The skin forms 10% of body mass and is considered a protective barrier, preventing water loss and maintaining homeostasis (
1). Despite this barrier function, pharmacologic agents can penetrate into and through skin layers (
2), an effect that can be potentiated by various penetration enhancers (
3). The convenience and accessibility of the transdermal route of drug administration have influenced the increasing prominence of topical medication in human and veterinary clinical practice.
An uncertainty with using topical medications is exactly how much active constituent or vehicle of a formulation will penetrate through skin and be available in both local tissues and systemically in the target species (
4). Several factors affect transdermal drug penetration, including skin blood flow (
5,
6), number and type of appendages such as sweat glands and hair follicles (
7,
8), and the physicochemical characteristics of the skin (
3). Therefore, interspecies extrapolation of the kinetics of transdermal drug penetration is impractical, and investigation of the rate and extent of drug movement through skin must be undertaken in the target species (
1).
One constant factor in transdermal drug penetration is the stratum corneum (SC). This layer of the epidermis is the main barrier to solute and water movement through skin (
2,
9). Many studies of transdermal drug penetration have used normal skin, particularly skin with an intact SC (
10). Although there is limited information on the specific effects of skin disease on transdermal drug penetration, it is reasonable to assume that progressive loss of the SC will greatly diminish the barrier function of skin (
11). A study of human skin used microdialysis to show that damage to the SC in vivo, induced by tape-stripping or by application of an irritant (sodium lauryl sulfate), or both, significantly increased the penetration of salicylate (Sa) (
12).
An important consideration when applying topical formulations to veterinary species, such as the horse, is that the integrity of the SC may be compromised by prior treatments used to prepare the skin site for topical applications. For example, application areas are shaved, as hair could be considered the initial barrier to topical application of medications in many species, reducing skin contact with the applied formulation (
3). More importantly, preparations used to clean or disinfect the skin, particularly those containing alcohols, can irritate and disrupt the SC (
13,
14). Any variation in transdermal drug penetration in the horse will have implications for efficacy, adverse effects, and, especially for competing animals, systemic bioavailability of the active drug, its metabolites, and vehicle components.
In this study, we used in vitro techniques to investigate the effects in the horse of commonly used skin pretreatments on the transdermal penetration of a topical anti-inflammatory formulation.
Horse skin was harvested from 5 Thoroughbred geldings that had been presented to the University of Queensland Veterinary School for euthanasia. The horses were euthanized by an intravenous injection of sodium pentobarbital, and the hair over the central thorax, approximately midway between the costochondral junction and the vertebrae, was removed by electric clippers (Cryotech blade, size 40; Oster Professional Products, Toronto, Ontario). The skin sections were dissected away, with care taken to trim off subcutaneous fat, and frozen at −20°C until required (
15). This protocol was approved by the Animal Ethics Committee of the University of Queensland (approval number SVS/087/04/RIRDC).
A 20% methylsalicylate (MeSa) gel (Dencorub Pain Relieving Gel; Church & Dwight Pty. Ltd., Brookvale, New South Wales, Australia) and concentrated chlorhexidine gluconate (Chlorhex C: 50 mg/mL chlorhexidine gluconate; Jurox Pty., Rutherford, New South Wales, Australia) were purchased from a pharmaceutical wholesaler. All other chemicals, including Sa, MeSa, p-toluic acid, and albumin (bovine fraction V) were purchased from a manufacturer (Sigma Pharmaceuticals, St. Louis, Missouri, USA). Solutions of aqueous chlorhexidine (0.1% w/v) and alcoholic chlorhexidine (0.5% w/v in methanol) were prepared according to the manufacturer’s recommendations.
The thawed thoracic skin samples from each horse were divided into 5 sections with a marker pen, and 1 of the following preparations was applied to each section in randomly selected order for each horse: (a) no preparation, as control; (b) 0.1% w/v aqueous chlorhexidine (Aq-C), applied with a ball of cotton wool that was gently rubbed over the skin site 3 times; (c) 0.5% w/v alcoholic chlorhexidine (Al-C), applied as for aqueous chlorhexidine; (d) shaving (Sh), in which a single-bladed disposable razor was applied once over the skin site to remove the remaining hair; and (e) tape-stripping (Ta), in which adhesive tape was firmly applied to the skin section and then removed. An average of 3 replicates was used for each treatment in each horse, with final results representing the mean of this figure for the 5 horses for each treatment and time point.
The skin was then cut into circular sections (approximately 2 cm in diameter) and mounted in Franz-type diffusion cells, with the SC uppermost. A measured volume (approximately 3.2 mL) of phosphate-buffered saline (PBS), pH 7.4, containing 4% bovine serum albumin (BSA) as a receptor fluid was added to the lower reservoir with a magnetic flea for stirring. Next, 1 mL of PBS was added to the donor reservoir. The skin cell was placed in a water bath containing a magnetic stirring plate and allowed to equilibrate at 35°C for 60 min. The surface temperature of the skin in the diffusion cell was approximately 32°C (confirmed with a digital thermometer). The PBS was removed from the donor reservoir and 2 g of MeSa gel added to each diffusion cell (time 0). A 200-μL sample was collected from the receptor fluid via a side port of each diffusion cell and immediately replaced with fresh solution at 1, 2, 3, 4, 6, and 10 h (
16). The sample was frozen at −20°C until analyzed for Sa and MeSa, with the use of high-performance liquid chromatography as described previously (
17), within 48 h. The reported values represent a mean of the 5 data points for each horse at each time point.
The area under the curve (AUC) was calculated for the MeSa and Sa measurements in the receptor fluid at each collection time with the use of curve-fitting software (GraphPad Prism, version 4.00 for Windows; GraphPad Software, San Diego, California, USA). Student’s t-test for unpaired data was then used to compare the differences in the AUC for MeSa and Sa for the different skin pre-treatments compared with control (clipped only). The initial rates of MeSa and Sa absorption (flux; mg/cm2·h) were calculated from linear regression analysis of the cumulative rates between 0 and 10 h. Lag times for penetration through skin were calculated from the intercept with the y-axis. The 2-tailed Student’s t-test for unpaired data was used to compare the means for different skin pretreatments compared with control (clipped only). The parameters for the combined transdermal penetration of MeSa and Sa were calculated for illustration only; no statistical analysis was performed.
Over the 10 h of the study, a significantly higher proportion of MeSa penetrated through skin treated with Al-C or Sh (
P < 0.01) and with Aq-C or Ta (
P < 0.05), and significantly more Sa was detected in the receptor phase of skin treated with Ta (
P < 0.01), Al-C, Aq-C, or Sh (
P < 0.05) compared with untreated skin (
Figure 1). Similarly, the absorption rate or flux of MeSa was significantly higher (
P < 0.05) after each of the treatments, compared with no treatment, although the rate of appearance of Sa was significantly higher (
P < 0.05) only for skin treated with Al-C or Ta compared with no treatment (
Table I). A significantly shorter lag time (
P < 0.05) was calculated for MeSa for all treatments, although the shorter lag time for Sa appearance in the receptor phase was significant (
P < 0.05) only after Al-C treatment (
Table I).
 | Figure 1Penetration over 10 h of methylsalicylate (MeSa, mg/L) and its active metabolite, salicylate (Sa, mg/L), through equine skin from 5 horses pretreated as follows: cleaned with alcoholic chlorhexidine (![[filled square]](corehtml/pmc/pmcents/x25AA.gif) ), tape-stripped (♦), cleaned with (more ...) |
 | Table I Penetration in vitro over 10 h of salicylate (Sa) and methylsalicylate (MeSa) through equine thoracic skin pretreated in various ways |
The results of this study conclusively demonstrate that common cleaning and preparation techniques applied to the skin can significantly affect transdermal drug penetration in the horse. It could be argued that clipping itself may also be a “pretreatment” of the skin, although removal of the upper hair coat primarily removes the physical barrier, which may impede contact of a topical formulation with the SC. Histologic examination of clipped equine skin revealed that hair follicles protrude marginally above the skin surface, compared with unclipped regions, with no loss or damage to the SC region (unpublished observations). Skin stored by freezing is frequently used during in vitro studies to investigate percutaneous penetration through human and animal skin. Freezing is unlikely to have affected the results in this study because the freezing process did not appear to have affected the integrity of the SC (
3). More importantly, the method used in the current study permitted simultaneous comparison of several treatment effects on skin permeability, greatly reducing interindividual and day-to-day variability.
Many reports in the literature on transdermal pharmacokinetic and pharmacodynamic studies have based findings on normal skin, particularly an intact SC (
2). However, it is acknowledged in the literature that disease processes can alter lipid composition and disrupt the structure of the epidermis (
18). Similarly, it has been shown that the SC could be inadvertently damaged by preparation of the skin before application of a topical agent (
19). In the current study, chlorhexidine significantly enhanced the rate and amount of MeSa penetration and the appearance of its active metabolite. This may be due to physical contact and removal of upper layers of the SC during application but is more likely related to a change in the constituents of the SC, such that the barrier to this relatively hydrophilic commercial preparation was decreased.
Extraction of intercellular lipids with various solvents causes a significant reduction in the barrier function of the SC (
20). Alcohols, particularly the more polar members, are solvents known to irritate the skin by delipidizing the membrane and disrupting the SC (
13,
14,
21). Small-chain alcohols, such as methanol and ethanol, may be in preparations used to clean skin in many veterinary and human applications, yet it is well known that the barrier function of skin can be greatly diminished by the loss of intercellular lipids and the subsequent disruption of the SC structure (
22). In our study, the addition of methanol to chlorhexidine induced a small but not significant increase in the rate and extent of MeSa and Sa recovery in the receptor phase compared with chlorhexidine alone. It is uncertain whether the methanol was acting synergistically or was affecting the SC by a mechanism different from that of chlorhexidine. Some alcohol- containing vehicles have been shown to enhance transfollicular delivery of drug molecules, and this has been related to the solvent nature of the alcohol, acting on sebum within the follicle (
23).
Direct removal of some of the SC by tape-stripping or shaving also significantly increased the transdermal penetration of MeSa and Sa through equine skin. Repeated tape-stripping is a useful technique to investigate transdermal drug penetration and the barrier function of different layers of the epidermis in vivo and in vitro (
11,
24,
25). Since progressive loss of the SC will greatly diminish the barrier function of skin (
11), the results of the current study indicate that procedures commonly used to prepare skin sites can result in significantly higher penetration of topically applied drugs. We did not specifically measure the extent of SC loss after pretreatment of the skin samples, although the procedures for shaving and the application and removal of adhesive tape were based on what might be expected during preparation of horse skin for clinical procedures. Furthermore, relatively more Sa and less MeSa was measured in the receptor fluid after tape-stripping, although the differences were not significant. The relative contribution of the different layers of skin to cutaneous metabolism and resistance to transdermal penetration, particularly in models of skin damage, warrants further investigation.
This study has shown that disruption or loss of the SC will reduce the barrier function of skin with regard to topically applied formulations. Once a drug has penetrated this principal defense against water loss and solute passage, several other factors (including skin blood flow, number and type of appendages, and the physicochemical characteristics of the skin) will determine the rate and extent of the systemic drug concentration. However, it is reasonable to assume that increased transdermal drug penetration will correlate with higher systemic drug concentrations. It should therefore be acknowledged that routine pretreatment of skin may result in higher efficacy and, more importantly, an increased potential for adverse effects when topical formulations are subsequently applied.
