Two-step in vitro-in vivo correlations: Deconvolution and convolution methods, which one gives the best...

Area of Science:

• Pharmaceutical Sciences
• Drug Development
• Biopharmaceutics

Background:

• Predictive dissolution tests and in vitro-in vivo correlations (IVIVCs) are vital for generic drug development, potentially replacing human bioequivalence studies.
• IVIVCs can be established using either a one-step or a two-step strategy.

Purpose of the Study:

• To compare deconvolution and convolution methods for developing two-step Level A IVIVCs.
• To determine if the linearity of in vitro and in vivo dissolution rates should influence the choice between one-step and two-step IVIVC approaches.

Main Methods:

• Comparative analysis of various deconvolution and convolution techniques for two-step IVIVC development.
• Evaluation of IVIVC validity and biopredictiveness under different dissolution rate relationships (linear vs. non-linear).

Main Results:

• Valid and biopredictive two-step IVIVCs were achieved when in vitro and in vivo dissolution rates exhibited a linear relationship.
• No single combination of deconvolution and convolution methods proved superior for two-step IVIVCs, as all prediction errors remained within acceptable limits.
• Two-step IVIVCs failed when the dissolution rate relationship was non-linear.
• The one-step IVIVC approach successfully generated valid IVIVCs irrespective of the linearity of the dissolution rate relationship.

Conclusions:

• The linearity of dissolution rates is a key factor in selecting an IVIVC development strategy.
• The one-step IVIVC approach offers greater flexibility, successfully accommodating both linear and non-linear dissolution rate relationships, making it a robust method for generic drug development.