Several sites are present within the follicle that represent potentially significant targets for drug delivery, but these may be limited by the structural aspects of the hair follicle and its chemical environment (14,15). Getting the drug to these sites is also limited by the chemical nature of the drug, for example, hydrophobic or hydrophilic. Aspects of the hair follicle structure that may impede drug delivery are the following.
1. The glass membrane surrounding the entire follicle and the keratinous layers of the inner and outer sheaths may physically restrict the passage of chemical agents deep within the follicle.
2. Sebum flow into the hair follicle may represent a physical barrier to drug delivery from some formulations. The effect of sebum composition on delivery of compounds into the follicles is not known. It is, however, likely that in the sebum-filled follicle, efficient drug delivery, and pharmacological activity would depend on the interaction of drug and sebum and the physicochemical properties of the vehicle.
3. Optimum particle size requirements may also be required to traverse the hair follicle. It was found (16) that delivery of polystyrene beads into the hair follicle was size-dependent. Beads between 5 and 7 mm in diameter were optimally deposited deep within the follicle, whereas smaller or larger beads were more likely to be localized in the SC and skin surface.
ROLE OF SEBUM IN THE FOLLICULAR ROUTE
Sebaceous glands connected to the hair follicle by ducts release sebum into the upper third of the follicular canal, creating an environment enriched in neutral, nonpolar lipids. Among mammals, sebum composition varies considerably. It has been reported that the sebum-enriched environment of the hair follicle that invagi- nates the epidermis into the dermis could provide significant lipoidal pathway for the transport of large lipophilic molecules. Chapter 15 shows a comparison of sebum composition among different species commonly used in percutaneous absorption studies (Table 2 in that chapter). Wide variations in sebum composition among species should be considered before extrapolating data from animal to human studies. However, knowledge of sebum lipid components and their physical chemical interactions helps us to consider how a penetrating vehicle will react upon entrance into the follicular canal.
The use of vehicles that are compatible with sebum to selectively transport drug via the transfollicular pathway is certainly not a new concept (17). The basis for understanding a vehicle’s effect on the delivery into the sebum-rich areas, such as the hair follicle, appears to be fully explained by conventional solubility properties. The Hildebrand coefficients for model sebum compositions demon- strate that sebum is overall a nonpolar, oily material, with a Hildebrand coefficient of approximately 7.5 to 8.0 cal1/2/cm3/2. Some authors believe (6) that many polar vehicles used in topicals, vehicles such as water, propylene glycol (PG), and ethanol would be too polar to be readily soluble in sebum. Hence, to deliver rela- tively polar drugs, if the vehicle is not specifically designed to solubilize sebum, there would be little chance to effectively deliver the drug to the deeper portions of the hair follicle. Hydrophilic or marginally hydrophobic drugs would not be
expected to be highly compatible with the sebum. For hydrophobic drugs, the transport of the drug is predominantly follicular and is brought about by cotran- sport of the drug dissolved in the oil phase into and across the PSUs (17). The effi- ciency of such a follicular enhancement in drug transport would depend on the solubility of the drug in the vehicle and the compatibility of the vehicle, with the sebum-rich lipid environment within the follicles. The deposition of a drug into sebaceous glands follows the rank order suggested in Table 1 (17). Thus, sebum therefore may represent a significant physical barrier to drug delivery from some formulations.
It has been reported that the highest pilosebaceous penetration occurred with vehicles that were mixtures of various solvents, interface active agents, coupling agents, and solubilizers (18). Studies identified isopropyl myristate, glyceryl dilau- rate (GDL), and polyoxyethylene-10-stearyl ether (POE-10) as being effective at delivering drugs into the pilosebaceous duct because of their compatibility with sebum. Studies have suggested that the follicular delivery may be dependent on the physicochemical properties of the drug and/or vehicle (19,20). The lipoidal environment of the follicular canal may favor certain drugs and manipulating formulation factors, and thereby may enhance delivery of vehicles and drugs by the follicular delivery route. Solvents, which interact with sebum, thus were opening the passageway for drug deposition within the follicle.
Since sebum is a mixture of different components, it is important to character- ize it. It is useful to know how variations in the components, their carbon chain length, and the ratios of unsaturation to saturation affect the physical but thermal behavior of sebum. This will provide direction on how to manipulate these proper- ties with topical delivery systems. The thermal behavior of sebum in response to these variables has been reviewed in Chapter 15. Vehicles that alter this behavior have also been reviewed in that chapter and in Motwani et al. (20). These are dis- cussed later in this chapter in relation to drug delivery.
MODELS USED TO STUDY FOLLICLES
A major limitation in elucidating the follicular pathway is the current lack of an adequate pharmacokinetic model that can clearly distinguish transfollicular from transepidermal percutaneous absorption. Animal models and human models can be used to study follicular delivery. The experiments can be in vivo or in vitro. These models can further be divided into models for studies of diseases and models to investigate follicular delivery. To date, the following models (Table 2) have been used and details on each model are given after the table, along with studies using these models in the next section.
Rabbit Ear Model
Kligman and Kwong (21) proposed this model for studying the comedogenic poten- tial of substances commonly found in cosmetic formulations. This model is very well-suited for assessing comedogenic compounds that are encountered in the etio- pathogenesis of acne vulgaris, acne cosmetica, chloracne, cutting oil acne, pitch acne, pomade acne, and tar acne (22). The procedure involves applying a comedo- lytic substance using a glass rod over the undersurface during a two-week period. After the animal is sacrificed, skin is sampled up to the level of the cartilage and is immersed in 608C water for two minutes, after which the epidermis is peeled off, with the follicular extensions still attached. The comedones can be observed under the microscope. For quantitative studies, the number of comedones can be counted and compared to comedones induced by other comedogenic substances.
Rhino Mouse Model
This model was used extensively by Kligman and Kligman (23). The skin of the rhino mouse, which is an allelic of the hairless mouse, contains deep dermal cysts and a huge number of utriculi filled with keratinous debris that resembles comedones. The horny cells do not contain P. acnes, but like in human acne, the sebaceous glands shrink as the pseudocomedones enlarge. They concluded that the rhino mouse is a suitable model for assessing chemicals that effect epithelial differentiation (retinoids) or that promote loss of cohesion between horny cells (des- caling agents). In the procedure, formulations are applied to the dorsal trunk of the mouse for a specified period and then the skin sections are histologically examined. The procedure involves application of the test substances to the entire dorsal trunk for a specified time.
Macaque Monkey Model
This model has been used to study drugs affecting hair growth. This species exhibits baldness due to hormonal and genetic factors, which is similar to andro- genic alopecia (19). The stumptailed macaque monkey model exhibits a species- specific frontal scalp baldness that coincides with puberty at a rate of nearly
100% in both sexes.
Uno and Kurata (24) investigated the effects of topically applied minoxidil and diazoxide on hair growth in bald macaque monkeys and they studied the activity of 5-a reductase inhibitors in nonbald monkeys. Hair and follicular growth were evaluated by obtaining phototrichograms (gross photographs and images of frontal scalps) after closely clipping the hairs and folliculograms (4 mm punch biopsies embedded in paraffin, serially sectioned and stained). The rate of DNA synthesis in the follicular cells was also determined. The phototrichograms revealed a conversion of short vellus hair to long terminal hairs, and the folliculo- grams showed enlarged follicles. The DNA synthesis studies indicated an increased rate of follicular cell proliferation. This model therefore appears to have promise for further work related to human androgenic alopecia.
Hamster Ear Model
Matias and Orentreich (25) examined this model by planimetry of peeled off skin instead of routine vertical histological sections. This model has been proposed as a model system for human sebaceous glands because of the similarities in mor- phology and in turnover time (26). In this case, each sebaceous gland unit consists of an infundibular canal, several large lobulated acini with 6 to 20 layers of sebo- cytes, and one centrally located piliary unit. Male hamsters are large, and the size of their sebaceous glands is similar to that of humans. Microscopic observation of the sebaceous glands from biopsies detects pilosebaceous drug deposition.
This model has been used extensively by Illel et al. (27). The isolation procedure produces both follicle-free and follicular skin. The hairless rat model is chosen because the sebaceous glands density and size are closely related to those of human forehead skin. The process involves inducing anesthesia to the animals, after which the animals are treated by immersion of the back in 608C water for one minute. The epidermis is removed from this treated area. The subsequent healing is monitored histologically and transepidermal water loss measurements (TEWLs) are done. After 9 to 10 weeks, the epidermis and horny layer regrow as a continuous layer, without the pilosebaceous structures, and the TEWL returns to normal, thereby indicating normal barrier properties. In this model, the SC looks normal, but the epidermis looks hyperplastic. The permeation is done in vitro to see the differences between follicle and follicle-free skin. However, aside from the pain and suffering inflicted on the animal and the time involved in developing the follicle-skin, there is the question of structural composition and permeability behavior of the newly regrown tissue compared with the normal epidermis of hairless rat skin and whether the comparison between the two tissue types is valid. A hairless rat has also been used without the treatment mentioned earlier, with autoradiography to give a visual impression of the follicu- lar route (28).
Wahlberg (29) used hairy and nonhairy skin of guinea pigs to demonstrate the delivery into the follicles. For the hairy skin, the back of the guinea pig was used, whereas for the nonhairy skin, the area behind the ears of guinea pigs, which is completely devoid of hair follicles and of sebaceous and sweat glands, was used.
However, the physiological differences between the two areas make it difficult to justify its use. Also, in some cases, different strains of guinea pigs (hairy and hair- less) (30) have been used to investigate the importance of follicles. However, direct comparison due to differences in species is difficult.
The human skin contains acne-prone areas with numerous huge sebaceous follicles—an ideal site for acne vulgaris and other such follicular diseases. The human model can be used for both investigating follicular diseases as well as follicular delivery. The human skin is generally investigated by the noninvasive cyanoacrylate technique or the skin biopsy technique, the details of which are given in the next section.
METHODS TO INVESTIGATE DEPOSITION INTO THE FOLLICLES
The skin is a multilayered organ, complex in structure and function, composed of outer epidermis and inner dermis. There are various potential pathways for per- meation through the SC, including the transepidermal (or “bulk”) and the trans- appendageal (or “shunt”) routes. The transepidermal route across the continuous SC and the epidermis comprises two routes of entry: the intercellular spaces, con- sisting of bilayers of lipids, and the intracellular or transcellular route through the keratin of the dead cornified cells (corneocytes). The transappendageal route com- prises transport via both the sweat glands, hair follicles with their associated sebaceous glands. Capillaries closely surround both the follicles and the sweat glands, and rapid absorption of permeating molecules is assumed once the mol- ecules have passed through the follicular walls. An important aspect to consider for transdermal delivery is the mechanism of penetration, that is, the fractional contribution of the bulk strata versus the appendageal pathway to the total absorption. Some of the techniques, both qualitative and quantitative, are sum- marized next.
Skin Biopsy Techniques
Noninvasive Cyanoacrylate Technique
Marks and Dawber (31) expanded the ancient technique to remove horny cell layers by presenting a very potent glue—cyanoacrylate—and termed it skin surface biopsy. This technique was used for excavation of follicular contents, analogous to the painful squeezing of sebaceous filaments. It involves placing a drop of adhesive on a clean region of the skin and then gently pressing a slightly moist glass slide on top of the area. After 30 to 40 seconds, the slide is removed along with the superficial horny layer of SC and follicular casts. Lavker et al. (32) have investigated the earlier events of comedo formation in relation to the bacterial con- tents using this technique. Mills and Kligman (33) used this model to assay comedolytic substances. Microcomedones were induced on the back of adult males by a two-week occlusive exposure to 10% crude coal tar. The test agents were then applied for two weeks and the reduction in the density of microcome- dones is determined by the noninvasive cyanoacrylate technique. This method was used to screen anti-acne formulations. Mills et al. (34) analyzed SA levels in sebaceous follicles after topically applying it in a 2% hydroalcoholic vehicle or gel base.
Azelaic acid (20%) (AzA) cream was used on human volunteers for the treat- ment of acne (35). Surface AzA was removed by washing with acetone and then col- lected the follicular casts using follicular biopsy method and analyzed by HPLC method. Initial samples were taken in five minutes and thereafter over a period of five hours, five times with no fixed intervals.
Often, a full thickness biopsy containing several sebaceous follicles is needed. Biopsies are inevitable if sebaceous glands are to be investigated. It has been used for human tissues where the tissue is used within 24 hours of excision (36). After the permeation experiments in vitro, the SC is removed by stripping and then the biopsies are taken from the formulation-treated skin. The biopsies are immersed in CaCl2 for a few hours and then the epidermis and dermis is separated. Under microscopic observation, the epidermis with the attached PSU is removed from the dermis. These sebaceous glands are cut off from the under side of the epidermis and then analyzed by a suitable analytical technique.
Mechanical Separation of the Follicles
This has been documented recently by Lieb (37). The procedure involved using the Syrian hamster ear as a model. The scheme involved mounting the ears, ventral side up on the diffusion cells, which maintained the temperature at 328C. An aliquot of the test preparation was then applied to the skin surface. At predetermined time intervals, the skin specimens were separated by gentle peeling or scraping into three layers: epidermis, ventral dermis, and dorsal dermis. The sebaceous glands in the ventral dermis were removed by gentle dragging of a dull scalpel across the bottom surface of the layer. The pilosebaceous material appeared as a milky material. The scraping process was considered complete when the skin areas pre- viously containing glands appear empty under the microscope. The scraped sebac- eous glands and skin layers were assayed separately by scintillation counting.
This has been explained earlier in the rat model. Illel and Schafer (38) reported the first systematic, quantitative in vivo studies of follicular delivery using this method. By employing the Franz diffusion cell method, it is possible to compare penetration of drug across both follicle-free and normal skin (30).
Skin Autoradiography by Computerized Image Analysis
In this technique, the spatial distribution of a radiolabeled substance within a biological specimen is detected by exposure of the specimen to radiation-sensitive film. The main drawback of qualitative autoradiography is that it merely enables visualization but not quantification of the penetrant in the skin. Furthermore, the signals appearing on the images generally represent the tissue-bound residues rather than the freely diffusable substance. The technique, therefore, only provides an indication of drug transport through the integument.
Some workers have used quantitative autoradiography (39). With this method, it is possible to visualize and measure by means of a software program the concentrations in the hair follicle, glands, and various skin layers without
resorting to mechanical horizontal sectioning. The general procedure involves freezing the test skin sample at 21308C, and then vertically sectioning them by cryo- stat in order to yield 6-mm thick slices. These sections are then fixated on glass slides, placed in film cassettes, and covered by tritium-sensitive autoradiography film. Following a four- to seven-week exposure period, the autoradiograms are developed. An imaging system consisting of a video camera, high-resolution light microscope, IBM-compatible computer, and appropriate software transforms the optical density readings of the experimental autoradiograms into drug concen- trations values and projects the resultant digitized images on a TV screen, where they are displayed in pseudocolor mode. This provides a more visually meaningful picture of drug localization. In addition to autoradiography, the original skin sec- tions are stained with hematoxylin and eosin dye. In the final step, the instrumen- tation is used to superimpose the autoradiography image with the histological image. The computerized system is then used to quantify drug deposition densities into drug concentration. These values require reference to calibration graph data, which are derived from a separate set of experiments. In these calibration experi- ments, good reproducibility was shown between the optical density readings of standard sections containing the same drug concentration (40).
Confocal Laser Microscopy
This method depends upon the analysis of re-emitted light, with the measuring depth dependent upon the beam wavelength. However, in fluorescence spec- troscopy, the remitted light is detected after passing a second monochromator, that is, only a specific emission wavelength is detected. The procedure involves application of a self-fluorescent marker in a formulation on the skin. After a prede- termined time, the diffusion cells are dismantled and the area of the skin exposed to treatment is punched out. The peeling and scraping technique previously described is used to mechanically isolate the sebaceous glands. The scraped glands are then suspended in a buffer and sonicated in order to release the fluorescent marker. This solution could then be assayed for the marker with a fluorescent microscope. This technique has been used quantitatively by Weiner et al. (18).
Horizontal Sectioning of Strata
This technique involves the application of the test compound (typically radiolabeled) to a specific area of skin. After a defined exposure time, the excess drug is removed. The horny layer is stripped off by the consecutive application of an adhesive tape. The underlying tissue can then be excised, and the epidermis and dermis are sectioned parallel to the skin surface with a freezing microtome. The quantity of drug in each horny layer strip or tissue slice can be determined by using a suitable analytical technique. From the derived data, it is possible to generate a profile representing drug concentration versus depth for the investigated tissue. A disadvantage of this technique is the fact that unless sufficient data is available from the literature about the thickness of each skin layer at the required regional site, preliminary experiments are needed to assess this. Another major limitation of this technique is that it does not yield any data about penetrant concentrations in the PSUs. It usually has to be used in conjunction with other techniques described earlier (41).