LB-022 - CAN THE POTENTIAL OF A DRUG TO INHIBIT URIDINE DIPHOSPHATE GLUCURONOSYLTRANSFERASE (UGT) ENZYMES BE PREDICTED? INSIGHTS FROM A RETROSPECTIVE ANALYSIS.

X. Yang¹, D. Volpe¹, V. Arya¹, M. Qusa¹; ¹US Food and Drug Administration, MD, US.

Background: UGTs play an important role in drug metabolism and disposition and have been included in the ICH M12 drug-drug interaction (DDI) guideline. Currently, the guideline does not provide an objective criterion which, in conjunction with in vitro data, can be used to predict the potential for a drug to inhibit UGTs in vivo. Our objective was to determine if the available published in vitro and in vivo DDI data related to UGT inhibition can be leveraged to identify a predictive criterion.

Methods: A literature survey was conducted to collect experimental (IC50 or Ki values) and pharmacokinetic data (AUCR [ratio of UGT substrate in the absence and presence of a UGT inhibitor], inhibitor Cmax) for inhibitor-substrate pairs from drug labels (Drugs@FDA), PubMed, Embase, Micromedex and Google Scholar. Predictive performance analyses were conducted using [Cmax,u]/Ki or [Cmax,u]/IC50 ratios and compared with the in vivo AUCR for each inhibitor-substrate pair. Different AUCR criteria (1.25, 1.5, 2.0) and cutoff values (0.001-0.2) were used in the analyses.

Results: Thirty-six inhibitor-substrate pairs from clinical DDI studies were identified as having corresponding in vitro inhibition information (103 data points). Irinotecan (SN-38, UGT1A1) and zidovudine (UGT2B7) were the most common substrates (47%) in the clinical studies. Two-thirds of the clinical studies had no interaction (defined as AUCR < 1.25 for the purposes of this work) with the remaining third having AUCR>1.25. Using the AUCR criteria of ≥ 1.25, positive and negative predictive values increased from 0.39 to 0.56 and 0.54 to 0.62, respectively, as the cutoff value increased. Using a cutoff value of 0.02, the AUCR criteria of ≥ 1.5 provided a balance of low false negatives (20%) and false positives (22%) and showed better accuracy (0.57), sensitivity (0.45) and specificity (0.65), as compared to AUCR ≥ 1.25 or 2.0.

Conclusion: In comparing the in vivo interaction and in vitro inhibition data for 36 inhibitor-substrate pairs, an AUCR criteria of ≥ 1.5 and [Cmax,u]/Ki or IC50 ratio cutoff value of 0.02 reasonably predicted clinical DDIs from in vitro inhibition data for UGT interaction. Further analyses need to be performed to include DDIs mediated by other UGTs besides UGT1A1 and UGT2B7.