Eq. Dialysis Method (Matrix vs. Buffer) | Eq. Dilution Method (Diluted-Matrix vs. Buffer) | Eq. Competition Method (Matrix1 vs. Matrix2) | Flux Dialysis Method (Matrix vs. Matrix) | |
---|---|---|---|---|
Minimal dialysis time | (−) Establishment of equilibrium, (HTDialysis ≤4 h assuming minimal nonspecific binding) | (−) Establishment of equilibrium, (HTDialysis ≤4 h assuming minimal nonspecific binding) | (−) Establishment equilibrium, (hours, days, weeks, or months depending on fu) | (+) Receiver conc >> LLOQ, (always faster than equilibrium dialysis assuming same LLOQ |
Impact of device nonspecific binding | (−) Increased time to equilibrium, (although no impact on fu) | (−) Increased time to equilibrium (although no impact on fu) | (+) Nonspecific binding is minimized | (+) Nonspecific binding is minimized |
Impact of compound matrix instability | (−) Invalid fu | (−) Invalid fu | (−) Invalid fu if degradation rate is different between matrix1 and matrix2 | (+) No impact on fu |
Sample composition | (−) Matrix and buffer, two standard curves or sample blanking/matching require; Additional GLP validation and lower LLOQ may be required | (−) Matrix and buffer, two standard curves or sample blanking/matching require; additional GLP validation and lower LLOQ may be required | (−) Matrix1 and matrix2, two standard curves or sample blanking/matching require; Additional GLP validation may be required | (+) All samples identical matrix, single standard curve; preexisting GLP assay likely can be used |
LLOQ requirement | (−) LLOQ ≤ free buffer | (−) LLOQ ≤ free buffer, although free buffer increases roughly proportional with dilution factor | (+) LLOQ ≤ initial matrix conc assuming similar fu between matrix1 and matrix2 | (+) LLOQ ≤ 1/2 initial donor concentration |
Additional advantage/ disadvantages | (+) Most commonly used dialysis method | (−) Risk for saturating protein binding; (−) dilution does not reduce equilibrium time | (−) Discrete fu not determined, only fu ratio between matrix1 and matrix2 | (−) Unconventional time point(s) and data may require nonlinear curve fitting |
GLP, Good Laboratory Practice.