H. Loth et al., Statistical testing of drug accumulation in skin tissues by linear regression versus contents of stratum corneum lipids, INT J PHARM, 209(1-2), 2000, pp. 95-108
This investigation is a contribution to standardization in in vitro drug pe
netration measurements using excised human skin and to statistical treatmen
t of the observations. The wide variations observed in measurements of drug
accumulation in and drug permeation through the stratum corneum are caused
not only by analytical errors but also by the variability of the horny lay
er lipid composition. The last-mentioned systematic influence can be compen
sated for by stepwise (multiple) linear regression using the contents of th
e main lipid classes as independent variables. In consequence, the S.E. of
estimate given by the regression calculation is lower than the S.E. of the
means of the observations. Significant differences in drug quantities accum
ulated in skin tissues (stratum corneum and dermis) are sensitively detecte
d by Chow's P-test of structural change. Accumulation data of flufenamic ac
id and hydrocortisone penetrated from different bases are given as examples
. The calculation mode is exemplarily explained and discussed. The results
of the test for structural change, two-independent-groups t-test and paired
-samples t-test are compared. The F-test of structural change proves to be
a helpful statistical method suitable to the assessment of biopharmaceutica
l quality parameters and to measurements using biological materials. (C) 20
00 Elsevier Science B.V. All rights reserved.