A COMPARISON OF IN-VITRO SKIN-PENETRATION CELLS

A new low-volume flow-through diffusion cell (LVFC) was designed to pr ovide accurate determinations of penetrant flux across skin while mini mizing the dilution of penetrant in receptor fluid and eliminating the need for magnetic stirring. The performance of the 0.3-mL LVFC was co mpared to a magnetically stirred, 4.3-mL high-volume flow cell (HVFC) and to a magnetically stirred, manually sampled 7.5-mL static cell (SC ) with hydrophilic and lipophilic penetrants. The clearance of C-14-la beled benzoic acid from the LVFG and HVFC followed an exponential prof ile expected for complete mixing when the LVFC and HVFC were run at fl ow rates of 0.4-0.9 and 4.0-5.2 mL/h, respectively. The in vitro dispo sitions of C-14-labeled benzoic acid and estradiol were determined in the LVFC and HVFC by applying the compounds to splitthickness pig skin at a 4 mu g/cm(2) dose. Additionally, the effects of receptor fluid f low rate (1.2 vs 3.5 cell volumes/h) and method of skin attachment (O- ring vs compression) were determined on disposition in the HVFC. The p ercutaneous penetration of benzoic acid and the residue of estradiol w ithin skin did not differ between the LVFC and HVFC. However, the perc utaneous penetration of benzoic acid increased significantly (p < 0.05 ) using the O-ring attachment as compared to compression at a flow rat e of 1.2 cell volumes/h. The in vitro permeation of benzoic acid-satur ated water and 17 beta-estradiol-saturated propylene glycol monolaurat e through human epidermis was compared between the LVFC, HVFC, and SC. The LVFC and HVFC had flow rates of 0.9-1.0 mL/h. The static cells we re manually sampled; both receptor and donor chamber contents were rep laced with fresh liquid at 8-h intervals. The flux profiles calculated from the output of the LVFC accurately represented those of the SC. T he flux profiles calculated from the output of the HVFC were only comp arable following correction for penetrant remaining in the cell. The c learance of benzoic acid, percutaneous penetration of benzoic acid, an d penetration of estradiol into skin were similar in the LVFC and HVFC , but the LVFC was superior in requiring no magnetic stirring and appr oximately 1/15 the volume of receptor fluid. Additionally, skin-flux p rofiles from the LVFC output accurately represented those from the oth er cells, but without the inconvenience of manual sampling (SC) or cor rection of residual penetrant (HVFC).