Single concentration loss of activity assay provides an improved assessment of drug–drug interaction risk compared to IC50-shift

Abstract

1. The utility of two abbreviated, higher-throughput assays [IC50-shift and the loss of activity (LOA) assay] to evaluate time-dependent inhibition (TDI) of 24 structurally related compounds was compared.

2. Good correlation (R²  = 0.90) between % inhibition and k_inact/K_I suggested that the LOA assay has utility as an indicator of TDI potential. Weaker correlation was observed for the shifted IC50 (IC50_(T = 30)) (R² = 0.61) and the fold-shift in IC50 (R² = 0.17).

3. Primary mechanism for poor correlation was depletion of active enzyme at concentrations > 1 μM leading to greater than predicted inhibition in the IC50-shift assay.

4. Previously reported strong correlations between IC50_(T = 30) and k_inact/K_I were found to be dependent on potent TDI compounds with k_inact/K_I > 30; correlation was reduced for moderate inhibitors (k_inact/K_I < 30). LOA assay maintained good correlation even when strong TDI compounds were excluded.

5. LOA assay (% Inhibition at 30 min, 10 μM) was a good predictor of in vivo DDI (AUCr), providing a graded response with low potential for false negatives or positives. IC50-shift assay had bias for over-predicting in vivo DDI and was more likely to identify false positives.

• Drug–drug interaction
• IC50
• IC50-shift
• mechanism-based inactivation
• P450
• risk assessment
• time-dependent inhibition

Acknowledgments

The authors thank Larry Wienkers for helpful discussion and comments.

Declaration of interest

The authors report no conflicts of interest.