`
`www.elsevier.com/locate/ijpharm
`
`Review
`Drug delivery to the nail following topical application
`
`Sudaxshina Murdan *
`Department of Pharmaceutics, The School of Pharmacy, Uni6ersity of London, 29-39 Brunswick Square, London WC1N 1AX, UK
`
`Received 31 August 2001; received in revised form 10 December 2001; accepted 13 December 2001
`
`Abstract
`
`The absorption of drugs into the nail unit, following topical application to the nail plate, is highly desirable to treat
`nail disorders, such as onychomycosis (fungal infections of the nail). Nail permeability is however quite low and limits
`topical therapy to early/mild disease states. In this paper, the recent research into ungual drug delivery is reviewed.
`The nail unit and the two most common diseases affecting the nail—onychomycosis and nail psoriasis—are briefly
`described to set the scene and to give an overview of the nature and scope of the problem. The factors, which affect
`drug uptake and permeation through the nail plate such as solute molecular size, hydrophilicity/hydrophobicity,
`charge, and the nature of the vehicle, are then discussed, followed by ways of enhancing drug transport into and
`through the nail plate. Finally, drug-containing nail lacquers which, like cosmetic varnish, are brushed onto the nail
`plates to form a film, and from which drug is released and penetrates into the nail, are reviewed. © 2002 Elsevier
`Science B.V. All rights reserved.
`
`Keywords: Ungual drug delivery; Nail; Nail lacquers; Topical application
`
`1. Introduction
`
`to claws and
`The human nail, equivalent
`hooves in other mammals, evolved as our manual
`skills developed and protects the delicate tips of
`fingers and toes against trauma, enhances the
`sensation of fine touch and allows one to pick up
`and manipulate objects. The nail is also used for
`scratching and grooming, as a cosmetic organ and
`sometimes, to communicate one’s social status
`(Barron, 1970; Dawber and Baran, 1984; Chap-
`man, 1986; Gonzalez-Serva, 1997). The nail plate
`is the most visible part of the nail apparatus,
`
`* Tel.: +44-20-7753-5810; fax: +44-20-7753-5942.
`E-mail address: sudax.murdan@ulsop.ac.uk (S. Murdan).
`
`consists of tightly packed dead cells and is highly
`keratinised. It is also very variable among individ-
`uals. The plates can be small, large, wide, narrow,
`hard, smooth, ridged, thin, etc.
`Disorders of the nail unit range from relatively
`innocuous conditions such as pigmentation in
`heavy smokers, to painful and debilitating states
`where the nail unit can be dystrophied, hypertro-
`phied, inflamed, infected etc. Such conditions af-
`fect patients physically as well as socially and
`psychologically and can seriously affect the qual-
`ity of life. Many nail diseases are notoriously
`difficult to cure, need a long duration of treatment
`and relapse is common. Oral therapy has the
`inherent disadvantages of systemic adverse effects
`and drug interactions while topical therapy is
`limited by the low permeability of the nail plates.
`
`0378-5173/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved.
`PII: S 0 3 7 8 - 5 1 7 3 ( 0 1 ) 0 0 9 8 9 - 9
`
` 1
`
`CFAD Exhibit 1028
`
`1
`
`
`
`2
`
`S. Murdan /International Journal of Pharmaceutics 236 (2002) 1–26
`
`Topical therapy is an attractive option how-
`ever, due to its non-invasiveness, drug targeting to
`the site of action, elimination of systemic adverse
`events and drug interactions,
`increased patient
`compliance and possibly reduced cost of treat-
`ment. Topical therapy can be optimised by the use
`of: (i) potent drugs to ensure that effective drug
`concentrations are achieved at the site of action;
`(ii) drugs with the correct physico-chemical prop-
`erties for permeation into the nail plate; (iii) pene-
`tration enhancers
`to facilitate ungual drug
`permeation; and by (iv) appropriate formulations
`which aid ungual drug uptake, are easy to use,
`and which stay in contact with nail plates, releas-
`ing drugs continuously over long periods of time.
`In their review on the topical delivery of anti-
`fungal drugs for onychomycosis treatment, Sun et
`al. (1999), have concluded that topical treatment
`of onychomycosis remains a drug delivery prob-
`lem. In this paper, the research into drug delivery
`to the nail unit following topical application is
`reviewed, in an attempt to establish how topical
`therapy can be optimised. The nail unit and its
`most common disease states are briefly described
`to set the scene and understand the nature of the
`problem. Nail permeability to drugs is then dis-
`cussed, with respect to factors that influence drug
`permeation and ways of enhancing ungual drug
`
`penetration, including the novel delivery vehicles,
`drug-containing nail
`lacquers. Finally, conclu-
`sions are drawn on how one can optimise topical
`drug delivery to the nail unit.
`
`2. The nail unit
`
`The nail apparatus, schematically shown in Fig.
`1, is composed of the nail folds, nail matrix, nail
`bed and the hyponychium, which together form
`the nail plate (Zaias, 1990). The nail plate, pro-
`duced mainly by the matrix, emerges via the
`proximal nail fold and is held in place by the
`lateral nail folds. It overlays the nail bed and
`detaches from the latter at the hyponychium (skin
`under the free edge of the plate). The nail plate is
`a thin (0.25–0.6 mm), hard, yet slightly elastic,
`translucent, convex structure and is made up of
`approximately 25 layers of dead, keratinised, flat-
`tened cells which are tightly bound to one another
`via numerous intercellular links, membrane-coat-
`ing granules and desmosomes. The cells at the
`dorsal surface of the plate overlap (Fig. 2a) and
`produce a smooth surface. In contrast, the palmar
`surface of the nail plate is quite irregular (Fig.
`2b). Fig. 2c shows a cross-section view of the nail
`plate. The latter can be divided into three macro-
`
`Fig. 1. Schematic structure of the nail apparatus. Reproduced from Ref. De Berker et al. (1995a), with kind permission from
`Blackwell Science Ltd.
`
`2
`
`
`
`S. Murdan /International Journal of Pharmaceutics 236 (2002) 1–26
`
`3
`
`Fig. 2. Scanning electron micrograph of: (a) the dorsal surface of human nail plate; (b) the nail undersurface; and (c) a cross-section
`view of the nail plate.
`
`intermediate and ventral.
`scopic strata—dorsal,
`The dorsal layer is a few cells thick, while the
`intermediate plate is a softer, more flexible,
`thicker layer and accounts for the majority of the
`nail thickness. The ventral
`layer is very thin,
`consists of a few layers of cells and connects the
`nail plate to the nail bed. Kobayashi et al. (1999),
`calculated that the thickness ratio of each layer
`i.e. dorsal:intermediate:ventral is 3:5:2.
`Chemically, the nail plate consists mainly of the
`fibrous proteins, keratins, 80% of which is of the
`‘hard’ hair-type keratin, the remainder comprising
`the ‘soft’ skin-type keratin (Lynch et al., 1986).
`
`The keratin fibres are oriented into three layers,
`which are associated with the dorsal, intermediate
`and ventral nail
`layers. The hair-like keratin
`filaments are only present in the intermediate nail
`layer and are oriented perpendicular
`to the
`growth axis, while the skin-type keratin filaments
`are found in the dorsal and the ventral layers and
`are oriented in two privileged directions, trans-
`verse and perpendicular to the growth axis (Gar-
`son et al., 2000). The keratin fibres are thought to
`be held together by globular, cystine-rich proteins
`whose disulphide links act as glue (Fleckman,
`1997). The plate also contains water at 10–30%,
`
`3
`
`
`
`S. Murdan /International Journal of Pharmaceutics 236 (2002) 1–26
`
`growing in length, nail plates also grow thicker as
`they progress from the lunula to the free margin,
`and as ventral nail layers are added to the grow-
`ing plate by the nail bed. This mechanism is
`thought to contribute to approximately 20% of
`the final nail mass. The thickening rate is slow
`and a mean value of 0.027 mm/mm nail length
`has been reported (Johnson and Shuster, 1993).
`The nail plate adheres closely to and overlays
`the nail bed—a thin, soft, non-cornified epithe-
`lium, which extends from the lunula to the hy-
`ponychium. The nail bed acts as a holder and
`slide for the growing nail plate, as well as con-
`tributing to the growth of the nail plate as men-
`tioned above. The nail bed, nail matrix and the
`tissues around the nail are well perfused by blood
`vessels (Flint, 1955; Hale and Burch, 1960; Sam-
`man, 1959). In addition, the nail bed has a rich
`supply of
`lymphatic vessels
`(Pardo-Castello,
`1960).
`
`3. Diseases affecting the nail and their treatment
`
`Nails can suffer from a very wide range of
`disorders. For example, nails can be discoloured
`(e.g. by certain systemic drugs), rendered brittle
`(e.g. by chronic use of detergents), chronic trauma
`to toenails from ill-fitting shoes can lead to in-
`growing nails, plates can thicken, be infected, lift
`off the nail bed, etc. Disorders of the nail may
`also reflect systemic diseases and may provide
`useful diagnostic clues. The two most common
`diseases affecting the nail unit are onychomycosis
`infections of the nail plate and/or nail
`(fungal
`bed) and psoriasis of the nails. In this review,
`these two disease states are briefly described for
`their high occurrence rate and for the fact that
`most of the research conducted into topical drug
`treatment of diseased nails has been focussed on
`these two conditions.
`
`3.1. Onychomycosis
`
`Onychomycosis, responsible for up to 50% of
`nail disorders (Ghannoum et al., 2000) is a very
`common problem, affecting 3–10% of the popula-
`tion in Europe, prevalence being higher in older
`
`4 w
`
`ater content is directly related to the relative
`humidity and is important for nail elasticity and
`flexibility (Forslind, 1970; Baden et al., 1973). In
`contrast, the nail plate contains small amounts of
`lipid, between 0.1 and 1.0%, most of which is
`organised into bilayers oriented parallel to the
`nail surface and is concentrated in the ventral and
`dorsal nail layers (Walters and Flynn, 1983; Gni-
`adecka et al., 1998; Kobayashi et al., 1999; Gar-
`son et al., 2000).
`The nail plate is a fairly strong structure. Its
`hardness and mechanical rigidity is thought to be
`due to the sandwich orientation of the keratin
`fibres,
`the presence of globular proteins that
`provide the ‘glue’ to hold keratin fibres together,
`adhesiveness of nail cells to one another, physical
`and chemical stability of the nail proteins (con-
`ferred by the stable disulphide links), the design of
`the plate (which is curved in both transverse and
`longitudinal axes) and its water content.
`The nail plate is formed by the nail matrix
`which is a highly proliferative epidermal tissue. It
`is also called the root of the nail, lies underneath
`the proximal nail fold and its distal portion is
`often visible through the transparent nail plate as
`a white, semilunar area, called the lunula. Cell
`division of the matrix results in the continuous
`formation of the nail plate, which grows through-
`out life. Growth rate is highly variable among
`individuals; average values of 3 mm per month
`(fingernails) and 1 mm per month (toenails) are
`used when treating nails. A normal fingernail
`grows out completely in about 6 months while a
`normal toenail
`in about 12–18 months (Fleck-
`man, 1997). Nail growth rate is also highly influ-
`enced by age (ageing slows the rate), gender (rate
`is higher in males), climate (slower in cold cli-
`mate), dominant hand (growth is faster), preg-
`nancy (faster), minor trauma/nail biting (increases
`growth rate), diseases (can increase or decrease
`rate e.g. growth is faster in patients suffering from
`psoriasis and slower in persons with fever), mal-
`nutrition (slower rate) and drug intake (may in-
`crease or decrease)
`(Hamilton et al., 1955;
`LeGros, et al., 1938; Bean, 1980; Geoghegan et
`al., 1958; Hewitt and Hillman, 1966; Gilchrist and
`Dudley Buxton, 1938–39; Landherr et al., 1982;
`Sibinga, 1959; Dawber et al., 1994). As well as
`
`4
`
`
`
`S. Murdan /International Journal of Pharmaceutics 236 (2002) 1–26
`
`5
`
`people (Roberts, 1999; Chabasse et al., 2000; Pier-
`ard, 2001). Occurrence seems to be on the increase
`due to a growing elderly population, the spread of
`HIV infection and AIDS, a higher frequency of
`iatrogenic immunosuppression due to the use of
`immunosuppressant drugs,
`lifestyle factors such
`as the wearing of tight-fitting clothing and shoes
`and the use of communal recreational facilities
`and healthclubs, as well as improved detection
`and higher public awareness (Gupta and Shear,
`1997; Daniel, 1991; Scher, 1996; Cohen and Scher,
`
`1994). Most (90–95%) of the infections are caused
`by dermatophytes, the rest being caused by yeasts
`and moulds. Toenails are affected more than
`fingernails (Midgley et al., 1994). Toenail ony-
`chomycoses are also more recalcitrant and have to
`be treated for longer durations.
`Clinically, onychomycosis can be divided into
`categories depending on where the infection
`begins:
`(i) Distal and lateral subungual onychomycosis
`(Fig. 3a): The fungal infection starts at the
`
`Fig. 3. Onychomycosis: (a) distal and lateral subungual onychomycosis; (b) superficial onychomycosis manifested as white spots; (c)
`total dystrophic onychomycosis. Reproduced from Ref. De Berker et al. (1995b), with kind permission from Blackwell Science Ltd.
`
`5
`
`
`
`6
`
`S. Murdan /International Journal of Pharmaceutics 236 (2002) 1–26
`
`hyponychium and the distal or lateral nail
`bed. The fungus then invades the proximal
`nail bed and ventral nail plate.
`(ii) Superficial white onychomycosis (Fig. 3b):
`The nail plate is invaded directly by the
`causative
`organism and white
`chalky
`patches appear on the plate. The patches
`may coalesce to cover
`the whole plate
`whose surface may crumble.
`(iii) Proximal
`subungual onychomycosis: The
`fungus invades via the proximal nail fold
`and penetrates the newly formed nail plate,
`producing a white discoloration in the area
`of the lunula.
`(iv) Total dystrophic onychomycosis (Fig. 3c):
`This is the potential endpoint of all forms
`of onychomycosis and the entire nail plate
`and bed are
`invaded by
`the
`fungus
`(Elewski et al., 1997; De Berker et al.,
`1995b).
`In the past, diseased nails were surgically ex-
`tracted and disease-free nails allowed to grow
`back. Surgical avulsion is, however, extremely
`traumatic and has largely been consigned to his-
`tory (Niewerth and Korting, 1999). Nowadays,
`diseased nails may be chemically removed using
`urea ointment. A urea formulation (containing
`as much as 40% urea) is applied onto the dis-
`eased nail plates under occlusive dressings. Urea
`softens the plate and, after 5–10 days, the entire
`nail plate may be lifted off the nail bed and
`trimmed behind the proximal nail fold (Farber
`and South, 1978). A disease-free nail may then
`grow back. Onychomycosis may also be treated
`systemically with new potent antifungal agents,
`such as terbinafine and itraconazole. Following
`oral administration and absorption into the sys-
`temic circulation,
`the drugs diffuse from the
`blood vessels into the nail plate via the nail bed.
`Itraconazole is administered either continuously
`(200 mg daily for 3 months) or as pulse therapy
`(400 mg/day for 7 days, subsequent courses—2
`or 3—are repeated after 21-day intervals). Ter-
`binafine is dosed at 250 mg daily for 6–12
`weeks.
`Unfortunately, a significant number—around
`20% of patients—do not respond to treatment
`(Roberts, 1999). Relapse is also common. Tosti
`
`et al. (1998), reported that 22.2% of patients
`whose toenail onychomycoses had been cured by
`oral terbinafine or itraconazole experienced re-
`lapse during a 3-year follow-up study. Systemic
`therapy also has inherent disadvantages such as
`adverse events and drug interactions. For exam-
`ple, hepatic function tests are recommended for
`patients who use terbinafine continuously for
`more than 6 weeks. Additional blood counts are
`recommended for patients with known or sus-
`pected immunodeficiency. Itraconazole has been
`associated with liver damage; liver function tests
`are required if continuous treatment exceeds 1
`month.
`Topical therapy avoids the problems associ-
`ated with systemic treatment. However, drug
`diffusion into the highly keratinised nail plate is
`poor, duration of
`treatment
`is long and cur-
`rently, topical therapy is only recommended for
`the early stages of the disease and when up to
`two nails are affected or when systemic therapy
`is contra-indicated. The most convenient topical
`preparations are the nail lacquers (nail varnish)
`containing the antifungal agents amorolfine (Lo-
`ceryl®) and ciclopirox (Penlac®, Loprox® in
`Canada). Like cosmetic nail varnish, these drug-
`containing lacquers are applied to nail plates
`with a brush and dry within a few minutes to
`leave a water-insoluble film. Drug is then re-
`leased from the film and permeates into the nail
`plate. Loceryl® is applied 1–2 times weekly to
`filed nail plates for up to 6 months (fingernails)
`and for 9–12 months for toenails. Penlac® is
`applied once daily, preferably at bedtime, for up
`to 48 weeks. Every 7 days the Penlac® film is
`removed with alcohol before re-application of
`the lacquer. The nail lacquers, novel drug deliv-
`ery formulations, are discussed in more detail in
`Section 5. Other topical antifungal formulations
`for onychomycoses include tioconazole nail solu-
`tion (Trosyl®), an undecenoate solution (Mon-
`phytol®) and salicylic acid paint
`(Phytex®).
`Trosyl® is applied to infected nails and sur-
`rounding skin twice daily for up to 6 months.
`This may be extended to 12 months. Phytex®
`and Monphytol® are also applied twice daily,
`but are not the first line of treatment and are
`only used in certain circumstances.
`
`6
`
`
`
`S. Murdan /International Journal of Pharmaceutics 236 (2002) 1–26
`
`7
`
`3.2. Psoriasis
`
`an inflammatory disease of
`is
`Psoriasis
`the skin and is characterised by epidermal thick-
`ening and scaling as a result of excessive cell
`division in the basal layers. It affects between 1
`and 3% of most populations, but, is most com-
`mon in Europe and North America (Schofield and
`Hunter, 1999). It is thought that 80% of patients
`with skin psoriasis also suffer from psoriasis of
`the nail (De Jong et al., 1996) while 1–5% of
`patients with nail psoriasis do not present
`any overt cutaneous disease (Del Rosso et al.,
`1997).
`folds
`The nail matrix, nail bed and nail
`may all be affected. The psoriatic nail matrix
`results
`in pitting (presence of
`small
`shallow
`holes in the nail plate), nail fragility, crumbling or
`nail loss while nail bed involvement causes ony-
`cholysis (separation of the nail plate from the nail
`bed, which may be
`focal or distal),
`sub-
`ungual hyperkeratosis and splinter haemorrhages.
`Psoriatic nail
`folds
`result
`in paronychia (in-
`flamed and swollen nail folds) which leads to
`ridging of the nail plate. When paronychia is
`severe, the matrix may be injured with consequent
`nail abnormalities
`(Del Rosso et al., 1997).
`Fig. 4a–d show some examples of psoriatic condi-
`tions.
`Like onychomycosis, nail psoriasis is a long-
`term condition,
`is difficult
`to cure,
`relapse
`is common and therapy has to be maintained for
`very long durations. Injection of corticosteroids
`into the nail folds is the mainstay of therapy.
`Such injections are extremely painful and need to
`be repeated monthly for a total of 4–6 times.
`Topical therapy consists of debridement of the
`nail plate
`followed by
`the
`application of
`drugs
`such
`as
`steroids
`and
`5-fluorouracil
`(5-FU). These drugs must permeate through the
`nail plate
`to reach the nail bed and the
`matrix
`target
`sites. Oral
`agents
`such as,
`methotrexate, etretinate, cyclosporine, have been
`beneficial in some cases (De Jong et al., 1999).
`The diseased nails may also be non-surgically
`extracted using urea as described earlier, followed
`by the application of topical medication to the
`nail bed.
`
`4. Perungual drug absorption following topical
`application
`
`At first glance, the highly keratinised, compact
`nail plate appears pretty impermeable. There is a
`significant body of evidence, however, for nail
`permeability and some of the direct and indirect
`evidence has been discussed by Walters and Flynn
`(1983). Absorption of water by the nail plate and
`the subsequent plate softening is well known.
`Diffusion of topically applied urea into nails,
`resulting in the separation of the nail plate from
`the nail bed was mentioned earlier. The topical
`treatment of certain nail disorders such as early
`stage onychomycoses attest to the permeability of
`nail plates. Toxicity, e.g. loosening of the plate
`from the nail bed, arising from the absorption of
`noxious chemicals present in early nail cosmetics
`has also been reported (Sulzberger et al., 1948),
`and confirm a certain permeability of nail plates.
`It must be stressed, however, that nail permeabil-
`ity is generally poor and drug flux through the
`nail plate is low.
`In the last 20 years, nail permeability has been
`systematically characterised. Permeation studies
`using modified Franz diffusion cells, measurement
`of nail swelling and drug uptake into nails when
`the latter are soaked in drug formulations have
`been the most common in vitro tests. Avulsed
`human cadaver nail plates, nail clippings from
`healthy volunteers and bovine hoof membranes
`have been used as sources and model of the nail
`plate. Bovine hooves have been used, as they are
`easier to obtain than human nail. Hooves are
`taken from freshly slaughtered cattle, the adhering
`non-hoof tissues are removed, the hooves are
`soaked in water, following which membranes (e.g.
`100 mm thick) are sectioned using a microtome
`(Mertin and Lippold, 1997b). Mertin and Lippold
`(1997a),
`reported that permeability coefficient
`through human nail plate could be predicted if
`permeability coefficient into bovine hoof mem-
`brane was known, using the following equation:
`log PN=3.723+1.751 log PH,
`where PN is the permeability coefficient through
`the nail plate and PH is the permeability coeffi-
`cient through the hoof membrane. Permeability
`
`7
`
`
`
`8
`
`S. Murdan /International Journal of Pharmaceutics 236 (2002) 1–26
`
`Fig. 4. Nail psoriasis: (a) pitting; (b) ‘oil drop’ formation, a manifestation of focal onycholysis; (c) nail plate crumbling from matrix
`involvement; (d) hyperkeratosis. Fig. 4a and b have been reproduced from Ref. De Berker et al. (1995c), with kind permission of
`Blackwell Science Ltd. Fig. 4c and d have been reproduced from Ref. Del Rosso et al. (1997), with kind permission from WB
`Saunders Company.
`
`8
`
`
`
`S. Murdan /International Journal of Pharmaceutics 236 (2002) 1–26
`
`9
`
`coefficient is defined as the product of the drug’s
`diffusion coefficient (D) through the nail/hoof
`barrier and the drug’s partition coefficient (K)
`between the vehicle and the barrier, P=DK (cm2/
`s).
`
`Care must be exercised however, when using
`the hoof as a model for the human nail plate. The
`hoof is much more permeable than the human
`nail plate, the hoof keratin network is thought to
`be less dense and when incubated in water, the
`hoof swells to a larger extent—36 versus 27%—
`compared to human nails (Mertin and Lippold,
`1997a). Hoof proteins have a significantly lower
`content of half-cystine and disulphide linkages
`compared to the human nail plate (Baden et al.,
`1973; Marshall and Gillespie, 1977). As a result,
`the hoof may be less susceptible to compounds,
`which break the disulphide linkages and which are
`being investigated as potential perungual penetra-
`tion enhancers. In such cases, enhancement of
`perungual absorption in the hoof may be less than
`the enhancement that could be achieved in human
`nail plates.
`Following topical application of a drug formu-
`lation onto the nail plate, the drug has to enter
`the nail plate and diffuse into the deeper nail
`layers and possibly into the nail bed. Walters et
`al. (1983), found that the nail plate behaves like a
`concentrated hydrogel rather than a lipophilic
`membrane (unlike other body membranes such as
`skin and gastro-intestinal tract mucosa). Subse-
`quently, Mertin and Lippold (1997a), with refer-
`ence to Lieb and Stein (1969), compared the
`diffusion of solute molecules through the nail
`plate ( hydrogel) to the diffusion of non-elec-
`trolytes through polymers and suggested that the
`thermal movement of the keratin fibres in the nail
`plate would create holes (‘pores’) which would
`then be occupied by the diffusing molecules.
`From the hydrogel literature, we can expect small
`solute molecules to diffuse at a faster rate com-
`pared to larger molecules. When solute size is very
`small compared to pore size, drug permeation
`may occur via convection as well as via diffusion.
`The geometry (linear, globular, branched, asym-
`metry) of the diffusing molecules will also have an
`effect on the movement of
`these molecules
`through the pores of the hydrogel (nail plate). In
`
`addition to the pore mechanism, solute transport
`through nails ( hydrogels) may occur through
`the partition mechanism. In this case, the solute
`would partition into the keratin polymer network
`and diffuse along the polymer segments. The par-
`tition coefficient would be influenced by interac-
`tions,
`such
`as
`hydrophobic
`and
`electrical
`interactions, between the network (keratin fibres)
`and the solute molecule.
`Drug transport into the nail plate is thus ex-
`pected to be influenced by the physico-chemical
`properties of the drug molecule (e.g. size, shape,
`charge, hydrophobicity), the formulation charac-
`teristics (e.g. nature of vehicle, pH, drug concen-
`tration),
`the
`presence
`of
`any
`penetration
`enhancers, nail properties (e.g. thickness, hydra-
`tion), as well as interactions between the permeat-
`ing molecule and the keratin network of the nail
`plate. Section 4.1 describes how drug transport
`through the nail plate, measured as permeability
`coefficient, is influenced by various factors. Per-
`meability coefficient has been defined by a num-
`ber of authors as the product of
`the drug’s
`diffusion coefficient (D) through the nail/hoof
`barrier and the drug’s partition coefficient (K)
`between the vehicle and the barrier, P=DK (cm2/
`s). Other authors have used the more conven-
`tional definition of permeability coefficient (also
`called permeability) which is the product of D and
`K, divided by the nail thickness h, P=DK/h
`(cm/s). In this review, permeability coefficient is
`used as the authors of the individual papers have
`intended while the unit of P (cm2/s or cm/s)
`indicates which definition is being used.
`
`4.1. Factors which influence drug transport into
`and through the nail plate
`
`4.1.1. Molecular size of diffusing molecule
`As expected, molecular size has an inverse rela-
`tionship with penetration into the nail plate. The
`larger the molecular size, the harder it is for
`molecules to diffuse through the keratin network
`and lower the drug permeation. Mertin and Lip-
`pold (1997a), demonstrated the decreasing perme-
`ability coefficients through human nail plate and
`through bovine hoof membrane with increasing
`molecular size of a series of alkyl nicotinates (Fig.
`
`9
`
`
`
`10
`
`S. Murdan /International Journal of Pharmaceutics 236 (2002) 1–26
`
`eight carbon atoms resulted in a decrease in per-
`meability coefficient, after which, increasing chain
`length (to C12) resulted in increased permeability
`coefficient (Fig. 6a). Increasing lipophilicity of the
`diffusing alcohol molecule reduces the permeabil-
`ity coefficient until a certain point after which
`further increase in lipophilicity results in increased
`permeation. The nail plate seems to be a hy-
`drophilic structure when the permeation of the
`lower alcohols (BC8) is considered. The authors
`concluded that the nail plate behaves like a con-
`centrated hydrogel. The permeation of neat alco-
`hols follows a similar trend as shown in Fig. 6b
`(Walters et al., 1985a). However, except
`for
`methanol, the permeability coefficient of neat al-
`cohols ( absence of water) was approximately
`five times smaller than the permeability coefficient
`of diluted alcohols. The authors suggest that this
`indicate a facilitating role of water towards the
`diffusion of the alcohol molecules. It is possible
`that when an aqueous formulation is used, nails
`swell as water is taken up into the nail plates.
`Consequently,
`the keratin network expands,
`which leads to the formation of
`larger pores
`through which diffusing molecules can permeate
`more easily.
`The increase in permeation of the higher alco-
`hols (C10 and C12) with increasing lipophilicity
`was suggested to occur through a lipidic pathway.
`Despite the low content of lipid (up to 1% of the
`total weight) in the nail plate, this lipid pathway
`seems to be important for the passage of very
`hydrophobic substances. Indeed, extraction of the
`nail lipid by incubating the nail plates in chloro-
`form/methanol mixture for 24 h reduced the per-
`meation of decanol and dodecanol even though
`the permeation of water, methanol, ethanol and
`butanol were increased (Table 1). The authors
`suggest that a minor lipid pathway exists in the
`nail plate, and becomes the rate-controlling bar-
`rier for hydrophobic molecules like decanol and
`dodecanol.
`The relationship between drug permeation and
`hydrophilicity/lipophilicity is however not
`so
`straightforward. Mertin and Lippold (1997b),
`found that the permeability coefficients of a num-
`ber of alkyl nicotinates across bovine hoof mem-
`brane did not change with increasing lipophilicity
`
`Fig. 5. Relationship between log of permeability coefficient (P)
`and the molecular weight of a range of compounds (n=3–8).
`P for the nail plate (, ); P for the hoof membrane (, ).
`P is expressed in cm2/s. (, ) methyl, ethyl, butyl, hexyl and
`octyl nicotinates; ( , ) other substances (paracetamol, phen-
`acetin, diprophylline, chloramphenicol,
`iopamidol). Repro-
`duced from Ref. Mertin and Lippold (1997a), J. Pharm.
`Pharmacol., with kind permission from The Pharmaceutical
`Press.
`
`5). Movement of larger solutes through the ‘pores’
`in the keratin fibre network is obviously more
`difficult than the movement of smaller molecules.
`Fig. 5 also shows that the nail plate is less
`permeable than the hoof membrane. The authors
`suggested that the nail plate have a denser net-
`work of keratin fibres. A higher concentration of
`keratin fibres would result in greater chain–chain
`interactions,
`smaller
`‘pores’, overlapping of
`‘pores’, thus a more tortuous path for a diffusing
`molecule with consequently lowered permeation.
`The slope of the nail plate curve is twice as steep
`as that of the hoof membrane. This means that
`the nail plate is twice as sensitive to changes in the
`size of a diffusing molecule compared to the hoof
`membrane. Again, the denser network of the nail
`keratin was thought to be the reason. A denser
`network means there would be fewer pores whose
`size could accommodate the larger diffusing
`molecules.
`
`4.1.2. Hydrophilicity/lipophilicity of diffusing
`molecule
`Walters et al. (1983), studied the permeation of
`a series of homologous alcohols (C1–C12), di-
`luted in saline, through avulsed human nail plates.
`Increasing the chain length from one carbon to
`
`10
`
`
`
`S. Murdan /International Journal of Pharmaceutics 236 (2002) 1–26
`
`11
`
`Fig. 6. (a) Relationship between log of permeability coefficient and the alkyl chain length of the alcohol. Reproduced from Ref.
`Walters and Flynn (1983), J. Pharm. Pharmacol., with kind permission from The Pharmaceutical Press. (b) Permeability coefficients
`of the n-alkanols through nail plate from dilute aqueous solutions ( ) and from neat alcohols () as a function of alkyl chain
`length. Error bars indicate standard deviation. * indicates significant differences (PB0.05) between the permeability coefficients of
`neat and dilute alcohols. Reproduced from Ref. Walters et al. (1985a), J. Pharm. Pharmacol., with kind permission from The
`Pharmaceutical Press.
`
`of the compounds (Fig. 7). The slope of the curve,
`log permeability coefficient versus log partition
`coefficient (octanol/water), was nearly zero, indi-
`cating that
`the permeability coefficient across
`bovine hoof membrane was independent of the
`lipophilicity of the diffusing molecule. The au-
`thors argue that such independence is expected if
`the bovine hoof, like the nail plate is considered
`to be a hydrophilic gel membrane and it will not
`behave like a partition membrane. Fig. 7 also
`shows that the permeability coefficients of the
`alkyl nicotinates across the human nail plate de-
`creased significantly with increasing lipophilicity.
`However, this also contradicts a partition mem-
`brane model of the nail, where increase in drug
`lipophilicity is expected to result in increased drug
`permeation. The reduced permeation was assigned
`to the increasing molecular size of the nicotinates
`along the homologous series and the resulting
`reduced diffusion coefficients.
`
`The studies by Walters et al. (1983, 1985a) and
`those by Mertin and Lippold (1997a), agree as far
`as the nail plate is characterised as a hydrophilic
`gel membrane. However, unlike Walters et al.,
`Mertin and Lippold did not find any indication of
`a lipophilic pathway through the