Abstract for Poster 46

Percutaneous absorption of lipophilic compounds in aqueous vehicles – influence of dose volume

S.C. Wilkinson*¹, R.E. Wilkinson², F.M. Williams²
¹Health Protection Agency, Newcastle upon Tyne, United Kingdom
²School of Clinical and Lab. Sciences, Univ. of Newcastle, Newcastle upon Tyne, United Kingdom

Background

There is currently considerable interest in the use of predictive models based on quantitative structure activity relationships (QSAR) for the prediction of skin permeability of chemicals. Most of these models predict permeability coefficient (k_p) from large collections of experimental data, the majority obtained from in vitro studies. The measurement of k_p requires steady state absorption to be achieved, and such experiments usually use a large volume of a saturated, or near-saturated aqueous solution of the compound of interest in order to achieve the maximum thermodynamic activity, and hence create conditions for steady state absorption. However, such conditions are rarely achieved during occupational exposure; the volume of liquid in contact with the skin is usually much lower, and with certain compounds the dose may become depleted before steady state absorption could be achieved. The aims of the present investigation were: to establish the influence of dose volume on percutaneous absorption from near-saturated aqueous solutions of lipophilic compounds (log P 2.36 to 4.76), and in particular to determine whether flux and lag time from a finite volume experiment could be predicted from an infinite volume experiment; and to determine whether logP influenced percutaneous absorption in the same way in both finite and infinite dose volumes.

Methods

Human female abdomen skin from at least two donors per study was dermatomed at 330 micrometers and mounted on PTFE Scott Dick flow through diffusion cells. Minimal essential medium Eagle (supplemented with 2% (w/v) BSA), pH 7.4 was pumped beneath the skin at 1.5 ml/h and 37°C (giving a skin temperature of approximately 32°C) and with continuous stirring. The dose solutions were prepared from aqueous solutions of malathion (log P 2.36), testosterone (3.32), parathion (3.83) and triclosan (4.76), all at 90% of saturation, containing ¹⁴C labelled analogues of the test compounds (at least 200 000 DPM applied per cell). The dose solutions were applied at infinite (1 ml/cm²) or finite dose (0.025 ml/cm²), with six replicate cells in each case, using the same combination of skin donors. All dose solutions were occluded using a PTFE cap, and charcoal filters were placed over the donor chambers to trap any volatilised material. Receptor fluid samples were collected every 20 minutes during the first two hours after dose application, then hourly thereafter. At the completion of the test period (24 h), the remaining dose solution was carefully removed from the donor chamber with dry tissue paper swabs. The donor chamber was then carefully removed and washed twice with 70% (v/v) ethanol (washes analysed separately). The skin was carefully swabbed with three pairs of wet (soaked in 3% aqueous teepol L) and dry tissue swabs, each pair analysed separately. The remaining skin was carefully tape stripped (up to 15 tape strips with 3M magic tape, the first tape analysed separately). The tape stripped skin was digested at room temperature for 72 h in 1.5 mol/l potassium hydroxide in 80% (v/v) ethanol:water. The receiver chamber was washed twice in 70% (v/v) ethanol:water (washes analysed separately). All samples were analysed by LSC after mixing of swabs and tape strips, and aliquots of receptor fluids, cell chamber washes and skin digests, with HiSafe 3 scintillation cocktail. Swabs and tape strips were left to desorb for 72 h prior to scintillation counting. Concentrations of the test compounds in receptor fluids were used to generate a cumulative absorption-time curve. In infinite dose studies, the slope of the linear portion of this curve was the steady state flux. The intercept of this portion on the time axis was the lag time. In finite dose studies, the maximum slope of the linear portion of the curve was measured and is referred to in the results as the maximum flux. The intercept of this portion on the time axis was measured and is referred to as the apparent lag time.

Results

The maximum flux measured in finite dose studies was significantly lower than the maximum flux (steady state flux) measured in corresponding infinite dose studies for all four compounds (Table 1). The maximum flux measured in finite dose experiments was approximately a tenth of the steady state flux for malathion, testosterone and parathion, and some thirty-fold lower for triclosan. However, the rank order of maximum flux with finite dose was similar to that for steady state flux. The apparent lag time measured in finite dose studies was also significantly reduced with all compounds except testosterone.

Table 1 Percutaneous absorption of lipophilic compounds from saturated aqueous solutions applied at infinite (1 ml/cm²) or finite (0.025 ml/cm²) dose. Asterisks show significance of difference between infinite and finite dose (ANOVA, n=6 cells per treatment). Figures are means ± SEM.
Compound	malathion	testosterone	parathion	triclosan
Concentration (mg/ml)	0.14	0.02	0.01	0.01
Steady state flux (ug/cm²/h)	3.12±0.05	0.20±0.03	0.11±0.01	0.12±0.01
k_p(x 10^-3 cm/h)	21.8 ± 3.5	9.5 ± 1.3	10.7 ± 1.5	12.5 ± 1.0
lag time (h)	1.44±0.09	2.76±0.40	2.49±0.15	3.42±0.36
Finite dose maximum flux (μg/cm²/h)	0.35±0.08 ***	0.03±0.00 ***	0.01±0.00 ***	0.003±0.000 ***
Finite dose apparent lag time (h)	1.06±0.10 *	1.96±0.24	0.79±0.05 ***	1.03±0.09 ***

The absorption time profile also differed significantly between infinite and finite doses (Figure 1a and 1b). The cumulative absorption time curve obtained in infinite dose studies exhibited a constant linear phase once steady state absorption had been established. In contrast, the maximum flux measured in finite dose studies is more short lived (Figure 1b), and the flux decreased markedly after 4-5 h with malathion and testosterone. With triclosan, the maximum flux was measured for only 1-2 hours. However, in all cases, some degree of absorption continued to occur throughout the timecourse, despite the fact that the application vehicle evaporated after about 4 h.

The distribution of lipophilic compounds when applied in finite dose (Table 2) showed an overall decrease in absorbed dose (receptor fluid plus receiver chamber wash) with increasing log P, though the cumulative dose absorbed was very similar between malathion and testosterone, despite the difference in log P and the difference in measured flux. Furthermore, the proportion of the dose remaining unabsorbed with malathion was greater than that with testosterone and parathion. The proportion of the dose measured in the stratum corneum also increased with log P. A study of the percutaneous absorption of testosterone during the first four hours of exposure to a finite dose (conditions as in Table 2) showed that the proportion of the dose remaining unabsorbed decreased from 100% to 73.2 ± 3.4% after 1 h, and decreased further to 59.0 ± 5.1% after 3 h. This indicated that the surface dose was rapidly depleted during exposure to finite dose.

Table 2. Distribution of lipophilic compounds in human skin applied in finite dose volume (0.025 ml/cm²). Figures are means ± SEM
Compound	malathion	testosterone	parathion	triclosan
Concentration (mg/ml)	0.14	0.02	0.01	0.01
Applied dose (μg/cm²)	143.0	21.2	9.9	10.0
% unabsorbed	49.6 ± 7.0	34.4 ± 3.1	42.6 ± 5.2	75.8 ± 4.5
% in stratum corneum	4.7 ± 1.8	3.2 ± 0.8	10.5 ± 2.0	14.3 ± 1.2
% in membrane	5.6 ± 2.0	13.6 ± 1.4	12.9 ± 2.4	4.4 ± 0.7
% absorbed	42.3 ± 11.0	42.2 ± 3.1	23.8 ± 3.7	15.3 ± 2.6

Conclusions

The maximum flux (calculated from cumulative absorption-time profiles) with a finite dose volume of a near saturated aqueous vehicle was considerably lower than the steady state flux measured with an infinite dose volume for all four chemicals, indicating that steady state conditions were not achieved with the finite dose volume. The timecourse of absorption differed between finite and infinite dose volumes and between chemicals with finite doses.

The rank order of maximum flux was similar to that of steady state flux, suggesting that these parameters responded similarly to physicochemical properties of the test chemicals.

Although the maximum flux was maintained for only a short time with finite volumes, absorption continued at a slower rate throughout the timecourse, despite the vehicle having evaporated after about 4 h exposure. This illustrates that the absorption did not cease when the vehicle was no longer present.

There was no absolute relationship between the proportion of the dose absorbed with finite doses and log P. Although the maximum flux for testosterone was significantly lower than for malathion, the proportion of the dose absorbed was similar in each case.

Data for testosterone suggested that the surface dose was rapidly depleted during finite dose experiments, and this is a possible explanation for the much reduced absorption rates measured with finite dose volumes.

Acknowledgements: This research was carried out as part of the EDETOX project funded by the European Union (QRLT-2000-00196).

Content last modified: 24 May 2005