L Endrenyi1, P Al-Shaikh. 1. Department of Pharmacology, University of Toronto, Ontario.
Abstract
PURPOSE: To develop a new method for the direct, sensitive evaluation of the equivalence of absorption rates in linear kinetic systems. METHODS: Concentrations are obtained before the earlier peak. Ratios of concentrations adjusted for the corresponding ratio of AUCs (area under the curve contrasting plasma concentration with time), or their logarithm, are extrapolated by linear regression to the time of drug administration. The intercept estimates the ratio of absorption rate constants (ka), or its logarithm. RESULTS: The intercept metric assesses the equivalence of absorption rates with very favourable characteristics. The metric reflects the ka-ratio specifically (i.e., not affected by other kinetic parameters), is approximately linear to it, exhibits high kinetic sensitivity and excellent statistical properties. With many observations, the intercept metric has near-ideal features, including high power for determining bioequivalence and the ability to detect a 25% difference between ka values. With only 3 or 4 measurements before the earlier peak, the performance of the metric depends on the preset regulatory conditions. Reasonably good power is noted if the bioequivalence limits determine a 50% difference between two metrics and, approximately, between two ka values. The intercept metric shows very high power with a wider bioequivalence range. The power declines only moderately with increasing intraindividual variation of ka. The equivalence of absorption rates is assessed with much higher power by the intercept metric than by Cmax. CONCLUSIONS: The excellent kinetic and statistical properties of the intercept metric enable the specific and sensitive determination of the equivalence of absorption rates.
PURPOSE: To develop a new method for the direct, sensitive evaluation of the equivalence of absorption rates in linear kinetic systems. METHODS: Concentrations are obtained before the earlier peak. Ratios of concentrations adjusted for the corresponding ratio of AUCs (area under the curve contrasting plasma concentration with time), or their logarithm, are extrapolated by linear regression to the time of drug administration. The intercept estimates the ratio of absorption rate constants (ka), or its logarithm. RESULTS: The intercept metric assesses the equivalence of absorption rates with very favourable characteristics. The metric reflects the ka-ratio specifically (i.e., not affected by other kinetic parameters), is approximately linear to it, exhibits high kinetic sensitivity and excellent statistical properties. With many observations, the intercept metric has near-ideal features, including high power for determining bioequivalence and the ability to detect a 25% difference between ka values. With only 3 or 4 measurements before the earlier peak, the performance of the metric depends on the preset regulatory conditions. Reasonably good power is noted if the bioequivalence limits determine a 50% difference between two metrics and, approximately, between two ka values. The intercept metric shows very high power with a wider bioequivalence range. The power declines only moderately with increasing intraindividual variation of ka. The equivalence of absorption rates is assessed with much higher power by the intercept metric than by Cmax. CONCLUSIONS: The excellent kinetic and statistical properties of the intercept metric enable the specific and sensitive determination of the equivalence of absorption rates.