Tropical corticosteroid bioequivalence testing comparison of chromameter and visual data

Abstract

The major criticism of the human skin blanching assay is the subjective nature of grading the response. Recently the American FDA released a Guidance document for topical corticosteroid bioequi valence testing. The guidelines require the use of a chromameter as a reliable method to estimate skin blanching. The purpose of this study was to evaluate the recommendations of this document for appropriateness by comparing visual and chromameter data. The visually-assessed blanching assay methodology routinely practised in our laboratories was modified to comply with the specifications of the Guidance. The preliminary trial indicated that the training period that is required for a novice to be classified as an experienced observer is not a major problem. The major trend that emerged from the pilot study was that visual assessment was better than the chromameter. Longer dose durations were found to be more discriminatory than shorter durations. The visual data were best described by the sigmoid Emax model and the chromameter data were best described by the simple Emax model. The pivotal results indicated that the D2/Dj criterion to determine sample size of "acceptable blanchers" produced only few subjects suggesting that the validity of this criterion requires extensive investigations. The estimates of the Locke's confidence interval method were simiJar to those for the general simple formula. However, due to undefined parameters of the Locke's method in the Guidance, the validity of the Locke's method requires evaluation. The chromameter b-scale parameter was the least sensitive in estimating skin blanching whereas the a- and L-scale parameters produced similar results. Poor correlation between visual and chromameter was noted indicating that the visual method is still the best method.