Analysis of chromameter results obtained from corticosteroid-induced skin blanching. I: Manipulation of data.

PURPOSE: One of the unresolved issues in the FDA Guidance document for topical corticosteroid bioequivalence testing is the method of manipulation suggested for the chromameter data. The purpose of this study was to manipulate the instrumental data from a typical blanching study in a number of ways to investigate the appropriateness of these procedures for comparison with the subjective visually-assessed results.
METHODS: The human skin blanching assay methodology routinely practiced in our laboratories was utilised and the vasoconstriction produced by two corticosteroid formulations of different potency was assessed visually and instrumentally by use of a Minolta chromameter. The instrumental data were corrected for zero-time and unmedicated site readings. In addition, Euclidean distances were calculated using all data generated by the instrument.
RESULTS: Individually the a-, b- and L-scale chromameter values are imprecise and there is negligible vasoconstriction response recorded for the moderately potent formulation. Arithmetical manipulation of the data as suggested by the FDA does not appear to improve the quality of the data in any way. Euclidean distance analysis more closely resembles the visual data and appears to have better precision.
CONCLUSIONS: It is clear that mathematical correction of chromameter data is unnecessary, especially since the instrumental data are extremely imprecise. Furthermore, the assessment of each individual chromameter index does not adequately characterise the blanching response profile. It is therefore suggested that Euclidean distance may be a better measure on which to base an analysis of bioequivalence than the truncated data set methodology currently suggested by the FDA.