transdermal delivery of the potent analgesic dihydroetorphine: kinetic analysis of skin permeation...

J. Pharm. Pharmacol. 2000, 52: 1437±1449 # 2000 J. Pharm. Pharmacol.Received March 27, 2000Accepted July 19, 2000

Transdermal Delivery of the Potent Analgesic Dihydroetorphine:Kinetic Analysis of Skin Permeation and Analgesic Effect in the

Hairless Rat

SATOSHI OHMORI, TERUAKI HAYASHI, MASAMI KAWASE, SETSUO SAITO, KENJI SUGIBAYASHIAND YASUNORI MORIMOTO

Faculty of Pharmaceutical Sciences, Josai University, Sakado, Saitama 350-0295, Japan

Abstract

Dihydroetorphine is an extraordinarily strong opioid analgesic. To assess its effectivenessafter topical application in hairless rats we have examined the kinetic analysis of skinpermeation through excised skin and the in-vitro reservoir effect of skin, and haveinvestigated the predictability of plasma concentration and analgesic effect following in-vivo transdermal application.

Dihydroetorphine was moderately permeable from an aqueous suspension throughexcised hairless rat skin. Dihydroetorphine ¯ux from drug-dispersed pressure-sensitiveadhesive tape was threefold that from the applied aqueous suspension. The ¯uxes throughthe abdominal and the dorsal skin during tape application ®tted the Fickian diffusionequation well after the tape was removed peeling off the outer layer of the stratum cor-neum. The relationship between the plasma concentration and the analgesic effect wasexamined for four different rates of infusion of dihydroetorphine. A non-linear pharma-cokinetic disposition was observed. Following abdominal (0�28 cm2, 20mg) and dorsal(0�50 cm2, 35 mg) applications of the dihydroetorphine tape, plasma concentration (0�2±0�8 ng mLÿ1) and analgesic effect were maintained at a suitable level, for more than 8 h,until removal of the tape. These pro®les were predictable using the combined equation forpercutaneous absorption, disposition and the analgesic effect, but the analgesic effect wasslightly lower than the predicted value.

The results show that it was possible to control the plasma concentration and theanalgesic effect of dihydroetorphine by topical application of the analgesic using pressure-sensitive adhesive tape in the hairless rat. It was possible to predict the result usingmathematical modelling.

There are many reports on the usefulness of thetopical application of opioid analgesics such asmorphine (Sugibayashi et al 1989), buprenorphine(Stinchcomb et al 1996) and fentanyl (Southam1995). Continuous pain relief is necessary for thetreatment of illness and for improvement in thequality-of-life for end-stage cancer patients.Transdermal drug delivery is the easiest way forcontinuous delivery if the drug is adequatelypermeable through the skin. However, the outerlayer of the skin, the stratum corneum, is a sig-ni®cant barrier for the introduction of many drugs(Franz et al 1992). The aforementioned opioid

analgesics are reported to be permeable through ratand human skin with the aid of permeationenhancers to produce suf®cient pharmacologicaleffects. However, severe enhancers often causeirritation to the skin (Franz et al 1992; Finnin &Morgan 1999). If a more potent opioid analgesicwas used for this therapeutic system, it may bepossible to achieve continuous relief of severe painwithout the aid of enhancers.

The novel opioid analgesic dihydroetorphine(Figure 1) has an analgesic effect more than 1000-fold stronger than morphine (Bentley & Hardy1967; Tokuyama et al 1993). Dihydro-etorphinebegan to be used clinically for pain relief in Chinain the 1980s (Wang et al 1999; Zang 1999), but itsabuse increased soon after it was marketed becauseof its dependent liability under unrestricted control

Correspondence: Y. Morimoto, Faculty of PharmaceuticalSciences, Josai University, 1-1 Keyakidai, Sakado, Saitama350-0295, Japan.

(WHO Expert Committee 1999). In 1999, theUnited Nations registered dihydroetorphine as oneof the most strictly controlled narcotic drugs.

Dihydroetorphine can be administered intra-venously or sublingually. A sublingual dose producesa quick onset of analgesic effect (within 20 min),but its duration is relatively short (within 1�5±5�3 h)after dosing (Wu & Sun 1991). Therefore, it has tobe administered several times per day for con-tinuous pain relief.

In a previous study (Ohmori et al 2000b), it wascon®rmed that dihydroetorphine was quickly dis-tributed to the central nervous system (< 2 min)and was eliminated smoothly (half-life 37�7 min)after an intravenous injection in hairless rat.Furthermore, after intracutaneous administra-tion dihydroetorphine was highly available(bioavailability> 0�7) and produced a quick-onsetof analgesic effect (< 5 min), unlike the avail-ability after oral administration which was low(bioavailability< 0�004) due to extensive ®rst-passmetabolism. Its rapid ef®cacy and eliminationpotential make it feasible for input-phase controlleddrug delivery. In addition, its potency and somephysiological properties (melting point and lipo-philicity) are suitable for transdermal delivery(Finnin & Morgan 1999).

Guo et al (1997) reported in-vitro snake skinpermeation of dihydroetorphine using the permea-tion enhancer Azone. Chen et al (1996) reportedthat the transdermal delivery of dihydroetorphineusing a drug-reservoir system with Azone showed a

stable blood concentration and analgesic effect for32 h in Wistar rats. However, Azone has a sig-ni®cant permeation enhancing effect and leads toskin irritation (Okamoto et al 1988). Etorphine,structurally similar to dihydro-etorphine but with a potency one-quarter ofdihydroetorphine, was permeable through mice andcadaver skin without permeation enhancers, but toaccomplish the daily dose an application area of50 cm2 was required (Jolicoeur et al 1992). Ifdihydroetorphine has an equivalent permeability toetorphine, its application area would be smallerwhich would enable practical use.

In this study we have kinetically analysed theskin permeation and reservoir effect of dihydro-etorphine in excised abdominal and dorsal skinof hairless rat in-vitro. We have compared ¯uxesthrough the skin from aqueous suspension anddrug-dispersed pressure-sensitive adhesive tape in-vitro. We have estimated the effective plasmaconcentration range of dihydroetorphine usingvarious rates of infusion in hairless rat in-vivo. Wehave evaluated also the plasma concentration andanalgesic effect following topical application ofdihydroetorphine tape, and their predictability bymathematical calculation in-vivo.

Materials and Methods

AnimalsMale hairless rats (WBN=ILA-Ht strain) (250±350 g; 12±16-weeks old) were supplied by LifeScience Research Center of Josai University (Sai-tama, Japan) and Ishikawa Experimental AnimalLaboratory (Saitama, Japan). The animals weremaintained at 24� 1�C under a 12-h light±darkcycle and had free access to a standard rodent dietand clean drinking water. The experiments wereperformed in accordance with the Guide forLaboratory Animal Experiment adopted by JosaiUniversity.

ChemicalsFree base dihydroetorphine was synthesized fromcodeine by reported procedures (Bentley & Hardy1967; Barber & Rapoport 1975). Styrene±isoprene-styrene block co-polymer (Cari¯ex TR-1107) waspurchased from Shell Chemical Co. (Tokyo,Japan). Rosin ester (KE-311) was purchased fromArakawa Chemical Co. (Osaka, Japan). Isopropylmyristate and p-hydroxybenzoic acid methyl esterwere purchased from Tokyo Chemical Industry Co.(Tokyo, Japan). Trypsin was purchased from Wako

Figure 1. Chemical structure of dihydroetorphine.

1438 SATOSHI OHMORI ET AL

Pure Chemical Co. (Kyoto, Japan). Buprenorphinehydrochloride was kindly supplied by OtsukaPharmaceuticals (Tokyo, Japan). All other reagentswere of analytical grade.

Preparation of dihydroetorphine aqueoussuspensionExcess amounts of dihydroetorphine and isopropylmyristate were added to distilled water and weredissolved suf®ciently by sonication. The mixturewas left overnight. The undissolved drug and iso-propyl myristate were separated just before use.

Preparation of pressure-sensitive adhesive tapeThe matrix-type pressure-sensitive adhesive tapewas prepared by the casting method (Morimoto etal 1992a). Dihydroetorphine, styrene±isoprene-styrene block co-polymer, rosin ester and isopropylmyristate were dissolved in chloroform, and thenthe mixture was gently stirred at ambient tem-perature. The mixture was cast onto polyethyleneterephthalate ®lm as the backing layer usingBaker's Film Applicator (Ueshima Seisakusho,Tokyo, Japan). The adhesive tape was kept at 60�Cfor 15 min and then at ambient temperature for 16 hto evaporate the chloroform. The tape was thencovered with siliconized polyethylene terephthalate®lm as a release liner. The tape was stored atambient temperature until use. Immediately beforeapplication, the tape was pulled out circularly to anappropriate area. The thickness of the adhesivelayer of the tape was 70mm. Dihydroetorphinecontent of the tape was 66mg cmÿ2, and at least80% of the drug in the tape was dispersed whenobserved under a light microscope. A placebo tapewas prepared by the same method excluding thedihydroetorphine.

Skin preparationThe rat abdominal and dorsal regions were care-fully shaved, and skin excised under diethyletheranaesthesia. Subcutaneous tissue was carefullyremoved from the skin (intact skin). Stripped skinwas obtained by stripping the stratum corneumfrom the skin with adhesive cellophane tape(Nichiban Co., Tokyo, Japan) 20 times. The dorsalsplit skin was prepared by cutting off the dermisside of the dorsal skin using a cryostat microtome.The thickness of each skin preparation was mea-sured with a slide calliper. Other pieces ofabdominal and dorsal skin (16 cm2) were immersedin 1% trypsin at 37�C for 24 h, and then the stratumcorneum was carefully isolated. After suf®cient

rinsing and desiccation, the isolated stratum cor-neum of both skins was weighed.

In-vitro skin permeation studiesIn the aqueous suspension application study, theexcised skin preparations were inserted betweentwo halves of side-by-side diffusion cells encom-passed by a 37�C water jacket with a volume of2�5 mL and 0�95 cm2 effective diffusion area. As apretreatment of vehicles, isopropyl myristate satu-rated solution and distilled water were ®lled intothe donor and receiver compartments of the diffu-sion cell which were in contact with the stratumcorneum side and dermis side of the skin, respec-tively. Applied vehicles in each compartment werestirred with a magnetic bar driven by a constant-speed synchronous motor (Scinics, Tokyo, Japan).After the overnight-pretreatment, donor and recei-ver solutions were replaced by dihydroetorphineaqueous suspension and fresh distilled water,respectively, and then sampling of the receiversolution was started as described below. In the tapeapplication study, the dermis side of the skin wasattached to one-half of the side-by-side diffusioncells. The dihydroetorphine tape was applied to thestratum corneum side of the skin, and then distilledwater was ®lled into the receiver compartment ofthe diffusion cell. In both application studies, thereceiver solution was withdrawn at appropriatetimes for measurement of the permeated drug, anddistilled water was added to maintain a constantvolume in the diffusion cell. The suspension andtape were removed from the skin at appropriatetimes, and the receiver solution was sampled forevaluation of the reservoir capacity in the skinpreparation.

Intravenous infusion of dihydroetorphineRats were anaesthetized with diethylether. Afemoral vein was cannulated with polyethylenetubing for infusion of dihydroetorphine and afemoral artery was cannulated for blood sampling.After surgery each rat was placed in a Bollmancage and allowed to recover from the anaestheticfor 2 h. After the basal analgesic effect was mea-sured, dihydroetorphine in saline was infused at0�9, 1�8, 3�6 and 9�0 mg hÿ1 kgÿ1 for 4 h through thevenous cannula (0�5 mL minÿ1). Analgesic effectswere measured at appropriate times during andafter the infusion, and in succession with eacheffect measurement, blood was withdrawn from thearterial cannula. To determine the lowest con-centration of dihydroetorphine a 1-mL blood sam-ple was required, whereas a blood sample less than

TRANSDERMAL APPLICATION OF DIHYDROETORPHINE 1439

1 mL was enough to determine the higher con-centration (Ohmori et al 2000a). Therefore, theblood sampling was set at ®ve to nine times peranimal dependent on the predicted concentration inthe samples, with the total volume of blood samplesnot exceeding 1% of the body weight. Bloodsamples were placed in heparinized tubes and theplasma was separated by centrifugation. The brainwas excised after decapitation, and then the cere-bellum was removed from the brain. The brain washomogenized with 2 vols methanol. Supernatant ofthe homogenate was separated by centrifugation(9000 g) at 4�C. Plasma and supernatant sampleswere stored at ÿ20�C until analysis.

In-vivo application of dihydroetorphine tape onskinRats were anaesthetized with diethylether and afemoral artery was cannulated with polyethylenetubing for blood sampling. The tip of the cannulawas drawn through the skin on the back of the neckso that the awakening rats were able to move freely.After surgery the animals were allowed to recoverfrom the anaesthesia for 2 h. After the basalanalgesic effect was measured, dihydroetorphinetape was applied onto the abdominal skin(0�28 cm2, 20 mg drug) for 8 h or onto the dorsalskin (0�50 cm2, 35mg drug) for 24 h. Analgesiceffects were measured at appropriate times duringand after the application of the tape, and in suc-cession with each effect measurement, blood waswithdrawn from the arterial cannula. Blood sam-pling was set at four to six times per animal, withthe total volume of blood samples not exceeding1% of the body weight. Plasma was separated bythe same manner as in the infusion study and wasstored at ÿ20�C until analysis.

Measurement of dihydroetorphine concentrationsin the receiver solutionA sample of the receiver solution (100mL) wasmixed with an equal volume of p-hydroxybenzoicacid methyl ester as an internal standard (10mg mLin acetonitrile). After centrifugation (9000 g), thesupernatant (20 mL) was applied to a high-perfor-mance liquid chromatography system (LC-10A,Simadzu Co., Kyoto, Japan). The phosphate buffer(1=30 M, pH 5�2)±acetonitrile±methanol (70 : 21 : 9)mobile phase was delivered at 1�0 mL minÿ1

through Nucleosil 5C18 (4�6 mm i.d.6 250 mm,Macherey Nagel, Germany) under 40�C. Dihy-droetorphine and internal standard were detected at215 nm.

Measurement of dihydroetorphine concentrationsin rat plasma and brainDihydroetorphine concentrations in the plasma andbrain supernatants were measured by liquidchromatography±tandem mass spectrometry(Ohmori et al 2000a). Brie¯y, an exact volume lessthan 0�5 mL of the sample was mixed with 0�1 mLbuprenorphine methanol solution as an internalstandard (10 ng mLÿ1 in methanol) and 50 mM

phosphate buffer (pH 6�0). These mixtures wereapplied to a Bond Elut Certify cartridge column(3 mL=130 mg, Varian, Harbor City, CA) that hadbeen conditioned with methanol and 50 mM phos-phate buffer (pH 6�0). The column was washedsequentially with 100 mM acetic acid and methanol.The eluent with 2% ammonium hydroxide in ethylacetate was collected and evaporated. The residuewas dissolved in acetonitrile±water (80 : 20) and asample was applied to a LC-MS-MS system (API-300, Perkin Elmer-SCIEX, Foster City, CA). Anacetonitrile±50 mM ammonium acetate (95 : 5)mobile phase was delivered at 0�3 mL minÿ1

through Inertsil ODS-2 (5mm, 2�1 mmi.d.6 150 mm, GL Science, Tokyo, Japan) under40�C. The analytical conditions of API-300 werethe same as reported by Ohmori et al (2000a). Thelimit of quantitation with acceptance criteria asintra- and inter-assay precision within 20% was0�05 ng mLÿ1 in plasma and 0�15 ng gÿ1 in brain.

Measurement of analgesic effectThe analgesic effect was determined using the tail-immersion test (Ouellet & Pollack 1997). In thedihydroetorphine infusion study the tail of the ratwas extended from the Bollman cage. In thetransdermal application study using the dihy-droetorphine tape, rats were loosely wrapped in acloth from which the tail was extended. Two-thirdsof the tail was immersed in hot water (50±55�C)and the latent time for ¯ick of the tail or strugglewas measured. Basal latency was approximately 2 sand did not exceed 3 s. A cut-off time of 10 s wasadopted to prevent tail damage. Analgesic effectwas expressed as % of the maximum possible effect(% MPE): % MPE� (post-drug latencyÿ pre-druglatency)=(cut-off latencyÿ pre-drug latency)6 100.

Calculation of kinetic parametersFigure 2 shows the scheme of the mathematicalmodel for skin permeation, disposition and theanalgesic effect of dihydroetorphine.

Parameters for skin permeation. It was postulatedthat dihydroetorphine permeation through the


stratum corneum (sc) and viable skin (vs) occurredin accordance with the mono- or bi-layer diffusionmodel based on Fick's Second Law without anyspeci®c binding and=or enzymatic degradation ofthe drug in each skin layer (Tojo 1987). Dihydroe-torphine concentrations in stratum corneum (CSC)and viable skin (CVS) at a position x and at time t,are represented as follows:

aCSC=at � DSCa2CSC=ax2 �1�

aCVS=at � DVSa2CVS=ax2 �2�where DSC and DVS are diffusivity in each layer ofskin. Initial and boundary conditions in the skinwere set as follows:

at t � 0 and 0 � x � LSC

� LVS;CSC � CVS � 0�3�

at t > 0 and x � 0;CSC � KSC=DCD �4�

at t > 0 and x � LSC;CVS � KVS=SCCSC and

DSCdCSC=dx � DVS dCVS=dx�5�

at t > 0 and x � LSC � LVS;CVS � 0 �6�where LSC and LVS are the thickness of each layer ofskin, CD is the effective drug concentration in thedonor compartment (D) as a solubility of the drug inthe suspension and in the tape, and KSC=D and KVS=SC

are the partition coef®cients of dihydroetorphinebetween layers. The ¯ux of dihydroetorphine throughthe skin into the receiver compartment in the diffu-sion cell, j, is represented as follows:

j � ÿDVS�dCVS=dx�x�Lsc�Lvs �7�Partial differential equations 1 and 2, and ordinarydifferential equations 5 and 7, are approximated bythe explicit ®nite-difference equation (Crank 1975).The number of spatial mesh in each layer was 10,which satis®ed the degree of accuracy in this study.

The ¯uxes before removing the donor were ®ttedto equation 7 by the weighted least-square method

Figure 2. Scheme of the mathematical model for skin permeation, disposition and production of the analgesic effect. CD, CSC,CVS, C1, C2, dihydroetorphine concentrations in each layer or compartment; KSC=D, KVS=SC, partition coef®cient of dihydroetor-phine between layers; DSC, DVS, diffusivity of dihydroetorphine in each layer; LSC, LVS, thickness of each layer; Rinf, infusion rateof dihydroetorphine; Vd1, Vd2, volume of distribution in each compartment; k12, k21, k10, ®rst-order rate constant for transferbetween compartments and elimination of dihydroetorphine; EC50, dihydroetorphine concentration producing 50% of themaximum effect; r, sigmoidicity factor. In the topical application study, the drug in the donor enters the receiver compartmentof the diffusion-cell in-vitro and enters the central compartment of the body in-vivo through the skin. The dotted line in the stratumcorneum indicates the outer layer that peels off with the removal of the donor. The dotted line in viable skin indicates the drug thatenters into the blood vessel (central compartment) in-vivo. In the intravenous infusion study, dihydroetorphine directly entered thecentral compartment in the body in-vivo.


using a quasi-Newton algorithm on solver-functionof Microsoft Excel 97. Initially, the ¯ux through thestripped skin was ®tted to equation 7 combinedwith the mono-layer diffusion model, a more sim-ple case than the equations described above, byusing ®xed LVS, and thereby KVS=DCD and DVS

were estimated. Secondly, the ¯ux through theintact skin was ®tted to equation 7 combined withthe bi-layer diffusion model by using a ®xed valuefor LVS, KVS=DCD, DVS and LSC, and substitutingKVS=SC by KVS=DCD=KSC=DCD, and therebyKSC=DCD and DSC were estimated. The ¯ux atsteady state and the lag time for the quasi-steadystate were calculated from theoretical values byFick's First Law (Franz et al 1992).

Pharmacokinetic parameters. Pharmacokineticparameters of dihydroetorphine in the infusionstudy were calculated by ®tting a linear two-compart-ment model with a zero-order-rate input function,equation 8, to the mean plasma concentration, Cp.

Cp � Rinf

Vd1�aÿ b� �k21 ÿ a

afexp�ÿat1� ÿ 1g exp�ÿat2��

�bÿ k21

bfexp�ÿbt1� ÿ 1g exp�ÿbt2�� 8�

where Rinf and Vd1 are the infusion rate of the drugand the distribution volume of the central com-partment, respectively. a and b are the slopes ofdistribution and elimination phase on the logarith-mic plasma concentration±time curve, respec-tively, and k21 is the ®rst-order rate constant fortransfer from the peripheral- to the central-com-partment. t1 and t2 are the time during infusion andafter infusion, respectively.

Pharmacodynamic parameters. Pharmacodynamicparameters were calculated by ®tting a sigmoidaleffect±concentration relationship, equation 9, tothe mean plasma concentration and the mean %MPE in each infusion group.

% MPE � 100� �Cp�r=��EC50�r � �Cp�r� �9�where the maximum effect (MPE) is 100, EC50 isthe dihydroetorphine concentration producing 50%of MPE and r is the sigmoidicity factor describingthe shape of the relation curve.

Simulation of percutaneous absorption, plasmaconcentration and analgesic effectThe release pro®les after the donor was removedin-vitro and the time course of plasma concentra-

tion and analgesic effect during and after the topi-cal application of the tape in-vivo were simulatedby numerical solution combined with skin per-meation, disposition and the analgesic effects. Afterthe donor was removed from the intact or strippedskin, the boundary condition as in equation 4 wasexpressed as follows (Tojo 1988):

at t > 0 and x � 0; dCSC=dx � 0 or dCVS=

dx � 0; respectively �10�On the assumption that the outer layer of the stra-tum corneum is peeled off by removing the donor,several mesh on ®nite-difference approximationwere deleted from the outer layer of the stratumcorneum, so that equation 10 was held at theappropriate position x, as an integral number, notzero. Half-life of the ¯ux after removing the donorwas calculated from the theoretical ¯ux by the log-linear regression.

Results

In-vitro skin permeation of dihydroetorphine froman aqueous suspensionFigure 3 shows the ¯ux and the cumulative amountof dihydroetorphine that permeated through theexcised abdominal skin from the aqueous suspen-sion. The ¯ux gradually increased after the appli-cation of the suspension and reached a steady-statelevel within a few hours. The steady-state ¯uxthrough the stripped skin was 3�6-fold higher thanthat of the intact skin. These ¯uxes and cumulativeamounts during the suspension application ®ttedwell the mono- or bi-layer diffusion model forstripped and intact skin, respectively (solid lines inFigure 3). The obtained kinetic parameters are lis-ted in Table 1. Initial concentrations at theboundary between the donor solution and strippedskin or intact skin, KSC=DCD or KVS=DCD, wereapproximately the same. However, the diffusivity,DVS, was 110-fold higher than DSC. After the sus-pension was removed at 8 h (vertical dotted lines inFigure 3), the ¯ux decreased exponentially whichwas well predicted by the diffusion model withequation 10 at position x� 0 (solid lines in Figure3). Half-lives of the ¯ux from the intact and strip-ped skin after removing the suspension wereapproximately the same (Table 1).

In-vitro skin permeation of dihydroetorphine fromthe tapeFigure 4 shows the ¯ux and cumulative amount ofdihydroetorphine that permeated through the


excised abdominal and dorsal skin preparationsfrom the tape formulation. The steady-state ¯uxthrough the intact dorsal skin was one-half of thatthrough the intact abdominal skin. The ¯ux was3�1-fold higher than that after the application of thesuspension. The kinetic parameters calculated fromthe diffusion models are listed in Table 1. Dryweight of the stratum corneum of the dorsal skinwas 2�8-fold higher than that of the abdominal skin,so that the thickness of these skins was estimated as15�9 mm for abdominal skin and 45�8 mm for dorsalskin. The KSC=DCD for abdominal and dorsal skinwas 11±47-fold higher than that of the suspensionapplied, although KVS=DCD was the same order.Although the difference between DVS values forabdominal and dorsal skin was less than 2, DSC fordorsal skin was sixfold that of abdominal skin. In

the stripped skin, the release pro®les and cumula-tive amount permeated following the removal ofthe tape were well predicted using the diffusionmodel with equation 10 at position x� 0 (solidlines in Figure 4). In intact skin the release pro®leswere preferably described by the model withequation 10 at position x� 0�6 LSC for abdominalskin and x� 0�7 LSC for dorsal skin (dashed lines inFigure 4). This postulated whether the outer six orseven layers of the stratum corneum were peeledoff when removing the tape. It was visible to peeloff of several outer layers of the skin whenremoving the tape. Half-lives of the theoretical ¯uxcalculated with peeling off the stratum corneumsigni®cantly decreased from that calculated withoutpeeling off the stratum corneum in the abdominaland dorsal skin. To con®rm the suitability of thesemodels, the simulation of the ¯ux and cumulativeamount through the dorsal split skin by usingkinetic parameters (Table 1) was determined. It waspossible to predict them satisfactorily from thediffusion model supposing that the peeling off ofthe stratum corneum by removing the tape occurredat position x� 0�4 LSC on the dorsal split skin(dashed line in panel C and D of Figure 4). Thedifference between the peeled position on the dor-sal intact skin and split skin might be caused by theapplication period of the tape in-vitro.

Plasma concentration and analgesic effect byinfusion of dihydroetorphineFigure 5 shows the plasma and brain concentrationsand analgesic effect during and after 4-h-con-tinuous infusion of dihydroetorphine at rates of 0�9,1�8, 3�6 and 9�0mg hÿ1 kgÿ1. Pharmacokinetic andpharmacodynamic parameters are listed in Table 2.Steady-state plasma concentrations increased withthe infusion rate from 0�9 to 3�6mg hÿ1 kgÿ1. At the9�0 mg hÿ1 kgÿ1 infusion, however, the plasmaconcentration promptly rose. As a result of thepharmacokinetic analysis, the volume of distribu-tion of the central compartment apparentlydecreased at the 9�0mg hÿ1 kgÿ1 infusion. The ratswere severely anaesthetized and their respirationwas extremely weak, suggesting that an abnormaldisposition of dihydroetorphine occurred. Dihy-droetorphine indicated the linear disposition below1�2 ng mLÿ1 of plasma concentration. However, thedihydroetorphine concentration in the brain wasapproximately threefold that in the plasma even at9�0 mg hÿ1 kgÿ1 (Figure 5, open symbols, con-centration data). The analgesic effects increasedwith the infusion rate, and reached 100% maximumeffect at 3�6mg hÿ1 kgÿ1 infusion. The mean EC50

Figure 3. In-vitro permeation of dihydroetorphine from sus-pension through the excised abdominal skin of hairless rats.Each point represents the mean and standard error (n� 4). d,Intact skin; m, stripped skin. Vertical dotted lines indicate thetime of removal of the donor solution. Solid lines indicate thecalculated mean values obtained by ®tting to the mono- or bi-layer diffusion model.


and r, calculated using equation 9, were0�708 ng mLÿ1 and 3�03, respectively.

Prediction of the plasma concentration andanalgesic effect following topical application ofdihydroetorphine tapeAfter abdominal (0�28 cm2, 20mg) and dorsal(0�50 cm2, 35mg) tape application, the plasmaconcentration and analgesic effect increased to theeffective concentration until tape removal (Figure6). The plasma concentration during the applicationof the tape was similar to the predicted value cal-

culated using the combined equation. However, theanalgesic effects were slightly lower than the pre-dicted values.

Discussion

Dihydroetorphine shows several suitable propertiesfor transdermal delivery, as summarized by Finnin& Morgan (1999): `̀ the daily systemic dose shouldbe �20 mg'', the sublingual dose of dihy-droetorphine is �160mg (Li et al 1996); `̀ themolecular weight should be < 500 Da'', dihy-

Table 1. Kinetic parameters for the skin permeation of dihydroetorphine through the excised skin of hairless rats.

Skinpreparation

Thickness(mm)

Initial concn(mg cmÿ3)

Diffusivity(cm2 hÿ1)

LSC LVSa

(in-vitro)LVS

(in-vivo)KSC=DCD

b KVS=DCDc DSC

b DVSc

Abdominal skin (suspension)Stripped Ð 550 Ð Ð 0�350 Ð 3�976 10ÿ4

Intact 15�4e 550 Ð 0�386 (0�350) 4�426 10ÿ6 (3�976 10ÿ4)

Abdominal skin (tape)Stripped Ð 443 Ð Ð 0�532 Ð 8�236 10ÿ4

Intact 15�4e 578 154g 18�3 (0�532) 2�816 10ÿ7 (8�236 10ÿ4)

Dorsal skin (tape)Stripped Ð 1205 Ð Ð 0�348 Ð 1�476 10ÿ3

Intact 43�4f 1437 434g 4�24 (0�348) 1�696 10ÿ6 (1�476 10ÿ3)Splitd 43�4f 499 Ð (4�24) (0�348) (1�696 10ÿ6) (1�476 10ÿ3)

Skinpreparation

Steady state¯uxh (mg hÿ1cmÿ2)

Lag time forquasi-steady

stateh (h)

Half-life of ¯ux after removal of the donorh (h)

Without peeling ofstratum corneum

With peeling ofstratum corneum

Abdominal skin (suspension)Stripped 2�52 1�23 2�36 ÐIntact 0�70 2�40 2�10 Ð

Abdominal skin (tape)Stripped10ÿ4 9�81 0�38 0�74 ÐIntact 2�14 3�06 5�47 2�14

Dorsal skin (tape)Stripped 4�16 1�52 3�05 ÐIntact 1�07 6�97 8�90 4�52Splitd 1�42 2�96 Ð 2�68

aMeasured by slide calliper. bParameters for the stratum corneum were obtained by ®tting the bi-layer diffusion model to themean values of the dihydroetorphine ¯ux through the intact skin using ®xed parameters for viable skin given in parenthesis.cParameters for viable skin were obtained by ®tting mono-layer diffusion model to the mean dihydroetorphine ¯ux through thestripped skin. dParameters obtained from the intact skin data were used for simulation. eSato et al (1991). fCalculated from thethickness of abdominal stratum corneum corrected by the dry-weight ratio of isolated dorsal stratum corneum(2�79 mg cmÿ2)=abdominal stratum corneum (0�99 mg cmÿ2). gCalculated from 10 times the thickness of stratum corneum.hParameters were calculated from theoretical values illustrated in Figures 3 and 4.


droetorphine is 413 Da; `̀ the log P (octanol=water)should be in the range 1±3'', log P is 2�13 for dihy-droetorphine (unpublished data); `̀ the melting pointshould be < 200�C'', the melting point is 205�C fordihydroetorphine (Bentley & Hardy 1967); `̀ directskin irritation should not be presented'', only a slightirritation is observed for dihydroetorphine with thepermeation enhancer Azone (Chen et al 1996). Finnin& Morgan mentioned another two factors that arerequired: `̀ hydrogen-bonding groups should be�2'',dihydroetorphine has ®ve groups, and `̀ immuno-genicity should not be presented'', this has not yetbeen proved.

In this study, it was con®rmed that dihy-droetorphine was permeable enough through hair-less rat skin to produce an analgesic effect. Usingdrug-dispersed pressure-sensitive adhesive tape, the¯ux of dihydroetorphine through the intactabdominal skin was enhanced threefold with a 47-

fold increase of the initial concentration in stratumcorneum and a 16-fold decrease of DSC in com-parison with application of suspension with satu-rated isopropyl myristate (Table 1). These resultsare different from our previous report (Morimoto etal 1992a) that the ¯ux of isosorbide dinitrate wasthe same between isopropyl myristate saturatedaqueous suspension and pressure-sensitive adhe-sive. It was suggested that hydration of the stratumcorneum caused a decrease of solubility and anincrease of diffusion resistance of dihydroetorphineowing to its hydrophobicity (log P 2�13) dissimilarto the more hydrophilic isosorbide dinitrate (log P1�34) (Morimoto et al 1992b). Isopropyl myristatehas been used widely as a vehicle in cosmetic andpharmaceutical formulations because of its emol-lient properties and safety (Suh & Jun 1996). In thisstudy, it was used as an emollient additive ratherthan a permeation enhancer because its enhancing

Figure 4. In-vitro permeation of dihydroetorphine from the tape formulation through the excised abdominal (A, B) and dorsal (C,D) skin of hairless rats. Each point represents the mean and standard error (n� 4). d, Intact skin; m, stripped skin; j, split skin.Vertical dotted lines indicate the time of removal of the tape. Solid lines indicate the calculated mean values obtained by ®tting tothe mono- or bi-layer diffusion model. Dashed lines indicate the calculated mean values predicted by the diffusion model afterpeeling off the outer layer of the stratum corneum when removing the tape.


ratio of dihydroetorphine permeation was less than2 (Ohmori et al 2000c).

From microscopic observations of the variouscontents of the dihydroetorphine tape, at least 80%of the drug was dispersed in the tape. The ¯ux ofdihydroetorphine was stable for 8 and 24 h until thetape was removed from the abdominal and thedorsal intact skin, in which cumulative amountstransferred into the skin were nearly 30mg cmÿ2,more than 40% of the contents in the tape (Figure4B and D). Therefore, it is postulated that theboundary concentrations on the skin surface wereregular during these periods. If the tape was appliedfor a longer time, the ¯uxes might continue untilthe dispersed drug was lost.

Steady-state ¯ux of dihydroetorphine through thedorsal skin was one-half of that through theabdominal skin (Table 1). We suggested that thisdiscrepancy was caused by the difference inthickness of the stratum corneum. As the result ofthe kinetic analysis, the diffusivity of dihy-droetorphine in the stratum corneum of dorsal skinwas sixfold that of abdominal skin. Bronaugh et al(1983) reported that the permeability of water, ureaand cortisone in a water vehicle through theabdominal skin of male Osborne-Mendel rats was3-10-fold that of dorsal skin, dependent on a 2�5-fold difference in the skin thickness as 13�8 mm ofabdominal and 34�7 mm of dorsal skin, but therewas no signi®cant difference in the lag time.Although several authors have investigated thepermeation properties and pharmacological effectsfollowing drug application to different sites of theskin, the relation of site-dependency to the per-meation rate of the drug should be considered. Inour previous report (Morimoto et al 1992b), thepermeability coef®cient of drugs with moderatelipophilicity, such as dihydroetorphine (log Pnearly 2), through human skin was slightly lowerthan that of rat abdominal skin. This suggests thatthe dihydroetorphine permeation rate through

Figure 5. Dihydroetorphine concentrations in plasma andbrain, and analgesic effect (% maximum effect; % MPE)during and after 4-h continuous infusion in hairless rats.Each point represents the mean and standard error (n� 4-5).ju, 0�9 mg hÿ1 kgÿ1; ds, 1�8mghÿ1 kgÿ1; mn,3�6mg hÿ1 kgÿ1; ^^, 9�0mg hÿ1 kgÿ1; 6 , saline infusion(1 mL hÿ1). Closed and open symbols of the concentrationdata indicate dihydroetorphine concentrations in the plasmaand brain, respectively. Vertical dotted lines indicate thetermination time of infusion. Solid lines indicate the calculatedmean values obtained by ®tting to the linear two-compartmentmodel with zero-order input and the sigmoidal effect-concen-tration equation, respectively.

Table 2. Pharmacokinetics and pharmacodynamic parameters of dihydroetorphine in hairless rats.

Infusion rate Pharmacokinetic parametersa Pharmacodynamic parametersb

(mg hÿ1 kgÿ1)

a (hÿ1) b (hÿ1) k21 (hÿ1) V1 (L kgÿ1) EC50 (ng mLÿ1) r

0�9 10�4 1�05 5�32 1�35 0�496 2�301�8 10�0 1�09 5�38 1�09 0�679 2�283�6 10�4 1�13 5�68 1�30 0�691 3�839�0 11�5 1�06 4�90 0�38 0�966 3�69Mean 10�3c 1�09c 5�46c 1�24c 0�708 3�03

aParameters were obtained by ®tting the linear two-compartment model with zero-order input to the mean dihydroetorphineconcentration during and after continuous infusion. bParameters were obtained by ®tting the sigmoidal equation to the plasmadihydroetorphine concentration and analgesic effect during and after infusion. cCalculated from parameters in 0�9, 1�8 and3�6mg hÿ1 kgÿ1.


human skin will approximate that through dorsalskin rather than abdominal skin of the hairless rat.

Kinetic analysis of skin permeation describedwell the ¯ux during topical application and after thedonor was removed. Mathematical modelling forthe release pro®le from the skin after removal ofthe transdermal delivery devices has been reportedby several authors (Tojo 1988; Kubota & Maibach1994; Roberts et al 1999). Kubota & Maibach(1994) reported that the half-life of the ¯ux afterremoving the donor was dominated by the thick-ness of the skin layer, the diffusivity and the par-tition coef®cient in the skin layer. When the tapewas used the half-life of dihydroetorphine waslarger than when the suspension was used, due to

the smaller diffusivity and the higher partitioncoef®cient in the stratum corneum by the tape.There are few reports regarding the mathematicalanalysis in consideration with peeling off the stra-tum corneum by using an adhesive device. In thisstudy, the dihydroetorphine tape caused the peelingoff of six or seven of the ten hypothetical layers ofthe stratum corneum. This indicated that 50% of thedihydroetorphine remaining in the stratum corneumwas lost with the tape (Figure 4B, D). These eventsmight occur in the in-vitro condition only, becausethe stratum corneum layer of excised rat skin waseasily removed after four days hydration of the skinin our previous report (Sugibayashi et al 1995). Itwas unclear if the dihydroetorphine tape wouldcause the peeling off of the stratum corneum layerduring in-vivo application.

The plasma concentration of dihydroetorphinewas maintained in the effective concentrationrange, but it was somewhat variable during theabdominal and dorsal applications of the tape(Figure 6). The variation of plasma concentration issuggested to have been caused by variation of thedrug input rate through the skin, which was in¯u-enced by the cutaneous perfusion rate and expan-sion or contraction of the skin in contact with thetape by movement of the conscious rats. However,the in-vivo thickness of viable skin, which was setat 10-fold of the stratum corneum thickness, wassuggested to be acceptable to estimate the in-vivo¯ux of dihydroetorphine, because the time lag forthe onset of plasma concentration after applicationof the tape was close to theoretical values.

The analgesic effect during application of the tapewas slightly lower than the theoretical value calcu-lated from 4-h infusion studies (Figure 6). The plasmaconcentration was similar to the theoretical value, sothat an analgesic tolerance might occur during theapplication of dihydroetorphine tape for 8 or 24 h.The relationship between the plasma concentrationand the analgesic effect during application of the tapehad a tendency towards a shift to the right from thetheoretical line, but it was not clear (data not shown).Tokuyama et al (1993) reported the development oftolerance to dihydroetorphine analgesia by repeatedsubcutaneous injection. However, although a con-tinuous infusion of morphine showed apparent tol-erance to its analgesic effect in normal rats (Ouellet &Pollack 1997), the tolerance of morphine did notoccur in pain-presented rats (Vaccarino et al 1993).We have con®rmed that continuous exposure ofdihydroetorphine develops analgesic tolerance andrewarding effect in hairless rats, but not in formalin-pain presented hairless rats (unpublished data).

In conclusion, it was con®rmed that dihy-droetorphine was permeable enough through the

Figure 6. Plasma dihydroetorphine concentration and analge-sic effect following topical application of the tape to abdom-inal and dorsal skin in hairless rats. Each point represents themean and standard error (n� 4±5). d, Abdominal skin; m,dorsal; 6 , placebo tape to dorsal skin. Vertical dotted linesindicate the time of removal of the tape. Solid lines indicate thecalculated mean values obtained from the combined equationof the diffusion model, compartment model and=or sigmoidaleffect±concentration equation. Dashed lines indicate the cal-culated mean values predicted by models after peeling off theouter layer of the stratum corneum with removal of the tape.


abdominal and dorsal skin of hairless rat to producean analgesic effect. Using drug-dispersed pressure-sensitive adhesive tape, dihydroetorphine wascontinuously delivered for 8 or 24 h. After the tapewas removed, the release pro®les of dihy-droetorphine from the skin were described by theFickian diffusion equation with peeling off of theouter layer of stratum corneum. Following theabdominal and dorsal applications of dihy-droetorphine tape, the plasma concentration andanalgesic effect were maintained at suitable levelsuntil the tape was removed. These results werepredictable by mathematical modelling.

AcknowledgementsThe authors would like to thank Mr YoheiSugiyama, Mr Liang Fang and Mr HidehikoMizuguchi for their technical assistance.

References

Barber, R. B., Rapoport, H. (1975) Synthesis of thebaine andoripavine from codeine and morphine. J. Med. Chem. 18:1074±1077

Bentley, K. W., Hardy, D. G. (1967) Novel analgesics andmolecular rearrangements in the morphine-thebaine group.III. Alcohols of the 6,14-endo-ethenotetrahydrooripavineseries and derived analogs of N-allylnormorphine and-norcodeine. J. Am. Chem. Soc. 89: 3281±3292

Bronaugh, R. L., Stewart, R. F., Congdon, E. R. (1983)Differences in permeability of rat skin related to sex andbody site. J. Soc. Cosmet. Chem. 34: 127±135

Chen, X. P., Guo, Q. D., Shi, T. S. (1996) Studies ontransdermal delivery system of dihydroetorphine hydro-chloride. Acta. Pharm. Sin. 31: 770±774

Crank, J. (1975) The Mathematics of Diffusion, SecondEdition. Oxford University Press, London

Finnin, B. C., Morgan, T. M. (1999) Transdermal penetrationenhancers: applications, limitations, and potential. J. Pharm.Sci. 88: 955±958

Franz, T. J., Tojo, K., Shah, K. R., Kydonieus, A. (1992)Transdermal delivery. In: Kydnieus, A. (ed.) Treatise onControlled Drug Delivery: Fundamental, Optimization,Applications. Marcel Dekker, New York, pp 341±421

Guo, Q. D., Chen, X. P., Shi, T. S. (1997) Effect of vehicle pH,drug concentration and Azone on the snake skin permeation ofdihydroetorphine hydrochloride. Chin. Pharm. J. 32: 349±352

Jolicoeur, L. M., Nassiri, M. R., Shipman, C., Choi, H. K.,Flynn, G. L. (1992) Etorphine is an opiate analgesic physi-cochemically suited to transdermal delivery. Pharm. Res. 9:963±965

Kubota, K., Maibach, H. I. (1994) Signi®cance of viable skinlayers in percutaneous absorption and its implication inmathematical models: theoretical consideration based onparameters for betamethasone 17-valerate. J. Pharm. Sci.83: 1300±1306

Li, Y. S., Li, J. L., Qu, F. L. (1996) Using dihydroetorphine tocontrol cancer pain. Chin. J. New Drug 5: 122±123

Morimoto, Y., Hatanaka, T., Oguchi, M., Sugibayashi, K.,Kobayashi, M., Kimura, M. (1992a) A screening method forpercutaneous absorption enhancers appropriate for adhesivematrix devices. S. T. P. Pharma Sciences 2: 253±258

Morimoto, Y., Hatanaka, T., Sugibayashi, K., Ohmiya, H.(1992b) Prediction of skin permeability of drugs: compar-ison of human and hairless rat skin. J. Pharm. Pharmacol.44: 634±639

Ohmori, S., Hayashi, T., Kawase, M., Saito, S., Morimoto, Y.(2000a) Quantitative determination of dihydroetorphine inrat plasma and brain by liquid chromatography-tandem massspectrometry. J. Chromatogr. B 740: 253±263

Ohmori, S., Hayashi, T., Kawase, M., Saito, S., Morimoto, Y.(2000b) Pharmacokinetic and Pharmacodynamic evalua-tions of a potent analgesic dihydroetorphine in hairless rat.J. Pharmacol. Exp. Ther. (in press)

Ohmori, S., Suguyama, Y., Morimoto Y. (2000c) A protectiveeffect against undesirable increase of dihydroetorphine per-meation through damaged skin by using pressure-sensitiveadhesive tape with an ethylene-vinyl acetate co-polymermembrane. Biol. Pharm. Bull. (in press)

Okamoto, H., Hashida, M., Sezaki, H. (1988) Structure-activ-ity relationship of 1-alkyl- or 1-alkenylazacycloalkanonederivatives as percutaneous penetration enhancers. J.Pharm. Sci. 77: 418±424

Ouellet, D. M. C., Pollack, G. M. (1997) Pharmacodynamicsand tolerance development during multiple intravenousbolus morphine administration in rats. J. Pharmacol. Exp.Ther. 281: 713±720

Roberts, M. S., Anissimov, Y. G., Gonsalvez, R. A. (1999)Mathematical models in percutaneous absorption. In:Bronaugh, R. L., Maibach, H. I. (eds) Percutaneous Absorp-tion. Marcel Dekker, New York, pp 3±55

Sato, K., Sugibayashi, K., Morimoto, Y. (1991) Speciesdifferences in percutaneous absorption of nicorandil. J.Pharm. Sci. 80: 104±107

Southam, M. A. (1995) Transdermal fentanyl therapy: systemdesign, pharmacokinetics and ef®cacy. Anticancer Drugs 6(Suppl. 3): 29±34

Stinchcomb, A. L., Paliwal, A., Dua, R., Imoto, H., Woodard,R. W., Flynn, G. L. (1996) Permeation of buprenorphine andits 3-alkyl-ester prodrugs through human skin. Pharm. Res.13: 1519±1523

Sugibayashi, K., Sakanoue, C., Morimoto, Y. (1989) Utility oftopical formulations of morphine hydrochloride containingAzone and N-methyl-2-pyrrolidone. Sel. Cancer Ther. 5:119±128

Sugibayashi, K., Kobayashi, D., Nakagaki, E., Hatanaka, T.,Inoue, N., Kusumi, S., Kobayashi, M., Kimura, M., Mor-imoto, Y. (1995) Differences in enhancing effect ofl-methanol, ethanol and their combination between hairlessrat and human skin. Int. J. Pharm. 113: 189±197

Suh, H. R., Jun, H. W. (1996) Effectiveness and mode ofaction of isopropyl myristate as a permeation enhancer fornaproxen through shed snake skin. J. Pharm. Pharmacol. 48:812±816

Tojo, K. (1987) Mathematical modeling of transdermal drugdelivery. J. Chem. Eng. Jap. 20: 300±308

Tojo, K. (1988) Concentration pro®le in plasma after trans-dermal drug delivery. Int. J. Pharm. 43: 201±205

Tokuyama, S., Takahashi, M., Kaneto, H. (1993) Antinocicep-tive effect of dihydroetorphine and its tolerance=dependenceliability in mice. Biol. Pharm. Bull. 16: 774±777

Vaccarino, A. L., Marek, P., Kest, B., Ben-Eliyahu, S., Couret,L. C., Kao, B., Liebeskind, J. C. (1993) Morphine fails toproduce tolerance when administered in the presence offormalin pain in rats. Brain Res. 627: 287±290

Wang, C. Y., Li, H., Cheng, N. S., Zhong, J. H., Wang, B.(1999) An investigation on the situation of utilization ofnarcotic medications for cancer pain patients. Chin. J. DrugDepend. 8: 210±213


WHO Expert Committee (1999) WHO Expert Committeeon Drug Dependence, Thirty-®rst Report. World HealthOrganization Technical Report Series 887, WorldHealth Organization, Geneva

Wu, G. Q., Sun, Y. (1991) Dihydroetorphine hydrochloride formoderate and severe cancer pain. Chin. J. Oncol. 13: 64±67

Zang, D. L. (1999) A survey on the utilization of narcoticanalgesic drugs in general hospitals of Qingdao. Chin. J.Drug Depend. 8: 204±206


transdermal delivery of the potent analgesic dihydroetorphine: kinetic analysis of skin permeation...

Documents