.I(o)l7I2.'\}~\L OF
`
`
`i’ harmaoy and
`Pharmacology
`
`
`
`
`
`n y.
`
`002243573‘
`
`JUN 0 71999
`
`HE.0LTf-4 SCJENCES L!EP.AR!ES
`
`
`;i>.’%
`I}
`:7 .
`
`1'“ Vial" H
`
`
`;~’.-~~‘:-Hygx 3..
`"($2-"
`
`
`
`,
`_
`_
`_
`Royal Pharmaceutical Soclety of Great Brltam
`
`
`
`Page 1 of 10
`
`Kaken Exhibit 2014
`Au.“ “Kaken
`IPR20l7-00190
`
`

`

`
`
`
`
`
`Journal of Pharmacy and Pharmacology
`
`Published by The Royal Pharmaceutical Society of Great Britain
`
`
`
`
`
`
`
`
`
`l Larnbeth High Street. London SE] TJN Telephone 0l7l~735 9141
`FAX DITI-320 3917
`
`
`
`
`
`
`
`
`
`
`
`
`
`E-Mai] jpp@rpsgb.org.u.k
`http:!."dspace.diaJ.pipex.comfjpp.F
`
`
`
`Editorial Oifice for Japan: Department of Pliannaceutics. Kanazawa University. Takara-machi I3-l, Kanazawa 920
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`EDITOR Dal CHAMBERLAIN
`
`
`
`
`
`ASSISTANT EDITOR DR A L SUGDEN
`
`
`
`
`
`
`
`
`EDITORIAL ASSISTANT G M MCMAHON
`
`
`
`
`
`
`
`EDITORIAL BOARD
`
`
`
`
`ASSOCIATE EDITOR {JAPAN}
`
`
`
`
`
`PROFESSOR A TSUJI
`
`
`
`
`
`K L Auous, Lawrence, Kansas
`
`
`
`
`
`L L AUGSBURGER, Baltimore, Maryland
`
`
`
`
`B W BARRY, Bradford
`
`
`
`
`E J BEJAR, San Diego. California
`
`
`
`
`
`E BEUBLER, Graz, Austria
`
`
`
`
`N BODOR, Gainesville, Florida
`
`
`
`
`D D BREIMER, Leiden, The Netherlands
`
`
`
`
`
`D J BURGESS, Storrs, Connecticutt
`
`
`
`
`
`
`Y W CHIEN, Piscataway. New Jersey
`
`
`
`
`D A COWAN, London
`
`
`
`
`
`
`
`S P DENYER (Chairman), Brighton
`F J EVANS, London
`
`
`
`
`J J FERRARA, New Orleans, Louisiana
`
`
`
`
`
`J L FORD, Liverpool
`
`
`
`
`D R FRIEND, Menlo Park, California
`
`
`
`
`
`D GANDERTDN, London
`
`
`E P GOLDBERG. Gainesville, Florida
`
`
`
`
`
`D J GREENBLATT, Boston, Massachusetts
`
`
`
`
`
`
`
`
`
`M J GROVES, Chicago, Illinois
`R H GUY, Archarnps, France
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`A HOLT. Edmonton, Alberta
`
`
`
`P G JENNER, London
`
`
`
`
`I W KELLAWAY, Cardiff‘
`
`
`
`
`
`
`
`
`W E LINDUP, Liverpool
`H I MAIBACH, San Francisco, California
`
`
`
`
`
`T MANDAL, New Orleans, Louisiana
`
`
`
`
`
`R J NAYLOR, Bradford
`
`
`
`
`S DIE, San Francisco, California
`
`
`
`
`.1 W PHILLIS, Detroit, Michigan
`
`
`
`
`
`K D RAINSFORD, Sheffield
`
`
`
`
`W R RAVIS, Auburn, Alabama
`
`
`
`
`
`J E RJVIERE, Raleigh, North Carolina
`
`
`
`
`
`J R Roamson, Madison, Wisconsin
`
`
`
`
`
`A ROLLAND, The Woodlands, Texas
`
`
`
`
`
`H H TAI, Lexington, Kentucky
`
`
`
`
`
`B TESTA. Lausanne, Switzerland
`
`
`
`
`E TOMLINSON, The Woodlands, Texas
`
`
`
`
`B WIDDOP, London
`
`
`
`D E WURSTER, Iowa City, Iowa
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`COPYRIGHT i_'.- 1999 Journal ofPharmac_1' and Phannacologjr.
`
`
`
`
`
`
`
`Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by the Journal of Phannacy
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`and Pharmacology for libraries and other users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service. provided that
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`the base fee of $15 per article is paid directly to CCC, 21 Congress St. Salem, MA D1970. USA.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`CODE: 0022-35'.l3/90.
`
`
`Annual joint subscriptions H999) to Journal of Phannacy and Phannacology and Pharmacy and Pharmacology Communications: UK £450; Europe
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`£470: America and Japan $790; rest of world £490. Single issues: UK £30; Europe and rest of world £35: America and Japan $56.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`All orders should he sent to: The Pharmaceutical Press. PO Box 15!, Wallingford. Oxon (JXIO SQU. UK. Telephone +44 {D}l49l S19 2'f2; FAX +44
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`(0)1491 829 292: E-Mail rpsgb@cahi.org
`
`
`
`
`
`Claims for missing copies cannot be considered unless received within three months of publication.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Page 2 of 10
`
`Page 2 of 10
`
`

`

`Phann. Pharmacol. I999. 5|: 27l-278
`lneeived August 25. 1998
`Amepted November 2. I998
`
`.1’
`
`I999 J. Phann. Pharmacol.
`
`Drug Permeation through the Three Layers of the
`Human Nail Plate
`
`YOICHI KOBAYASHI. MISAO MlYAMOTO*. KENJI SUGIBAYASHI
`AND YASUNORI MORIMOTO
`
`Faculty of Phamtaceutical Sciences, Josai University. [-1 Keyakidai, Sakado, Saitama 350-0295 and
`*Nissan Chemical Co., Ltd, 3-7-1 Kanda-Nishiki-cho, Chiyoda-ku, Tokyo 101-0054, Japan
`
`Abstract
`
`The in-vitro permeation characteristics of a water soluble model drug, 5-fiuorouraeil, and a
`poorly water soluble model drug, flurbiprofen, were investigated through three layers of the
`human nail plate (namely, the dorsal, intermediate and ventral nail plates), using a modified
`side-by-side diffusion cell. The dorsal-filed nail plate, the ventral-filed nail plate and the
`dorsal-and-ventral-filed nail plate were prepared to known thicknesses and then used with
`the full-thickness nail plate to investigate the penneation characteristics of each single
`layer.
`Most of the lipids in the human nail plate were found in the dorsal and ventral layers.
`The rank orders of the penneation fluxes for 5-fluorouraeil and flurbiprofen were both:
`dorsal-and-ventral—fi|ed nail plate > dorsal-filed nail plate > ventral-filed nail plate > full-
`thickness nail plate. With respect to 5-fluorouracil permeation through each single layer,
`the permeability coefficient of the intemiediate layer was higher than those of other single
`layers. However in the case of flurbiprofen, the permeability coefficient of the ventral layer
`was higher than other single layers. The diffusion coefficients of 5—fluorouracil and
`flurbiprofen in the dorsal layer were the lowest of any single layer. The drug concentration
`in each layer was estimated using each respective permeation parameter. The drug
`concentration in the nail plate was observed to be dependent on the solubility and the
`flux of the drug.
`From these findings, we suggest that the human nail plate behaves like a hydrophilic gel
`membrane rather than a lipophilic partition membrane and that the upper layer functions as
`the main nail barrier to drug permeation through its low diffusivity against the drugs.
`
`Although the human nail plate is generally con-
`sidered to be composed of either one layer or three
`layers,
`the
`three-layer model
`is more widely
`accepted (Spearman 1978; Dawber 1980). The
`upper,
`the middle and the lower layers of the
`human nail plate are called the dorsal. intermediate
`and ventral nail plates, respectively (Figure I). In
`addition, the tissue under the nail plate, called the
`nail bed, consists of viable epidennis. Information
`regarding the distribution of fungi
`in the human
`nail plate is required for the treatment of onycho-
`mycosis trichophytica. It has been reported that the
`dorsal nail plate,
`the ventral nail plate,
`the sub-
`ungual keratin and the eponychium are infected by
`fungi such as Trichophyton rubrum and Tricho-
`
`Correspondence: Y. Morimoto. Faculty of Pharmaceutical
`Sciences. Josai University.
`l-l Keyakidai. Sakado, Saitama
`350-0295, Japan.
`
`phyton mentagrophytes (Sagher 1948; Jillson &
`Piper 1957; Akiba 1971). Of these nail compo-
`nents, it is the ventral nail plate and the subungual
`keratin that are found to be infected in most cases.
`
`it is of great importance to understand
`Therefore,
`the specific properties of each layer in the human
`nail plate.
`Onychomycosis trichophytica has been treated
`mainly with oral antifungal medication (Piepponen
`et al 1992; Villars & Jones 1992). With systemic
`treatment, an antifungal drug may be delivered
`from the blood in the dermis under the nail bed and
`
`the nail matrix to the ventral nail plate and sub-
`ungual keratin (Matthieu et al 1991). On the other
`hand, antifungal drugs are not delivered to the
`ventral nail plate and the subungual keratin fol-
`lowing topical treatment because of low nail-plate
`permeability. Each layer has unique physical
`
`Page 3 of 10
`
`

`

`
`
`\TNCHlKUBAYASHlETAL
`
`
`
`
`
`
`E
`
`h"
`Dome mm
`
`
`
`Dorsal nail plate
`Intermediate nail plate
`¢
`\
`
`Ventral nail plate
`
`Figure 1.
`
`
`
`
`Schematic diagram of the human nail plate.
`
`
`
`
`
`
`
`
`
`
`
`
`Preparatiwr ofnai! plates
`
`
`
`
`
`
`
`
`
`Tip nail pieces were obtained from the fingers of
`
`
`
`
`
`
`
`
`
`healthy volunteers using nail clippers. A lot of nail
`
`
`
`
`
`
`
`pieces, clipped from the same volunteer
`(male.
`
`
`
`
`
`
`
`
`
`aged 24 years) were used in the permeation study
`
`
`
`
`
`
`
`and for microscopic observation. Forty nail pieces
`
`
`
`
`
`
`
`obtained from other healthy volunteers (10 males,
`
`
`
`
`
`
`
`
`
`mean age 24 years} were used in the measurement
`
`
`
`
`
`
`
`
`
`of total lipid. The thicknesses of nail pieces were
`
`
`
`
`
`
`measured with a micrometer
`l_Mitutoyo Corp,
`
`
`
`
`
`
`Japan) equipped with pointed metal attachments.
`
`
`
`
`
`
`
`
`
`
`The thickness ratio of each layer of the human nail
`
`
`
`
`
`
`plate (dorsal:intermediate:ventral) was assessed
`
`
`
`
`
`
`
`
`to be 3 :5 : 2 after examining several reports (Jillson
`
`
`
`
`
`
`
`& Piper 1957; Spearman l978; Stijttgen 8:’ Bauer
`
`
`
`
`
`
`
`1982). The dorsal-filed nail plate, the ventral—filed
`
`
`
`
`
`
`
`
`nail plate and the dorsal- and ventral-filed nail plate
`
`
`
`
`
`
`
`were prepared to known thicknesses with sand
`
`paper
`
`
`
`
`
`
`
`
`
`Deternrirrarrbrr ofsrrlubiliries and nail plate/tieltide
`
`
`
`
`partition r'0ejfic'ienIs qf drugs
`
`
`
`
`5-Fluorouracil and flurbiprofen aqueous suspen-
`
`
`
`
`
`
`
`
`sions were mixed with a magnetic stirrer at 37°C.
`
`
`
`
`
`
`
`
`After 24 h, each suspension was subjected to fil-
`
`
`
`
`
`
`
`tration (Ekicrodisc 3: German Sciences Japan, Ltd,
`
`
`
`
`
`
`
`Tokyo). The filtrate was immediately diluted with
`
`
`
`
`
`
`
`
`distilled water or methanol to obtain samples for
`
`analysis.
`
`
`
`
`
`
`
`The finger nail pieces (full-thickness nail plate.
`
`
`
`
`
`
`
`ventral-filed nail plate, dorsal-filed nail plate and
`
`
`
`
`
`dorsal-and-ventral-filed
`plate) were
`the
`nail
`
`
`
`
`
`
`
`weighed with an electronic balance (IL-200, Chyo
`
`
`
`
`
`
`Balance Corp.). They were immersed in half-con-
`
`
`
`
`
`
`
`centration solutions (lmL) of 5-fluorouracil or
`
`
`
`
`
`
`
`
`flurbiprofen solubility at 37°C for 48 h. After
`
`
`
`
`
`
`
`
`
`removal of the nail piece from each solution. the
`concentration of the solution was measured. The
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`in the nail pieces were extracted with
`drugs
`
`
`
`
`
`
`
`methanol (l0mL) three times. The extracted sol-
`
`
`
`
`
`
`
`
`vent was evaporated under nitrogen gas at 60"C.
`
`
`
`
`
`
`
`
`Subsequently,
`1 ml. of distilled water or methanol
`
`
`
`
`
`
`
`
`was added to obtain samples for analysis. We
`
`
`
`
`
`
`
`
`assumed that
`the weight (volume) ratio of each
`
`
`
`
`
`
`
`
`
`
`layer of the nail plate was 3:5:2. The apparent
`
`
`
`
`
`
`
`
`partition coefficient of the drugs was calculated as
`
`
`
`
`
`
`
`
`the concentration ratio in the nail plate/ vehicle at
`37°C.
`
`
`Permeation studies
`
`
`
`
`
`
`
`
`
`To overcome large variations in nail permeabilities
`
`
`
`
`
`
`
`due to individual differences in barrier properties,
`
`
`
`
`
`
`
`
`
`nail pieces (20-35 mg) from the third finger and
`
`
`
`
`
`
`
`
`
`
`the fifth finger of the same volunteer were used in
`
`
`
`
`
`
`
`
`
`therefore,
`and,
`properties
`drug perrneabilities
`
`
`
`
`
`
`
`
`
`through each layer may be different. Only a few in-
`
`
`
`
`
`
`
`vitro drug permeation studies have been performed
`
`
`
`
`
`
`
`
`
`on the human nail plate. Merlin & Lippold (l99?a)
`
`
`
`
`
`
`
`
`
`reported that the human nail plate and the keratin
`
`
`
`
`
`
`membrane from bovine hooves (the model mem-
`
`
`
`
`
`
`
`
`the human nail plate). behave like
`brane of
`
`
`
`
`
`
`
`hydrophilic gel membranes. Walters et al (1983,
`
`
`
`
`
`
`
`
`1985) also suggested that
`the human nail plate
`
`
`
`
`
`
`behaves like a hydrophilic gel membrane. Fur-
`
`
`
`
`
`
`
`thermore, they suggested that it has an additional
`
`
`
`
`lipophilic route for permeation.
`
`
`
`
`
`
`
`
`
`In the present study, the distribution of lipids in
`
`
`
`
`
`
`
`
`
`the human nail plate was examined in detail. In
`
`
`
`
`
`
`
`
`order to investigate the nail penneability of a water
`
`
`
`
`
`
`
`soluble model drug, 5-fluorouracil, and a poorly
`
`
`
`
`
`
`soluble model drug, flurbiprofen, a modified side-
`
`
`
`
`
`
`
`by-side diffusion cell was used. The drug permea-
`
`
`
`
`
`
`tion characteristics of each single layer were
`
`
`
`
`
`
`
`
`investigated by filing upper or lower layers, or both,
`
`
`
`
`
`
`
`of the human nail plate. 5-Fluorottracil and flurbi-
`
`
`
`
`
`
`
`
`profen perrneations, from the ventral to the dorsal
`
`
`
`
`
`
`
`
`nail plate, were compared with those from the
`
`
`
`
`
`
`
`
`
`dorsal to the ventral nail plate. The amounts of 5-
`
`
`
`
`
`
`
`fluorouracil and fiurbiprofen in the nail plate were
`
`
`
`
`
`
`
`also compared with those calculated from the per-
`
`
`
`
`
`meation parameters of each drug.
`
`
`
`
`
`
`
`
`Materials and Methods
`
`
`
`Materials
`
`
`
`
`
`
`
`
`5-Fluorouracil was obtained from Tokyo Kasei
`
`
`
`
`
`
`
`Kogyo Co., Ltd (Tokyo, Japan). Flurbiprofen was
`
`
`
`
`
`
`
`supplied by Kaken Pharmaceutical Co., Ltd (Chiba,
`
`
`
`
`
`
`Japan). Triolein (practical grade), cholesterol and
`
`
`
`
`
`
`
`vaniline (for measuring total
`lipid) and Sudan
`
`
`
`
`
`
`
`
`Black B (for lipid staining) were purchased from
`
`
`
`
`
`
`
`Wako Pure Chemical
`Industries, Ltd (Osaka,
`
`
`
`
`
`
`Japan). Other reagents were obtained from com-
`mercial sources.
`
`
`
`Page 4 of 10
`
`Page 4 of 10
`
`

`

`DRUG PERMEATION THROUGH HUMAN NAIL PLATE
`
`
`
`
`
`
`
`
`273
`
`
`
`
`
`
`
`
`the permeation studies of 5-fluorouracil and flur-
`
`
`
`
`
`
`biprofen, respectively. The full-thickness nail plate,
`
`
`
`
`
`
`
`
`the dorsal-filed nail plate, the ventral—filed nail plate
`
`
`
`
`
`
`and the dorsal-and-ventral~filed nail plate were
`
`
`
`
`
`
`sandwiched between 2 adapters made of poly-
`
`
`
`
`
`
`propylene with an O-shaped ring (effective diffu-
`
`
`
`
`
`
`
`
`sion area, 0-O49 cm2) and mounted in a side-by-side
`
`
`
`
`
`
`
`
`
`
`
`diffusion cell
`(2-5 mL or 1-5 mL) with a water
`
`
`
`
`
`
`
`
`jacket connected to a water bath at 37°C. 5-Fluor-
`
`
`
`
`
`ouracil- or flurbiprofen-suspension was applied to
`
`
`
`
`
`
`
`
`
`
`the dorsal nail plate side of the diffusion cell, and
`
`
`
`
`
`
`
`
`
`the ventral nail side was filled with distilled water
`
`
`
`
`
`
`
`
`or 40% polyethylene glycol 400 (to maintain the
`
`
`
`
`
`
`
`
`
`sink condition). In the case of vehicle application to
`
`
`
`
`
`
`the ventral nail plate side, 5-fluorouracil- or flur-
`
`
`
`
`
`
`
`biprofen-suspension was placed in the ventral nail
`
`
`
`
`
`
`
`
`plate side compartment, and the dorsal nail plate
`
`
`
`
`
`
`
`
`side was filled with distilled water or 40% poly-
`
`
`
`
`
`
`
`ethylene glycol 400. No preservative was added
`because the receiver solution was clear even at the
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`end of the experiment. The permeation was mea-
`
`
`
`
`
`
`
`
`sured by sampling the receiver side solution at
`
`
`
`
`
`
`predetermined times. The experimental period was
`
`
`
`
`
`
`
`‘i’ days (5-fluorouracil) or 19 days (flurbiprofen)
`
`
`
`
`
`
`
`
`because of the low nail permeability of drugs.
`
`
`
`
`
`Measurement of total lipid
`
`
`
`
`
`
`
`
`Total lipid was measured according to the method
`
`
`
`
`
`
`
`
`
`
`of Knight et al (1972). After nail pieces from the
`
`
`
`
`
`
`
`fingers of healthy volunteers were prepared into
`
`
`
`
`full-thickness, ventral-filed, dorsal-filed and dorsal-
`
`
`
`
`
`
`
`and-ventral-filed layers, they were immersed in a
`
`
`
`
`
`
`
`chloroform—methanol (3: l) mixture for 24h to
`
`
`
`
`
`
`
`extract lipids. The extracted solvent was evaporated
`
`
`
`
`
`
`
`
`
`under nitrogen gas at 50°C. Sulphuric acid (300 pl.)
`
`
`
`
`
`
`
`
`
`was added to the sample and heated in boiling
`
`
`
`
`
`
`
`
`
`Water for 5 min. After cooling to room temperature,
`
`
`
`
`
`
`
`
`
`2mL of phosphovaniline test solution (06% w/v
`
`
`
`
`
`
`aqueous vaniline solution/phosphoric acid; 1:4)
`was added and the mixture was incubated for
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`15min at 37°C. Then,
`the absorbance was mea-
`
`
`
`
`
`
`
`sured at 535nm using a spectrophotometer (UV-
`
`
`
`
`
`IGOA, Shimadzu Seisakusho, Kyoto, Japan). A
`
`
`
`
`
`
`
`
`mixture of cholesterol and triolein (3 : 1) was used
`as a reference standard.
`
`
`
`
`
`
`
`
`
`
`
`
`Observation of iipid dimibution
`
`
`
`
`
`
`
`
`
`finger of a healthy
`A nail piece from the first
`
`
`
`
`
`
`
`
`volunteer was used to observe nail plate lipid dis-
`
`
`
`
`
`
`
`
`tribution. After the nail plate was frozen in an
`
`
`
`
`embedding medium (Tissue-Tek; Sakura Fine-
`
`
`
`
`
`
`
`
`
`technical Co., Ltd, Tokyo),
`it was sliced with a
`
`
`
`
`
`microtome (IBC Model Minotome Microtome
`
`
`
`
`
`International Equipment Company). A
`Cryostat:
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`section of this nail plate was stained with Sudan
`
`
`
`
`
`
`
`
`
`Black B in isopropyl alcohol
`(5 mgmL_') for
`
`
`
`
`
`
`
`
`
`
`30 min. After being washed with 50% methanol for
`
`
`
`
`
`
`
`
`
`about 303,
`it was observed under a microscope
`
`
`
`
`
`(New VANOX. Olympus, Tokyo).
`
`
`
`Analytical method
`
`
`
`
`
`5-Fluorouracil and flurbiprofen levels were deter-
`
`
`
`
`
`mined by high-performance liquid chromatography
`
`
`
`
`
`
`
`(HPLC). Sample solutions were injected into the
`
`
`
`
`
`
`
`HPLC instrument. which was composed of a pump
`
`
`
`
`
`system (LC-10A, Shimadzu Seisakusho, Kyoto,
`
`
`
`
`
`
`
`Japan), a UV detector (SPD-IOA, Shimadzu), a
`
`
`
`
`
`fluorescence detector (RF-IOAXL, Shimadzu), a
`
`
`
`
`
`Chromatopack (C-RSA. Shimadzu), a system con-
`
`
`
`
`
`
`(SCL-IOA, Shimadzu), an auto injector
`troller
`
`
`
`
`
`
`
`(SIL-10A, Shimadzu) and a reverse phase column
`
`
`
`
`
`
`(Inertsil ODS 250 mmx-4-6 mm i.d., GL Sciences
`
`
`
`
`
`
`
`
`Ine.. Tokyo). The mobile phase for flurbiprofen was
`
`
`
`
`
`
`
`
`
`01% phosphoric acid/acetonitrile (40:60),
`the
`
`
`
`
`
`
`
`
`
`flow rate was 1 mLn1in' 1 and fluorescence detec-
`
`
`
`
`
`
`
`
`tion was conducted at an excitation wavelength of
`
`
`
`
`
`
`
`
`
`260nm and an emission wavelength of 313 nm.
`
`
`
`
`
`
`
`
`The mobile phase for 5-fluorouracil was 0-1%
`
`
`
`
`
`
`
`phosphoric acid acetonitrile (98:2), the flow rate
`
`
`
`
`
`
`
`
`
`
`was 1 mL min‘ , and UV detection was conducted
`
`
`
`
`
`
`at a wavelength of 270 nm.
`
`
`
`Results and Discussion
`
`
`
`
`
`
`
`
`
`Lipid distribution in the nail plate
`
`
`
`
`
`
`
`
`
`
`The lipid content in the human nail plate was found
`to be much lower than that in the stratum corneum
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`of skin (Walters & Flynn 1983). However, various
`
`
`
`
`
`
`
`
`lipids such as long-chain fatty acids,
`free fats,
`
`
`
`
`
`
`cholesterol, squalene and phospholipids are present
`
`
`
`
`
`
`
`
`
`in the human nail plate (Spearman 1978; Hirose et
`
`
`
`
`
`
`
`
`
`
`al 1990). It is thought that the lipid distribution and
`
`
`
`
`
`
`
`
`concentration in the human nail plate may affect
`
`
`
`
`
`
`
`the drug permeation, particularly the partition of
`
`
`
`
`drugs into the membrane.
`
`
`
`
`
`
`Table l shows the total-lipid concentration (CL)
`
`
`
`
`in the ventral-filed, dorsal-filed, dorsal-and-ventral-
`
`
`
`
`
`
`filed and full—thickness layers. The ventral-filed and
`
`
`
`
`
`
`dorsal—filed layers had a high total-lipid con-
`
`
`
`
`
`
`
`
`centration; about l-5 and 3-0 times those observed
`
`
`
`
`
`
`in the dorsal-and-ventral-filed layer. The total-lipid
`
`
`
`
`
`
`
`concentrations of each single layer were calculated
`
`
`
`
`
`
`
`from the experimental data of the ventral-filed,
`
`
`
`
`dorsal—filed
`and dorsal-and-ventral-filed
`layers
`
`
`
`
`
`
`
`(Table 1). The total-lipid concentration in the
`
`
`
`
`
`
`
`
`
`ventral
`layer was the highest
`in the human nail
`
`
`
`
`
`
`
`
`
`plate. In contrast, the intermediate layer had a low
`
`
`
`
`
`Page 5 of 10
`
`Page 5 of 10
`
`

`

`274
`
`YOICHI KOBAYASHI ET AL
`
`Table l. Experiniental (a) and calculated lipid (b) concentration in the nail plate.
`
`Experimental Layer
`(thickness ratio)
`
`CL (fig mg l)
`
`Calculated Layer
`(thickness ratio)
`
`c._ (pg mg ‘ ')
`
`Ventral-tiled
`(3:5)
`
`4-33:‘..:U-45
`
`Dorsal
`(3)
`
`6-82
`
`Dorsal-filed
`(5:2)
`
`8~7():t l~l5
`
`lntennediatc
`(5)
`
`2-83
`
`Dorsal-and-ventral-filed
`(5)
`
`2-83:l:0-28
`
`Ventral
`(2)
`
`236
`
`I
`
`Full-thickness
`(3 15:2)
`
`7-5811-Z0
`
`Full-thickness
`(3 :5: 2)
`
`850
`
`CL: total lipid concentration. Each value in Table 1 represents the meanis.e.m. (n : IO).
`
`total—lipid concentration. The rank order of the total
`lipid concentration in each single layer was ventral
`layer > dorsal layer > intemtediate layer. The total
`lipid concentration in the full-thickness nail plate
`(Table l) was calculated using those of each single
`layer. This lipid concentration was very similar to
`the value for the full-thickness nail plate obtained
`from the experiment.
`Figure 2 shows a micrograph of a human nail
`plate cross section. The lipids in the nail plate were
`stained with Sudan Black B. Sudan Black B can
`
`lipids. such as neutral fats and phos-
`stain most
`pholipids. Most lipids were observed in the upper
`and lower parts of the human nail plate. Moreover.
`uneven pans of the ventral side contained a lot of
`lipids. whereas no lipids were observed in the
`intemiediate part of the nail plate.
`From these results. we suggest that the dorsal and
`ventral layers in the human nail plate contain some
`lipids. whereas the intermediate layer, which is the
`main nail body of the human nail plate, contains
`few lipids.
`
`
`
`Figure 2.
`Micrograph of a human nail plate. D: dorsal nail
`plate side. V: ventral nail plate side. This nail plate section was
`stained with Sudan Black B.
`
`Permeation parameters of each layer
`The data analysis is based on Fick‘s first law:
`
`dQ/dt = DmAC/h
`
`(l)
`
`where dQ/dt is the steady state permeation rate. D",
`is the diffusion coefficient of the drug in the
`membrane. AC is the concentration differential of
`
`the drug in the membrane, and h is the membrane
`thickness.
`ln the case of a sink condition. AC in
`
`Equation I can be replaced with the product of the
`solubility of the drug in the donor solvent (CV) and
`the membrane/donor vehicle partition coefficient
`of the drug (Km):
`
`dQ/dt = DmKmCV/h
`
`The pemteability coefficient of the drug (P)
`given by:
`
`P = DmKm/h
`
`(2)
`
`is
`
`(3)
`
`Figure 3 shows the permeation profiles of 5-fluor-
`ouracil and flurbiprofen through the ventral-tiled,
`dorsal-tiled.
`dorsal-and-ventral-tiled
`and
`full-
`
`layers. The rank orders for the 5-
`thickness nail
`fluorouracil and fiurbiprofen fluxes were both:
`dorsal-and-ventral-filed layer > dorsal-tiled layer >
`ventral—filed layer>full-thickness layer. The lag
`times for 5-fiuorouracil and fiurbiprofen permea-
`tions through full-thickness nail plates were about
`2-5 and 11 days, respectively. Mertin & Lippold
`(l997a) also suggested that steady-state perrnea—
`tions of nicotinic acid esters through human nail
`plates were obtained after
`l0—80h.
`In addition,
`they reported that
`steady-state permeation of
`chloramphenicol, from an aqueous suspension or a
`nail lacquer. through the human nail plate occurred
`after a lag time of 200h and 400 h, respectively
`(Mertin & Lippold 1997b).
`Table 2 shows the 5-fluorouracil and fiurbiprofen
`pemteation parameters of the ventral-filed. dorsal-
`filed, dorsal-and-ventral-filed and full-thickness
`
`nail layers. The apparent pemieability cocfficicnts
`of each drug through these nail plates were calcu-
`
`Page 6 of 10
`
`

`

`DRUG PERMEATION THROUGH HUMAN NAIL PLATE
`
`as§
`
`
`
`
`
`5-fluorouracilpermeated(ygcm‘2)
`
`
`
`
`
`Flurbiprofenpermeated(pgcm‘2)
`
`_. N O
`
`6o
`
`O)O
`
`N4:.onooo
`
`4
`
`Time (days)
`
`10
`
`1 5
`
`Time (days)
`
`Figure 3.
`Pcmication profiles of 5-fluorouracil (a) and flurbiprofcn (b) through ventral-filed (A). dorsal—filcd (U). dorsal-and
`ventral-filed (O) and full-thickness (C) nail layers. Each value represents the mean:l:s.e.m. (n =3-5).
`
`lated using Equations 2 and 3. In all permeation
`studies. CV values are given in saturated con-
`centrations of 5—fluorouracil (17-lmg mL") and
`flurbiprofen (27-7;tg mL' ')
`at
`37“C. because
`donor vehicles are aqueous suspensions. Although
`the ventral-filed nail layer had approximately the
`same 5-fluorouracil permeability coefficient as the
`dorsal-filed layer, the dorsal-and-ventral-filed layer
`had a higher permeability coefficient
`than those
`layers. On the other hand. the flurbiprofen penne-
`ability coefficient of the dorsal-and-ventral-filed
`layer was also higher than those of the ventral-filed
`and dorsal-filed layers. The nail plate/vehicle par-
`tition coefficients of each drug were calculated as
`concentration ratios in the ventral-filed, dorsal-
`filed, dorsal-and-ventral-filed and full-thickness
`layers/water at 37"C, respectively. The nail plate/
`vehicle partition coefficients of 5-fluorouracil and
`flurbiprofen were not very different between the
`ventral-filed. dorsal-filed and dorsal-and-ventraL
`
`filed layers. The diffusion coefficients of each drug
`in the_ ventral-filed, dorsal-filed. dorsal-and-ventrah
`filed and full-thickness nail layers were calculated
`using Equation 3. The diffusion coefficients of 5-
`fluorouracil and fiurbiprofen in the ventral-filed
`layer were lower than those in the dorsal-filed and
`dorsal-and-ventral-filed layers.
`
`The permeation parameters of each single layer
`(dorsal,
`intermediate or ventral) were computed
`using those of the ventral-filed, dorsal-filed and
`dorsal-and-ventral-filed layers. Flynn et al (1974)
`proposed that the diffusional resistance. R,.
`in the
`ith layer can be defined by:
`
`R=1/P;=h;/(D;K.)
`
`(4)
`
`In multiple layers. the total diffusional resistance
`(RT) may be computed by:
`
`R1" = 1/Pr = hi/(DIKI) + hz/lD2K:)
`
`(5)
`
`+...'l-ll"/(DnKn')
`
`Table 3 shows the permeation parameters of 5-
`fluorouracil and flurbiprofen for each single nail
`layer. The pemieability coefficient of each single
`layer was calculated using Equation 5. For 5-
`fiuorouracil,
`the intermediate layer had a high
`penneability coefficient compared with the other
`single layers.
`In contrast, the ventral
`layer had a
`high permeability coefficient for flurbiprofen. The
`permeability coefficients for 5-fluorouracil and
`flurbiprofen through the dorsal layer were low. The
`nail plate/vehicle partition coefficients of the drugs
`for each single layer were calculated using those
`of the ventral-filed. dorsal-filed and dorsal-and-
`
`Page 7 of 10
`
`

`

`276
`
`
`Table 2.
`
`
`
`Permeation parameters of 5-fluorouracil and fiurbiprofen through the human nail plate.
`
`
`
`
`
`
`
`
`
`
`
`
`
`YOICHI KOBAYASHI ET AL
`
`
`
`
`
`
`
`Drug Parameter
`
`
`
`Layer (thickness ratio)
`
`
`
`
`Ventral—filed {3 :5}
`
`
`
`
`Dorsal- filed {S : 2}
`
`
`
`
`Dorsal—and-ventral-filed (5)
`
`
`
`Full—thickness (3 : 5 1 2]
`
`
`
`
`
`
`
`5—Fluorouracil
`h (pm)
`
`
`P (X lU?cms ’ ')
`
`
`
`Km
`1
`
`
`Din {X lO8cm's" ')
`
`
`
`
`Flurbiprofen
`h turn)
`
`
`
`P (xl{l°cms"1'}
`
`
`Km
`3
`
`
`
`
`Elm (x l0"crn‘s‘ ')
`
`
`396.0
`
`2-5! $048
`
`0-53 :1: 0-04
`
`1-87
`
`
`34?-2
`
`1-83 d: 0-] l
`
`l-07:t 0-15
`
`
`5-92
`
`
`346-5
`
`3-07:l:0-22
`
`0-49:l:U-OI
`
`2-18
`
`
`303-8
`
`2-76:l:0-43
`l«l3:t0«24
`742
`
`
`
`
`
`2475
`
`7- l 2 :l: 0-65
`0-56 :I: 0-02
`3- l3
`
`
`
`
`
`217-0
`
`3-59:t 1-28
`
`
`0-98:t0«22
`
`796
`
`
`495-l}
`
`1-49i{l-2'?
`
`O-S4-4.-.0-06
`
`1-3?
`
`
`434-0
`
`l-45 i 0-54
`
`l-47 i 032
`
`4-29
`
`
`
`Each value represents the mean :E s.e.m. (n = 3—-5). Apparent penrteability coefficients were calculated using Equations 2 and 3.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`h is the nail membrane thickness; P is the permeability coefficient: Km is the membranefdonor vehicle partition coefficient of the
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`drug; Dm is the diffusion coefficient of the drug in the nail membrane.
`
`
`
`
`
`
`
`
`
`
`
`ventral-filed nail layers and the weight ratio of each
`
`
`
`
`
`
`
`layer {3:5:2). The partition coefficient for 5-
`
`
`
`
`
`
`
`
`fluorouracil
`in the ventral
`layer was lower than
`
`
`
`
`
`
`
`
`those in the dorsal and intermediate layers. The
`
`
`
`
`
`
`
`rank order of the partition coefficients for S-fiuoro-
`
`
`
`
`
`
`
`uracil was
`interrnediate layer :» dorsal
`layer >
`
`
`
`
`
`
`
`
`ventral layer, whereas the rank order of the flurbi—
`
`
`
`
`
`
`
`
`profen partition coefficients was the reverse of this.
`
`
`
`
`
`
`
`In addition, rank orders of these partition coeffi-
`
`
`
`
`
`
`
`
`
`
`cients for each drug are correlated with those of the
`
`
`
`
`
`
`
`
`
`total lipid concentrations in the three layers of the
`
`
`
`
`
`
`
`
`human nail plate. The diffusion coefficients of the
`
`
`
`
`
`
`
`
`drugs in each single layer was calculated using
`
`
`
`
`
`
`
`
`Equation 3. The dorsal
`layer had low diffusion
`
`
`
`
`
`coefficients
`for 5-fluorouracil
`and fiurbiprofen
`
`
`
`
`
`
`
`compared with the other single layers. The diffu—
`
`
`
`
`
`
`sion coefficients for 5-fluorouracil and fiurbiprofen
`
`
`
`
`
`
`
`
`
`in the intermediate layer were the highest of those
`
`
`
`
`
`
`
`
`for any single layer. The permeation parameters of
`
`
`
`
`
`
`
`
`
`the full—thickness nail plate (as shown in Table 3)
`
`
`
`
`
`
`
`
`
`
`
`
`
`were calculated by substituting those of each single
`
`
`
`
`
`
`
`layer, according to Equation 5. They agreed with
`
`
`
`
`
`
`the experimental permeation parameters of the full-
`
`
`
`
`
`
`thickness nail plates obtained from the permeation
`
`
`
`
`
`
`
`
`
`studies for both drugs (as shown in Table 2).
`
`
`
`
`
`
`From these results,
`the drug permeation char-
`
`
`
`
`
`
`
`
`acteristics of each single layer can be summarized
`
`
`
`
`
`
`
`
`
`as follows:
`the dorsal layer is characterized by a
`
`
`
`
`
`
`
`
`low diffusivity of drugs; the intermediate layer is
`
`
`
`
`
`
`
`characterized by low lipophilicity; and the ventral
`
`
`
`
`
`
`
`layer is characterized by high lipophilicity. We
`
`
`
`
`
`
`
`
`suggest that the main nail barrier to drug pemiea-
`
`
`
`
`
`
`
`
`
`
`tion may be the low diffusivity of drugs in the
`
`
`
`
`
`
`
`
`dorsal layer. The difference between the nail plate/
`
`
`
`
`
`
`vehicle partition coefficients for 5-fluorouracil and
`
`
`
`
`
`
`
`
`flurbiprofen was small because of the low lipid
`
`
`
`
`
`
`
`
`in the human nail plate. Therefore.
`this
`content
`
`
`
`
`
`
`
`
`
`suggests that the human nail plate behaves like a
`
`
`
`
`
`
`
`hydrophilic gel membrane rather than a lipophilic
`
`
`
`
`
`
`partition membrane. This suggestion agrees with
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Table 3. Calculated permeation parameters of 5—fluorouracil and flurbiprofen through the human nail plate.
`
`Drug Parameter
`
`
`
`
`Layer (thickness ratio)
`
`
`
`
`Intermediate
`(5)
`
`
`
`
`247-5
`
`T-12
`
`056
`
`3-13
`
`
`217-0
`
`3-59
`
`0-93
`
`7-96
`
`
`Ventral
`{3}
`
`
`
`
`99-0
`
`5-39
`
`0-30
`
`180
`
`
`86-8
`
`1 1-90
`
`1 -5 1
`
`3-42
`
`
`Full—thi-ckness
`(3 :5 : 2)
`
`
`
`
`
`495-0
`
`l-Tl
`
`0-48
`
`I-75
`
`
`434-0
`
`1-58
`
`1- I6
`
`5-9]
`
`
`Dorsal
`( 3)
`
`
`
`
`148-5
`
`3-37
`0-48
`l-2|
`
`
`
`I302
`
`3-21
`1-22
`2-63
`
`
`
`5-Fluorouracil
`h (pm)
`
`
`
`
`P (x 10? cms" '}
`
`Km
`Dm (x1o“cm2.-:' ‘i
`
`Flurbiprofen
`
`h (um)
`
`
`
`P (xIo“cms‘ ')
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Equation
`Each value represents the mean i s.e.m. (n : 3~5}.Perrneability coefficients of each single layer were calculated using
`
`5. h is the nail membrane thickness: P is the permeability coeffieient: Km is the membraneldonor vehicle partition coefficient of the
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`drug; Drn is the diffusion coefficient of the drug in the nail membrane.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Page 8 of 10
`
`Page 8 of 10
`
`

`

`DRUG PERMEATIDN THROUGH HUMAN NAIL PLATE
`
`
`
`
`
`
`
`
`2??
`
`
`
`
`
`
`
`
`
`
`
`where [CA — CA}, (Cg - CB) and (Cc — Clcl are
`
`
`
`
`
`
`
`
`the concentration differentials of that drug in each
`
`
`
`
`
`
`
`layer {A =dorsal; B‘: intermediate; C : ventral).
`
`
`
`
`
`
`
`
`
`
`
`In a sink condition. C C can be assumed to be equal
`
`
`
`
`
`
`
`
`
`to zero. Therefore, the total amount of a drug (QT)
`
`
`
`
`
`
`
`
`
`in the full-thickness nail plate can be defined as
`follows:
`
`
`
`
`
`
`(7)
`
`
`
`
`
`
`QT/S ‘—‘ (CA ‘ Cirslhn/2 + (CH ‘ Cialha/2
`
`
`+ Cchc/2
`
`
`
`
`
`
`
`
`
`
`where S is the surface area of the nail plate. We
`
`
`
`
`
`
`calculated the mean drug concentration in each
`
`
`
`
`
`
`
`single layer using Equations 6 and 7.
`Table 4 shows the amounts and mean con-
`
`
`
`
`
`
`
`
`
`
`
`
`centrations of 5-fluorouracil and flurbiprofen in
`
`
`
`
`
`
`each layer,
`as estimated from the permeation
`
`
`
`
`
`parameters. The estimated amount of 5-fluorouracil
`
`
`
`
`
`
`
`
`in the full-thickness nail plate was similar to the
`
`
`
`
`
`experimental data. The 5-fluorouracil and flurbi-
`
`
`
`
`
`
`
`profen concentrations in the ventral nail layer that
`
`
`
`
`
`
`
`resulted from drug permeation data from the ven-
`
`
`
`
`
`
`
`
`
`tral nail plate were 4-2 and 45 times higher than
`
`
`
`
`
`
`
`those from the dorsal plate. Compared with the
`estimated 5-fluorouracil con