ANALYSIS OF PERCUTANEOUS PERMEATION DATA .2. EVALUATION OF THE LAG TIME METHOD

To evaluate the lag time method, the permeation of four drugs, propran olol (PR), triamcinolone acetonide (TA), physostigmine (PHY), and tetr ahydroaminoacridine (THA), was studied across two skin membranes, hair less mouse skin (HMS), and human cadaver skin (HCS). The permeation st udies were conducted with vertical Franz diffusion cell in the static mode, using an infinite dose technique. The resulting permeation profi le for each individual drug was analyzed by the lag time method to est imate parameters such as the steady-state flux (J(ss)), lag time (T-la g), diffusion coefficient (D) and skin/donor-phase partition coefficie nt (K-m). The diffusion coefficient (D) and skin/donor phase partition coefficient (K-m) were used to regenerate the entire permeation profi le (pre-steady and steady states) using an equation based on Pick's la ws of diffusion. The fit of the observed data to the regenerated curve was estimated by linear regression with the observed data as the inde pendent variable. In experiments where steady state should have been a chieved at the end of the experiment (experimental duration > 3 x T-la g), R(2) was greater than 0.993, the slope was unity and the y-interce pt zero for the regenerated profile vs the observed data regression (e .g., PR 72 h, TA with 2% Atone (AZ) pretreatment and 2% AZ as copenetr ant, PHY, and THA experiments). For the PR 27 h experiment, TA control and TA with 2% propylene glycol (PG) pretreatment experiments, the sl ope was significantly greater than unity for linear regression of the regenerated profile vs the observed data. In these experiments, the ex perimental duration was less than 3 x T-lag and steady state may not h ave been achieved, which could have resulted in inaccurate estimates o f D and K-m and hence, the positive bias in the regenerated profiles. Overall, the lag time method was successful in the estimation of perme ation parameters from which the entire profile could be regenerated fo r all the drugs studied. However, the lag time method may result in in accurate estimates of D and K-m if steady state is not achieved as dem onstrated by a positive bias (slope > 1) in the regenerated profile. T he approach used in this study could be used to verify whether D and K -m obtained from a permeation experiment are precise, accurate and hen ce reliable.