P315 Comparison of Digoxin Efflux Inhibition in Caco-2 and MDR1-LLC-PK1 Cell Monolayers

Elke S. Perloff , BD GentestSM Contract Research Services, Woburn, MA
Andrew K Mason , BD GentestSM Contract Research Services, Woburn, MA
Sudarshan Kapadnis , BD GentestSM Contract Research Services, Woburn, MA
Wendy Khun , BD GentestSM Contract Research Services, Woburn, MA
Lisa G Fox , BD GentestSM Contract Research Services, Woburn, MA
David M. Stresser , BD GentestSM Contract Research Services, Woburn, MA
Objective: Assessment of P-gp inhibition in vitro has become a routine aspect of drug development, however, literature data as well as the recent results of the P-gp IC50 Working Group Industry Initiative indicate large variability in IC50 values for the same inhibitor, presumably due to differences in test systems, experimental methodology, and data processing approaches between labs. In this study, we compared the inhibition of digoxin efflux by 16 inhibitors in two different models, Caco-2 and MDR1-LLC-PK1 cell monolayers, using identical experimental methods in the same lab. Methods: Bidirectional permeability of digoxin as a probe substrate for P-gp was determined in 21- to 25-day Caco-2 and 7-day MDR1-LLC-PK1 cell monolayers. Cells were cultured under cell line specific conditions, however, all experimental work was performed using identical methodologies for both cell lines (HBSS buffer pH 7.4 with 10 mM HEPES, 90 min incubation, 5 µM digoxin). Bidirectional transport of digoxin was assessed in the presence or absence of six concentrations of 16 reported P-gp inhibitors. IC50 values were determined based on the decrease in digoxin efflux ratio and decrease in net secretory flux using non-linear regression to a 2-parameter Hill equation in XLfit v5.2. Results: IC50 values for inhibition of digoxin efflux ratio obtained for the 16 inhibitors ranged from 0.34 µM (elacridar in Caco-2) to 36 µM (troglitazone in MDR1-LLC-PK1) and varied up to 3.9-fold between Caco-2 and MDR1-LLC-PK1 cells. Nine out of the 16 inhibitors showed less than 2-fold variation between models. IC50 values for inhibition of net secretory flux showed a similar degree of variability (up to 4.6-fold), but were on average 2.5-fold higher than corresponding IC50 values based on efflux ratios. Baseline activity for positive (25 µM ketoconazole) and negative (no inhibitor) control digoxin transport was reproducible with inter-day variability in Papp values and efflux ratios ranging from 8% to 30% CV. Conclusions: Variability in IC50 values determined using different cell models but identical experimental methodologies in the same lab did not exceed 5-fold for 16 inhibitors tested. These findings suggest that while intrinsic differences between test systems such as different P-gp expressing cell lines do contribute to the variability in IC50 values observed in the literature and in the results from the P-gp IC50 Working Group Industry Initiative, other factors, including differing experimental methodology and technique, as well as alternative data processing approaches likely play a large role in inter-laboratory variability in P-gp inhibition data.