Subjects
Five healthy volunteers (caucasians, 4 fe-male and 1 fe-male, skin types III to IV), with no history of dermatological disease, aged 24-52 years (mean 33 years), participated in this study (approved by the local university ethical commitee), after written consent was ob-tained. The sites of drug administration were the relatively non-hairy surfaces of the ventral forearms. The volunteers refrained from any cosmetic treatment of the test areas during at least 12 hours prior to the experiment.
Chemicals
TBF was supplied by Novartis Pharma (Basel, Switzerland), and was included at 500 mg/ml in the vehicle (used as control formula-tion) which consisted of a 50:50 (v/v) mixture of ethanol (EtOH) (Fluka, Buchs, Switzer-land) and isopropyl myristate (IPM) (Sigfried, Zofingen, Switzerland). The test formulations included one of the following penetration en-hancers: 5% w/v oleic acid (OA) (Sigma, Steinheim, Germany), 10% w/v 2-pyrrolidone (2P) (BASF, Ludwigshafen, Germany), and 1% w/v urea (UR) (Sigma, Steinheim,
Ger-many). The enhancers were selected as model vehicle components, representative of differ-ent types of substances which have been most frequently studied in the past. The concentra-tions were selected somewhat empirically but were again consistent with previously pub-lished penetration-enhancing effects.
For TBF extraction and HPLC analysis, acetonitrile (HPLC grade), tetrahydrofuran (THF) (HPLC grade), tetramethylammonium hydroxyde pentahydrate (TMAH), and aque-ous triethylamine (TEA), all from Sigma-Aldrich, (Steinheim, Germany) were used.
Topical treatments
A cellulose pad (Tela, Basel, Switzerland) measuring 7 cm2 was soaked with 700 µL (corresponding to 50 mg/cm2 of TBF) of each formulation, and then applied to the ventral forearm skin via an adhesive polyurethane film (Opsite®, Smith-Nephew, Hull, UK), internally coated with an occlusive polyester film (Scotchpak® , 3M, St. Louis, MN). After 2 hours treatment, the patch was removed and the excess formulation gently blotted away (without solvents) by three fresh cellulose pads, without rubbing.
Tape stripping
This non-invasive technique permitted thin samples of SC (0.5-1 µm thickness) to be col-lected from the treated area. To allow uptake of any unremoved, non-absorbed superficial
formulation, which could lead to an overesti-mation of the amount of TBF in the first tape strip, a period of 30 min was allowed to pass before the first stripping. Non-treated skin surrounding the application site was protected with adhesive tape, so that the stripped area corresponded rigorously to the treated area.
The SC was sequentially stripped up to 20 times with a commercially available contact adhesive tape (Scotch® Book Tape 845, 3M, St. Louis, MN), which was pressed to the skin by rubbing six times back and forth with a spatula. The direction of stripping was alter-nated. This procedure was not normalized, since the inconsistent cohesion of the corneo-cyte layers [9] means that reproducible amounts of SC (within and between subjects) cannot be removed.
Up to 15 data points were obtained for each formulation and subject: up to 10 strips, the tapes were collected individually; the fol-lowing 10 were collected in pairs, in order to ensure a sufficient analytical sensitivity for the deeper SC layers, where the TBF concen-tration is very low.
The amount of SC removed was deter-mined by individually weighing the tapes be-fore and after stripping on a semi-micro bal-ance (precision 10 µg, Mettler-Toledo AT 261, Greifensee, Switzerland). Assuming that the SC density is constant (1 g/cm3) across its thickness [10], then the SC weights can be converted to volumes and, given that the area stripped is constant, to the SC thickness
re-moved. Finally, transepidermal water loss (TEWL) measurements, performed during the sequential tape stripping process allowed, as described elsewere [11], the total SC thick-ness to be determined and permitted, thereby, concentration profile data from different sub-jects to be presented and compared as a func-tion of relative depth into the barrier.
Measurement of TBF concentrations within stratum corneum
The in vivo spectroscopic determination was carried out similarly to a previously pub-lished study [6], in the attenuated total reflec-tance mode, by placing the treated site of the arm directly on the trapezoidal ZnSe crystal (reflectance angle of 45°) of a Nicolet 730 FTIR spectrometer (Nicolet, Madison, WI), equipped with a liquid nitrogen cooled mer-cury-cadmium-telluride detector. A spectrum of the skin was recorded before each removed tape strip, and the TBF level was inferred from the area of the characteristic absorbance peak centered at 774 cm-1 (due to the stretch-ing vibration of the aromatic C-H) where no major interfering SC absorbances were ob-served (Figure 2). However, since the contact of the arm with the ZnSe crystal is difficult to reproduce (resulting in variation of the spec-tral intensities independent of the amount of drug present), the area of the TBF peak was normalized with respect to the areas under the amide I (1645 cm-1) and II (1545 cm-1) bands
originating, respectively, from the carbonyl stretching and the N-H bending vibrations from SC keratin [8,12,13].
As FTIR allowed only a relative quantifi-cation of TBF in the SC, we also extracted and analyzed the tape strips using HPLC to obtain an absolute measure of the drug in the target tissue and, at the same time, to validate the spectroscopic measurements.
For each formulation and subject, all tapes (including the first one) were extracted during 16 hours in a 80:20 mixture of acetonitrile and aqueous TEA buffer 0.72 mM at pH 2.5.
The solution was then passed through a poly-amide syringe filter (Nalgene® 0.45 µm, Nalge, Rochester, NY) and the filtrate was
transferred into HPLC glass vials for quantifi-cation. The chromatographic system consisted of a model 600 pump, an autosampler 717 Plus (Waters-Millipore, Milford, MA), and a 12-cm Partisphere RP-18 column (Whatman, Clifton, NJ).
The mobile phase was isocratic, and com-prised a volumetric mixture of 50% acetoni-trile, 14.3% THF, and 35.7% TMAH buffer 1.59 mM at pH 7.8. The flow rate was 2 ml/min, and the analyte was detected at 280 nm. Under these conditions, the retention time of TBF was about 6 min. Comparison of the chromatograms of extracted blank tapes and tapes+SC did not reveal any interfering peaks with TBF. Drug concentrations were
deter-Figure 2: FTIR spectrum of human skin in vivo, after 2 hours treatment with the control formulation (TBF 500 mg/mL in EtOH:IPM 50:50). The inset shows limited spectra of (a) the skin before treatment, (b) the skin following treatment with TBF as stated above, and (c) TBF alone. For the semi-quantification of TBF, the peak located at 774 cm-1 was used.
mined using the AUC method from TBF cali-bration plots generated with the neat com-pound in the same solvent used for extraction.
The detection limit was 10 ng on column. The recovery of TBF was validated by spiking tape-stripped samples of untreated SC with 100 µl of a 10 mg/ml solution of TBF in ace-tonitrile (corresponding to 1 mg TBF per tape) and submitting them to the protocol de-scribed above. The extraction efficiency was 96.6 ± 1.9% of TBF (n = 5).
Analysis of the concentration profile data The concentration profile of TBF across the SC was fitted to the solution of Fick's sec-ond law of diffusion for the case of transport in a plane sheet [14]. The SC was assumed to be a homogeneous barrier, and the following boundary conditions applied (Figure 3):
1. At t > 0 and x=0 (at the SC surface)
The resulting solution for the concentration profile as a function of normalized position (x/L) in the SC and time (t) is:
where K is the apparent partition coefficient of TBF between the SC and the applied vehi-cle, Cv is the TBF concentration in the vehi-cle, and the ratio D/L2 represents the charac-teristic diffusion parameter of TBF across the SC. Thus, the best fits of Eq. 1 to the experi-mental results (infrared and HPLC) were de-termined (using Grafit software, version 3.03, Erithacus) and the values of K and D/L2,which optimized these regressions, were obtained.