STRUCTURE AND PERMEABILITY OF HUMAN NAIL 371 controlling lipid pathway for the higher alcohols was interfered with by extraction of the very low, but necessary, lipid level in the nail plate (34). In a study by Soong (35), the in vitro penetration of acetic acid (molecular weight = 60), benzoic acid (molecular weight = 122), and suprofen (molecular weight -- 260) through nails was investigated. Steady-state conditions were achieved in these experiments, and the permeated drug was assayed by HPLC. The results of this study showed that as the size of the permeant increased, the lag time increased and diffusion coefficients de- creased. The data from this detailed investigation indicated that small, lipophilic mol- ecules would be good candidates for topical nail therapy, both in terms of shorter lag times and faster penetration into the nail plate (due to effective drug partitioning). However, the author cautions against using a very highly lipophilic molecule since water has been shown to facilitate drug transport across the nail plate (33). In order to ascertain whether the dorsal part of the nail plate (which contains tight intercellular junctions) constituted the main barrier to drug penetration, the permeability of benzoic acid through intact as well as shaved (dorsal part removed by scraping) nails was studied (35). No dramatic increase in the permeability coefficient occurred with the shaved nails (which was expected if the dorsal nail constituted the main barrier). Thus, the author speculated that the intracellular structured keratin was the main barrier to the perme- ation of drugs. A recent study has further attempted to explain the transport mechanism of permeants through the nail. In this investigation, Mertin and Lippold (36) studied the in vitro permeation of homologous nicotinic acid esters (with octanol/water partition coefficients ranging from 7 to 51000) through human nail and a keratin membrane derived from bovine hooves. The diffusion cells used by these researchers were of the Franz-type, temperature was maintained at 32øC, and drug analysis was performed spectrophoto- metrically for studies with the hoof membrane, and by HPLC for studies with the nail plate. It was found that the permeability coefficients through both human nail and bovine hooves did not increase with increasing partition coefficient or lipophilicity. Thus, the authors stated that the nail appears to behave as a hydrophilic gel membrane rather than as a lipophilic partition membrane. Penetration of acetaminophen and phenacetin showed that maximum drug flux was a function of solubility in water or the swollen keratin matrix. This study showed that water solubility of the drug is an important consideration in formulating a topical product for nail disorders, as increased water solubility can enhance maximum drug flux. The relationship between the molecular weight (MW) of several drugs (including a series of antimycotics) and permeability through the nail plate and hoof membrane was also investigated (37). The donor compartment consisted of a drug suspension in ethanol (42% v/v), pH 8.1. Due to the limited aqueous solubility of antimycotics, the same medium (without drug) was used as the receptor fluid. The effect of molecular size on the permeability coefficient of antimycotics through the hoof membrane (P•oooe) is de- picted in Figure 3. A linear relationship with a negative slope between log PHoooe and log MW was reported for the antyimycotics (Figure 3). Similar trends were seen with other drugs (nicotinic acid esters, acetaminophen, phenacetin, etc.) for permeability coeffi- cients through both nail plate (P) and hoof membrane (PHoooe). However, the permeability coefficients of the drugs tested were higher through the hoof membrane than through the nail in all cases. This was considered to be due to the denser keratin network (and therefore fewer pores and increased tortuosity) in the nail as compared with the hoof

372 JOURNAL OF COSMETIC SCIENCE Log MW 2.00 -6.50 -7.00 -7.50 -8.00 -8.50 -9.00 -9.50 2.20 2.40 2.60 2.80 3.00 Figure 3. Relationship between the logarithm of molecular weight (MW) of various antimycotics and the logarithm of their permeability coefficients through a bovine hoof membrane (PHoof). Adapted from reference 37. membrane, and possibly due to distinct differences between the two keratinous tissues. However, due to the limited availability of nails and the similarity between nails and hoof membranes (both act as hydrophilic gel membranes), the authors suggested that the hoof membrane may serve as an appropriate in vitro model to predict nail plate perme- abilities. The authors derived the following empirical equation to relate permeability coefficients through the nail and hoof membrane for different drugs: log P = 3.723 + 1.751 log PHoof (1) where P represents the permeability coefficient through the nail plate, and PHoo/repre- sents the permeability coefficient through the hoof membrane. Thus, if PHoo/values are experimentally determined, P values can be calculated. However, due to inherent dif- ferences between the two keratinous membranes, caution should be exercised in extrapo- lating data from the hoof membrane to the human nail. As stated earlier, the human nail plate is made of dorsal, intermediate, and ventral layers. The in vitro permeation of a water-soluble drug (5-fluorouracil) (5-FU) and a water- insoluble drug (flurbiprofen) (FB) through the different layers of the nail plate, using a modified side-by-side diffusion cell, has been reported (38). The thickness ratio of each layer (dorsal:intermediate:ventral) was assessed to be 3:5:2. The dorsal-filed, ventral- filed, and dorsal-and-ventral-filed layers of the nail plate were prepared by filing fin- gernail clippings with sandpaper to a known thickness. The drug concentrations were measured by HPLC. In this study, it was found that lipids are predominantly present in