cutaneous disposition of Assoc. 35 (1999) 95–100.topically-exposed chemicals. I. Studies with parathion mix- [69] C.M. Egger, T. Duke, J. Archer, P.H. Cribb, Comparison oftures in isolated perfused porcine skin, Toxicol. Appl. plasma fentanyl concentrations by using three transdermalPharmacol. 141 (1996) 473–486. fentanyl patch sizes in dogs, Vet. Surgery 27 (1998) 159–

[61] G.L. Qiao, S.K. Chang, J.E. Riviere, Effects of anatomical 166.site and occlusion on the percutaneous absorption and [70] J.R. Varvel, S.L. Shafer, S.S. Hwang, P.A. Coen, D.R.

14residue pattern of C-ring-2,6-parathion in vivo in pigs, Stanski, Absorption characteristics of transdermally adminis-Toxicol. Appl. Pharmacol. 122 (1993) 131–138. tered fentanyl, Anesthesiology 70 (1989) 928–934.

[62] G.L. Qiao, J.E. Riviere, Significant effects of application site [71] D.D. Lee, M.G. Papich, E.M. Hardie, Comparison of phar-and occlusion on the pharmacokinetics of cutaneous penetra- macokinetics of fentanyl after intravenous and transdermaltion and biotransformation of parathion in vivo in swine, J. administration in cats, Am. J. Vet. Res. 61 (2000) 672–677.Pharm. Sci. 84 (1995) 425–432. [72] G.L. Carroll, R.N. Hooper, D.M. Boothe, S.M. Hartsfield,

[63] P.L. Williams, D. Thompson, G.L. Qiao, N.A. Monteiro- L.A. Randoll, Pharmacokinetics of fentanyl after intravenousRiviere, R.L. Baynes, J.E. Riviere, The use of mechanistical- and transdermal administration in goats, Am. J. Vet. Res. 60ly defined chemical mixtures (MDCM) to assess component (1999) 986–991.effects on the percutaneous absorption and cutaneous dispo- [73] J.E. Riviere, M. Heit, Electrically-assisted transdermal drugsition of topically-exposed chemicals. II. Development of a delivery, Pharm. Res. 14 (1997) 691–701.general dermatopharmacokinetic model for use in risk as- [74] B.H. Sage, J.E. Riviere, Model systems in iontophoresis —sessment, Toxicol. Appl. Pharmacol. 141 (1996) 487–496. transport efficacy, Adv. Drug Deliv. Rev. 9 (1992) 265–287.

[64] M. Dagorn, P. Guillot, P. Sanders, M. Laurentie, P.L. [75] M. Heit, P.L. Williams, F.L. Jayes, S.K. Chang, J.E. Riviere,Toutain, Deconvolution study of the absorption rate and Transdermal iontophoretic peptide delivery. In vitro and indisposition kinetic values of lindane in sheep, Am. J. Vet. vivo studies with luteinizing hormone releasing hormoneRes. 51 (1990) 1993–2000. (LHRH), J. Pharm. Sci. 82 (1993) 240–243.

[65] M. Dagorn, M. Poul, P. Guillot, P. Sanders, Residues of [76] J.E. Riviere, B.S. Sage, P.L. Williams, The effects oflindane in plasma and fat of sheep after dipping, Food Addit. vasoactive drugs on transdermal lidocaine iontophoresis, J.Contam. 5 (1987) 51–58. Pharm. Sci. 80 (1991) 615–620.

[66] A.E. Kyles, M. Papich, E. Hardie, Disposition of transder- [77] J.E. Riviere, N.A. Monteiro-Riviere, R.A. Rogers, D. Bom-mally administered fentanyl in dogs, Am. J. Vet. Res. 57 mannan, J.A. Tamada, R.O. Potts, Pulsatile transdermal(1996) 715–719. delivery of LHRH using electroporation. Drug delivery and

[67] A.E. Kyles, E.M. Hardie, B.D. Hansen, M.G. Papich, skin toxicology, J. Control. Release 36 (1995) 229–233.Comparison of transdermal fentanyl and intramuscular oxy- [78] J.E. Riviere, T. Martin, S. Sundlof, A.L. Craigmill, Inter-morphone on post-operative behaviour after ovariohysterec- species allometric analysis of the comparative phar-tomy in dogs, Res. Vet. Sci. 65 (1998) 245–251. macokinetics of 44 drugs across veterinary and laboratory

[68] T.M. Robinson, K.T. Kruse-Elliot, M.D. Markel, G.E. species, J. Vet. Pharmacol. Ther. 20 (1997) 453–463.

Potential and Problems of Developing

Documents

Transcript of Potential and Problems of Developing