Prediction of Transporter-Based Drug Clearances and Tissue Concentrations: Relative Expression Factor Approach and Protein-Mediated Uptake Effect

Abstract

Clinical drug development often fails due to insufficient efficacy and safety, which can be attributed to not reaching the desired drug concentrations at their target sites (Smietana et al., 2016).Therefore, for successful and cost-effective development of a new drug, it's crucial to accurately measure or predict its concentrations at the site of actions. This becomes especially critical for drugs that are active transporter substrates or undergo metabolism in the target tissue. For such drugs, in vivo pharmacokinetic (PK) studies cannot yield reliable predictions of tissue drug concentrations (G Patilea-Vrana and Unadkat, 2016). Moreover, direct measurement of tissue drug concentrations using imaging techniques is not always feasible (Billington et al., 2019). These challenges highlight the need to accurately predict (not measure) tissue drug concentrations. To achieve this goal, all clearances (CLs) pathways mediating drug’s entry and exit from the tissue, including transporter-mediated and metabolic CLs must be accurately predicted. While metabolic drug CL predictions have been relatively successful, predicting transporter-mediated drug CL, using primary cells, remains challenging. Given this, we have proposed the proteomics-informed relative expression factor (REF) approach to predict transporter-based drug CL. This approach utilizes the transporter-expressing cells (TECs) or vesicles (TEVs), which are more readily available. We have verified this approach by successfully predicting transporter-based CLs and tissue concentrations for several transported drugs (Storelli, Yin, et al., 2022). However, the prediction fell short for the hepatic uptake CL of rosuvastatin (RSV, an organic anion-transporting polypeptide (OATP) substrate) (Kumar et al., 2021). In addition, the underprediction of in vivo hepatic uptake CL has been widely reported for OATP-substrates (Kim et al., 2019). Numerous investigators introduced “protein-mediated uptake effect (PMUE)” and indicating that the inclusion of plasma proteins enhances the apparent intrinsic uptake CL of OATP-substrate drugs, thereby improving the prediction of hepatic CL for these drugs (Schulz et al., 2023). Therefore, the primary goal of my dissertation work is to investigate if the addition of human plasma proteins can bridge the IVIVE gap of in vivo hepatic uptake CL of OATPs substrates, such as statins. To do so, in Chapter 2, I investigated PMUE on five statins (atorvastatin (ATV), cerivastatin (CRV), fluvastatin (FLV), pitavastatin (PTV), rosuvastatin (RSV)) using OATP1B1-expressing and mock HEK293 cells with varying (0, 1%, 2%, and 5%) human serum albumin (HSA) concentrations. However, the results showed that the observed PMUE was largely an artifact of the residual statin-albumin complex remaining with the cells or labware when the uptake experiment is terminated. The residual albumin-statin complex was estimated by quantifying the residual albumin using quantitative targeted proteomics (QTP). This residual statin-albumin complex has not been quantified by others and therefore its presence has been erroneously interpreted as actual drug uptake resulting in the false conclusion of an apparent PMUE on statin uptake. Since human hepatocytes (plated or suspended) are widely used by researchers to measure OATP-mediated uptake of drugs, in Aim 2 (Chapter 3), we investigated if the reported PMUE on statin uptake by human hepatocytes could also be explained by the residual statin-albumin complex. We hypothesized that if this apparent PMUE on human hepatocytes is also an artifact, it should be reduced when suspended hepatocytes are used. This is because when terminating uptake of drugs, the suspended hepatocytes are centrifugated through an oil layer potentially reducing the amount of residual statin-albumin complex remaining with the cells. Indeed, the results showed that the apparent PMUE on statins, when using human hepatocytes, was indeed an artifact and was much reduced when the oil-spin method was used. With such observations in Chapter 2 & 3, we investigated in Aim 3 (Chapter 4) if the underprediction of in vivo hepatic uptake CL was true for all OATP-transported drugs or specific only to RSV, using TECs/TEVs/REF approach. In Chapter 4, we extended the TECs/TEVs/REF approach to predict THE hepatobiliary CLs and hepatic concentrations of two additional OATP-substrates, glyburide (GLB) and pitavastatin (PTV). Then, we verified these predictions using their human PET imaging data. The hepatobiliary CLs of both GLB and PTV were well predicted (i.e. within two-fold of the observed values). In addition, the predicted hepatic concentration-time profiles for GLB and PTV, both fell within 2-fold range. These results, together with our previous successes, indicate that the REF approach can be used with confidence to predict transporter-based drug CLs and tissue concentrations. Such predictions may be useful to inform dose selection in clinical trials to improve successful development of drugs.