BACKGROUND: Incurred sample reanalysis (ISR) is the most recent in-study validation parameter that regulatory agencies have mandated to ensure reproducibility of bioanalytical methods supporting pharmacokinetic/toxicokinetic and clinical studies. The present analysis describes five representative case studies for macromolecule therapeutics. METHOD: Single ISR acceptance criteria (within 30% of the averaged or original concentration) and a modified Bland-Altman (BA) approach were used to assess accuracy and precision of ISR results. General concordance between the two criteria was examined using simulation studies. RESULTS: All five methods met the ISR criteria. The results indicated that thorough method development and prestudy validation were prerequisites for a successful ISR. The overall agreement between the original and reanalyzed results as determined by BA was within 20%. Simulation studies indicated that concordance between the ISR criteria and BA was observed in 95% of the cases. Dilution factors had no significant impact on the ISR, even for C(max) samples where 1:100 or higher dilutions were used. CONCLUSION: The current ISR acceptance criteria for macromolecules was scientifically and statistically meaningful for methods with a total error of 25% or less.
BACKGROUND: Incurred sample reanalysis (ISR) is the most recent in-study validation parameter that regulatory agencies have mandated to ensure reproducibility of bioanalytical methods supporting pharmacokinetic/toxicokinetic and clinical studies. The present analysis describes five representative case studies for macromolecule therapeutics. METHOD: Single ISR acceptance criteria (within 30% of the averaged or original concentration) and a modified Bland-Altman (BA) approach were used to assess accuracy and precision of ISR results. General concordance between the two criteria was examined using simulation studies. RESULTS: All five methods met the ISR criteria. The results indicated that thorough method development and prestudy validation were prerequisites for a successful ISR. The overall agreement between the original and reanalyzed results as determined by BA was within 20%. Simulation studies indicated that concordance between the ISR criteria and BA was observed in 95% of the cases. Dilution factors had no significant impact on the ISR, even for C(max) samples where 1:100 or higher dilutions were used. CONCLUSION: The current ISR acceptance criteria for macromolecules was scientifically and statistically meaningful for methods with a total error of 25% or less.