A Clinical Practice Perspective on the Evaluation of OATP1B1‐associated Drug Interactions

The liver uptake transporter OATP1B1 has been recognized by drug agencies as a determinant of drug–drug interactions. Nevertheless, labeling of OATP1B1 interactions appears limited considering available in vitro and in vivo data. This makes it difficult to evaluate OATP1B1-associated interactions in practice. Furthermore, genetic predisposition in SLCO1B1 (OATP1B1) affects systemic drug exposure and therefore tolerability, particularly in the case of statins. The evaluation of drug–drug–gene interactions in statin therapy may therefore be of relevance. Transmembrane transport plays an essential role in the pharmacokinetics (PK) of drugs, as it affects absorption, distribution, and excretion processes. One transporter that facilitates the hepatocellular uptake of both endogenous and xenobiotic substrates is the organic anion transporting polypeptide 1B1 (OATP1B1, encoded by the gene SLCO1B1). The transport of statins into the liver, for example, is regulated by OATP1B1.1 It is known that individuals with a genetic predisposition to impaired function of OATP1B1 have an increased systemic exposure under treatment with statins. This increases the risk of dose-dependent adverse drug reactions such as statin-associated musculoskeletal symptoms (SAMS). Today, we can consider pharmacogenetic dosing guidelines to mitigate the SAMS risk in clinical practice.2 Nevertheless, in our experience, there are patients who are affected by SAMS, even if the aforementioned dosing guidelines are followed. One reason for this could be drug–drug interactions (DDIs). Patients taking statins are often concurrently treated with several other drugs (polypharmacy), due to diseases such as cardiovascular conditions and diabetes. Already in 2008, the International Transporter Consortium (ITC) has acknowledged the clinical relevance of OATP1B1 in DDIs. Since then, the ITC has published several white papers to inform and advise on the consideration of OATP1B1 in the drug development and authorization process (e.g., refs 3, 4). In the meantime, based on the aforementioned ITC recommendations, drug authorities such as the U.S. Food and Drug Administration (FDA), the Japanese Pharmaceuticals and Medical Devices Agency (PMDA), and the European Medicines Agency (EMA) have issued corresponding guidance for drug interaction studies on OATP1B1. It is currently advised to assess OATP1B1 transport in vitro for drugs with a substantial hepatic elimination (≥ 25%).5-7 Further clinical DDI studies are warranted if the in vitro assay cannot exclude an in vivo inhibition of OATP1B1 based on Ki and IC50 (e.g., Ki ≤ 25*Iu,inlet,max; R ≥ 1.1).5-7 The Drug Interaction Database DIDB® (Copyright Certara USA, formerly University of Washington, www.druginteractionsolutions.org) compiles datasets of human in vitro and in vivo drug interaction studies, including drug authorization studies. However, since the use of such a database can be complex and time-consuming, as well as the service being subject to a charge, it is rarely used by healthcare professionals for the evaluation of DDIs in clinical practice. In practice, healthcare professionals rather consult the summary of product characteristics (SmPC) when advising patients and in particular when evaluating DDIs. Therefore, we systematically analyzed the information on OATP1B1 in the Swiss SmPCs as well as the available data in the DIDB® indicating OATP1B1 affinity (Figure 1) of drugs authorized in Switzerland. To search for information concerning OATP1B1 in Swiss SmPCs we applied natural language processing (NLP) with 70 predefined search terms using the application AmiKoWeb® (ywesee GmbH, https://amiko.oddb.org). To search for drug substances with OATP1B1 affinity we analyzed the DIDB® based on predefined PK threshold values (Km ≤ 10 μM, IC50 ≤ 10 μM, AUCR ≥ 1.25). Finally, we compared the overlap of information for the identified in vivo OATP1B1 substrates and/or inhibitors between the two sources (Figure 2). At this point, it seems noteworthy, that the Swiss drug agency Swissmedic refers to the recommendations of the EMA and the FDA for their own guidelines concerning drug interaction studies. The comparison of the available data in the DIDB® with the information in the Swiss SmPCs shows that the overlap of designated OATP1B1 substrates and inhibitors is low (Figure 2), which implies that only limited information gathered on the interaction with OATP1B1 is transferred into the SmPCs. As an illustration, compiled in the DIDB® there is clinical DDI data of sacubitril with the OATP1B1 substrate pemafibrate, indicating OATP1B1 inhibition (AUCR ≥ 1.25). However, when consulting the respective Swiss SmPC of the marketed combination product with Valsartan (Entresto©), there is no information on the in vivo OATP1B1 inhibition potential of sacubitril. Taken together, this finding suggests that as a clinician it is challenging to evaluate the currently used drugs for their OATP1B1 interacting properties based on information from the SmPCs in daily practice. Statins, which are frequently prescribed and well-known OATP1B1 substrates, are a long-term therapy for the permanent reduction of cholesterol levels and thus for the prevention of cardiovascular diseases. In Switzerland alone, around 11% of the population have been treated with a statin in 2019.8 However, adverse effects such as SAMS jeopardize patients' medication adherence and frequently lead to therapy discontinuation. An analysis of first-time statin users registered with the Swiss health insurer Helsana (n = 16,668) found that within 18 months of treatment start, 11% switched to a different statin, 2% switched to a different lipid-lowering substance class and 23% discontinued the lipid-lowering therapy completely.8 At this point, it seems plausible that, in addition to genetic predisposition (SLCO1B1), also non-genetic factors such as DDI can affect statin tolerability and may therefore need to be cumulatively evaluated (drug–drug–gene interactions) to optimize lipid-lowering therapy. Considering our data from the DIDB® and the Swiss SmPCs, there are a number of patient populations at risk for OATP1B1-involving drug–drug–gene interactions (DDGI) in statin therapy. The identified drug substances interacting with OATP1B1 include: antivirals, antibiotics, antimycotics, immunosuppressants, antineoplastics, antithrombotics, and antihypertensives. In this context, it seems reasonable to pay particular attention to patients who are treated with a combination of statins and antithrombotics, and/or antihypertensives, as these combinations should be relatively common based on the indications for these drug classes. The European Society of Cardiology (ESC) and the European Atherosclerosis Society (EAS) recommend statins as first-line therapy for the primary prevention of artherosclerotic and cardiovascular diseases (ASCVD) in high-risk patients as well as in secondary prevention of ASCVD.9 Patients with a Human Immunodeficiency Virus (HIV) infection were found to have an increased risk for ASCVD and additionally, certain antiretroviral drugs have been associated with dyslipidemia.10 Therefore, the ESC/EAS guideline recommends to consider statin therapy for HIV patients with dyslipidemia.9 Notably, in our analysis, we found multiple antiretroviral substances associated with OATP1B1 inhibition (Table S1). For most of them, their SmPC mentions interactions with statins without mentioning the involved mechanism but provides recommendation concerning statin selection and dosing. However, in addition to drug–drug interactions (DDI), specific attention should be paid to the patients' genetic predisposition, considering that SLCO1B1 is one of the genes recommended for testing in statin therapy to prevent SAMS,2 whereby rendering the evaluation of such a drug combination, to the next level of a drug–drug–gene interaction (DDGI). Taken together, it remains challenging to evaluate OATP1B1-associated drug interactions in clinical practice. Moreover, in view of potential DDGI, it is still not entirely clear how a combination of genetic information and drug interaction data should be considered in the evaluation of SAMS risk. In addition to these pharmacokinetic concerns, we may need to consider, that a decreased function of OATP1B1 could also affect the efficacy of statins in lowering cholesterol, although this is currently not well studied. Overall, our analysis outlines incomplete information transfer and perhaps also indicates missing data concerning OATP1B1 interaction studies. Therefore, we want to highlight the need for further initiatives to collect data on drug–drug–gene interactions, particularly in the context of statin therapy. As afore indicated, this would be particularly important for patients with cardiovascular risk factors under polypharmacy, which generally represents an elderly population. For such an endeavor, data from a large patient population is needed, which will only be feasible with continuous and structured digitalization of patient data. In Switzerland, for example, a national electronic patient record is only just about to be implemented. No funding was received for this work. Open access funding provided by Universitat Basel. The authors declared no competing interests for this work. Table S1 Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.