Human Induced Pluripotent Stem Cell-Derived Hepatocyte-Like Cells as a Potential New Tool to Understand Small Molecule Kinase Inhibitors Induced Hepatotoxicity

• FDA

Disclaimer: This article is not an official guidance or policy statement of the US Food and Drug Administration (FDA). No official support or endorsement by the US FDA is intended or should be inferred.

Introduction

Drug-induced liver injury (DILI) remains a persistent problem in medical practice and medical product development, complicated by the multiple mechanisms by which liver injury (hepatotoxicity) can be elicited as well as differences between species.¹ Accordingly, several approaches using human cell-based in vitro models have been explored,^1,2 ranging from cell lines derived from human hepatomas to donor-derived primary human hepatocytes.³ Over the last decade and a half the potential of deriving hepatocyte-like cells from human stem cells has been explored, to the extent that now human induced pluripotent stem cell-derived hepatocyte-like cells (HiPS-HLC) are routinely derived,^4,5 and are being rapidly commercialized. At least five vendors are now available in the US, and some universities begin to provide contract services to help prepare HiPS-HLC starting from human blood samples. These cells possess some characteristics of primary cultured hepatocytes, the current gold standard for drug metabolism and toxicity study under in vitro conditions.^3,6,7 However, detailed characterizations of HiPS-HLC are limited,^7,8 and suggest that in general these cells do not reflect characteristics of primary hepatocytes freshly isolated from adult human livers. Nonetheless, a promising application of HiPS-HLC lies in the screening of compounds for hepatotoxicity,^3,7,9 a bottleneck issue in drug discovery and development. A small number of drugs have been tested for this purpose with some success.7 This review will discuss the utility of HiPS-HLC in the study of small molecule kinase inhibitors (KIs) induced hepatotoxicity.

KIs are a group of relatively new drugs developed mainly for cancer treatment. However, they have much higher risks of hepatotoxicity than drugs in other categories.^10,11 Approximately 20% (6 out of 31) of FDA approved KIs have a black box warning (BBW), the strongest safety warnings issued by the FDA, for hepatotoxicity in the product labels.¹² In contrast, only about 3% (55 out of 1701) of all FDA approved new molecular entities have a BBW for hepatotoxicity.^13,14 Additionally, almost 71% (22 out of 31) of KIs have Warnings and/or Precautions for hepatotoxicity in the label,¹² while this number is only about 49% for 975 oral drugs.¹³ Since many KIs are designed to selectively target kinases that are aberrantly expressed only in tumor tissues, the expectation is that normal healthy cells will be unaffected, with the assumption that KI-induced organ toxicity will be minimal and hepatotoxicity can be overlooked. It is not uncommon that KIs are recommended for life-time use or to be used until unacceptable toxicity occurs. Thus the reality that these drugs can and do present a risk for life-threatening clinical liver injury¹¹ demands that the molecular and cellular mechanisms by which they cause hepatotoxicity be explored and screening approaches developed.

Mitochondrial Injury and KI Hepatocyte Toxicity

Despite their clinical significance, the mechanisms and risk factors for KI hepatotoxicity are poorly understood. We recently found that regorafenib, a newly approved KI with a BBW for hepatotoxicity, caused uncoupling of mitochondrial oxidative phosphorylation and cytochrome c release at clinically relevant concentrations, and these initial events triggered downstream reactions including adenosine triphosphate (ATP) depletion, mitochondrial inner membrane depolarization and permeability transition leading to oncotic necrosis in primary cultured rat hepatocytes.^12,13 The potential drawback of this study is that the human relevance is unknown, as drug hepatotoxicity in general is believed to be poorly predictable by animal models and it is likely that rat hepatocyte data may not have direct clinical significance. We therefore examined regorafenib effects in HiPS-HLC and found that the majority (but not all) of findings in rat hepatocytes can be reproduced. As demonstrated in some previous reports, the capacity of mitochondrial oxidative phosphorylation in HiPS-HLC appeared to be lower than primary human hepatocytes.¹⁵ This might be an underlying reason that HiPS-HLC seemed more sensitive to ATP depletion than primary rat hepatocytes after KI treatment. Interestingly, ATP shortage due to the uncoupling effect of regorafenib promoted the activation of AMP-activated protein kinase (AMPK) mobilizing alternative energy production pathway to sustain cell survival in primary rat hepatocyte,¹⁶ and this was also found to be true in HiPS-HLC after regorafenib treatment. The finding is in line with a recent report showing that AMPK pathway is well preserved in HiPS-HLC.¹⁷ This is of particular importance for the understanding of KI hepatocyte toxicity, as AMPK can be inhibited by 9 out of 17 tested KIs approved by FDA,¹⁸ and AMPK inhibition is a well-documented mechanism for KI organ toxicity.

It is unknown if KI hepatotoxicity can be predicted using various in vitro models that are being highly promoted recently.² Our most recent study showed that mitochondrial liability of 31 FDA approved KIs can help identify hepatotoxic KIs but not non-hepatotoxic KIs.¹² The result casts some doubts on the generalized application of using mitochondrial toxicity to accurately predict drug hepatotoxicity. Our preliminary data showed that the prediction power can be enhanced by using primary rat hepatocytes, likely due to the presence of drug metabolizing enzymes in hepatocytes. It can be anticipated that using HiPS-HLC would further increase the prediction of KI hepatotoxicity, as the hepatocyte toxicity of some KIs appeared to be highly dependent on CYP mediated metabolism. The species difference in CYP mediated metabolism is a well-recognized caveat in extrapolating non-clinical data to clinical settings, and HiPS-HLC can at least partially alleviate this concern. Indeed, it has been demonstrated that HiPS-HLC possessed similar CYP profiles as the primary cultured hepatocytes from the same donor,¹⁹ though this still needs to be confirmed by independent groups, as some investigators found that CYP activities were significantly lower in HiPS-HLC than in primary human hepatocytes.^3,6,15

Drug Metabolism and KI Hepatocyte Toxicity

Among various functions of hepatocytes drug metabolizing enzymes are arguably the best characterized in HiPS-HLC,^3,6 and are a key aspect for the study of KI hepatotoxicity. As shown in Table 1, 90% (28 out of 31) of KIs are mainly metabolized by CYP3A, and CYP2D6 only plays a minor role in metabolizing 4 KIs. This is in sharp contrast to drugs in other groups, as over 50% and 25% of marketed drugs are mainly metabolized by CYP3A4 and CYP2D6, respectively.^20,21 Of note, several KIs are mainly metabolized by non-CYP enzymes that have not been characterized in HiPS-HLC. Interestingly, the metabolites of nine KIs are pharmacologically active, and the metabolites of four KIs do not have pharmacological effects. This should be taken into considerations when studying the contribution of drug metabolism to KI hepatotoxicity.

Table 1. Drug metabolizing enzymes responsible for KI biotransformation

There has been substantial evidence that the hepatotoxicity of some KIs is related to CYP3A mediated metabolism. For example, lapatinib, a KI with a BBW for hepatotoxicity, is metabolized by CYP3A to produce a more toxic metabolite O-debenzylated lapatinib in hepatocytes of human origin,²² and CYP3A inducer dexamethasone indeed enhanced the hepatotoxicity of lapatinib in clinical practice.²³ The blood concentration of lapatinib is nearly 10-fold higher than its metabolite, and it is lapatinib but not the metabolite that exerts the major pharmacological effects. These findings raised the possibility of better monitoring and managing lapatinib hepatotoxicity by measuring and modulating CYP3A activities. One the other hand, we recently found that the hepatocyte toxicity of regorafenib was enhanced when CYP3A was inhibited in primary hepatocytes, indicating that the metabolites of regorafenib are less toxic than the parent drug. Of particular interest is that the two major metabolites of regorafenib are of similar blood concentrations in patients and have similar pharmacological effects as compared to its parent drug. This implies that regorafenib metabolites hold the promise of being developed into new drugs with reduced liver risks. We found that modulating CYP activities in HiPS-HLC can affect the cytotoxicity of these two KIs as expected, indicating that HiPS-HLC is a useful tool to study the contribution of drug metabolism to KI hepatotoxicity.

Table 1 shows that afatinib is barely metabolized, indicating that gene polymorphisms of CYPs are unlikely to affect a patient’s susceptibility to afatinib. Indeed, a recent study found that a patient with UGT1A1 and CYP3A polymorphisms (poor metabolizer phenotypes) developed serious hepatotoxicity after treatment with gefitinib and erlotinib, two KIs mainly metabolized by CYP3A. However, this patient showed no hepatotoxicity when switched to afatinib, the only KI that is not metabolized (Table 1).²⁴ These data indicate that gefitinib and erlotinib are likely detoxified by CYP3A mediated metabolism. However, studies with human microsomes found that gefitinib is metabolized by CYP3A to produce a reactive metabolite that may contribute to gefitinib hepatotoxicity.²⁵ Detailed mechanistic studies with HiPS-HLC may help clarify the role of CYP3A mediated metabolism in gefitinib hepatotoxicity.

Future Directions

Currently only a very limited number of KIs have been studied in hepatocytes of human origin. A comprehensive study involving all FDA approved KIs (currently 31) will significantly enhance the understandings of KI hepatotoxicity. However, primary hepatocytes from a single human donor are usually not enough for testing multiple compounds with multiple endpoints, and pooled human primary hepatocytes make it impossible to examine individual responses. The unlimited supply of HiPS-HLC provides a unique advantage for this purpose. A recent study successfully produced HiPS-HLC from primary human hepatocytes, and found that these HiPS-HLCs retained the key functions of the original primary hepatocytes, including the inter-individual differences in responses to drug treatment.¹⁹ Thus, while our preliminary experience with select KIs was based on HiPS-HLC from a single heathy donor, the possibility exists of exploring the KI responses of HiPS-HLC derived from different donors.⁴ While standardized protocols for generating HiPS-HLC have been published,^4,5 comparing the drug response of HiPS-HLC from various commercial vendors is complicated by the differences in HiPS-HLC derivation as well as the genetic background of the donors. Relevant to that concern, patients with genetic polymorphisms in manganese superoxide dismutase appeared to have a higher likelihood of developing hepatotoxicity after being treated with drugs causing mitochondrial injury or generating toxic metabolites,²⁶ as many KIs do. HiPS-HLCs provide a useful tool for the manipulation of interested genes for detailed mechanistic investigations into such genetic effects¹⁷ and would give insight into the genetic factors critical in determining patient susceptibility to KI hepatotoxicity.

Conclusion

HiPS-HLC can recapitulate the response of primary hepatocytes after KI treatment and may provide additional benefits such as unlimited availability for large scale tests and the possibility of manipulating selective genes for detailed mechanistic study. HiPS-HLC generated from human blood cells deserve to be examined for the prediction of individual susceptibility to KI hepatotoxicity. Despite the challenge that current HiPS-HLC derivation approaches do not yet yield fully mature hepatic phenotypes, these cells offer great promise for advancing our understanding of the cellular basis for clinical hepatotoxicity elicited by KI treatment.

References:

1. Atienzar FA, Blomme EA, Chen M, et al. Key Challenges and Opportunities Associated with the Use of In Vitro Models to Detect Human DILI: Integrated Risk Assessment and Mitigation Plans. Biomed Res Int. 2016;2016:9737920.
2. Sistare FD, Mattes WB, LeCluyse EL. The Promise of New Technologies to Reduce, Refine, or Replace Animal Use while Reducing Risks of Drug Induced Liver Injury in Pharmaceutical Development. ILAR J. 2016;57(2):186-211.
3. Kia R, Sison RL, Heslop J, et al. Stem cell-derived hepatocytes as a predictive model for drug-induced liver injury: are we there yet? Br J Clin Pharmacol. 2013;75(4):885-896.
4. Asplund A, Pradip A, van Giezen M, et al. One Standardized Differentiation Procedure Robustly Generates Homogenous Hepatocyte Cultures Displaying Metabolic Diversity from a Large Panel of Human Pluripotent Stem Cells. Stem Cell Rev. 2016;12(1):90-104.
5. Gieseck RL, 3rd, Vallier L, Hannan NR. Generation of Hepatocytes from Pluripotent Stem Cells for Drug Screening and Developmental Modeling. Methods Mol Biol. 2015;1250:123-142.
6. Goldring C, Antoine DJ, Bonner F, et al. Stem cell-derived models to improve mechanistic understanding and prediction of human drug-induced liver injury. Hepatology. 2017;65(2):710-721.
7. Lu J, Einhorn S, Venkatarangan L, et al. Morphological and Functional Characterization and Assessment of iPSC-Derived Hepatocytes for In Vitro Toxicity Testing. Toxicol Sci. 2015;147(1):39-54.
8. Gao X, Liu Y. A transcriptomic study suggesting human iPSC-derived hepatocytes potentially offer a better in vitro model of hepatotoxicity than most hepatoma cell lines. Cell Biol Toxicol. 2017.
9. Mann DA. Human induced pluripotent stem cell-derived hepatocytes for toxicology testing. Expert Opin Drug Metab Toxicol. 2015;11(1):1-5.
10. Teo YL, Ho HK, Chan A. Formation of reactive metabolites and management of tyrosine kinase inhibitor-induced hepatotoxicity: a literature review. Expert Opin Drug Metab Toxicol. 2015;11(2):231-242.
11. Shah RR, Morganroth J, Shah DR. Hepatotoxicity of tyrosine kinase inhibitors: clinical and regulatory perspectives. Drug Saf. 2013;36(7):491-503.
12. Zhang J, Salminen A, Yang X, et al. Effects of 31 FDA approved small-molecule kinase inhibitors on isolated rat liver mitochondria. Arch Toxicol. 2016.
13. Weng Z, Wang K, Li H, Shi Q. A comprehensive study of the association between drug hepatotoxicity and daily dose, liver metabolism, and lipophilicity using 975 oral medications. Oncotarget. 2015;6(19):17031-17038.
14. Chen M, Vijay V, Shi Q, Liu Z, Fang H, Tong W. FDA-approved drug labeling for the study of drug-induced liver injury. Drug Discov Today. 2011;16(15-16):697-703.
15. Yu Y, Liu H, Ikeda Y, et al. Hepatocyte-like cells differentiated from human induced pluripotent stem cells: relevance to cellular therapies. Stem Cell Res. 2012;9(3):196-207.
16. Weng Z, Luo Y, Yang X, et al. Regorafenib impairs mitochondrial functions, activates AMP-activated protein kinase, induces autophagy, and causes rat hepatocyte necrosis. Toxicology. 2015;327:10-21.
17. Im I, Jang MJ, Park SJ, et al. Mitochondrial Respiratory Defect Causes Dysfunctional Lactate Turnover via AMP-activated Protein Kinase Activation in Human-induced Pluripotent Stem Cell-derived Hepatocytes. J Biol Chem. 2015;290(49):29493-29505.
18. Uitdehaag JC, de Roos JA, van Doornmalen AM, et al. Comparison of the cancer gene targeting and biochemical selectivities of all targeted kinase inhibitors approved for clinical use. PLoS One. 2014;9(3):e92146.
19. Takayama K, Morisaki Y, Kuno S, et al. Prediction of interindividual differences in hepatic functions and drug sensitivity by using human iPS-derived hepatocytes. Proc Natl Acad Sci U S A. 2014;111(47):16772-16777.
20. Sevrioukova IF, Poulos TL. Current Approaches for Investigating and Predicting Cytochrome P450 3A4-Ligand Interactions. Adv Exp Med Biol. 2015;851:83-105.
21. Wang B, Yang LP, Zhang XZ, Huang SQ, Bartlam M, Zhou SF. New insights into the structural characteristics and functional relevance of the human cytochrome P450 2D6 enzyme. Drug Metab Rev. 2009;41(4):573-643.
22. Hardy KD, Wahlin MD, Papageorgiou I, Unadkat JD, Rettie AE, Nelson SD. Studies on the role of metabolic activation in tyrosine kinase inhibitor-dependent hepatotoxicity: induction of CYP3A4 enhances the cytotoxicity of lapatinib in HepaRG cells. Drug Metab Dispos. 2014;42(1):162-171.
23. Teo YL, Saetaew M, Chanthawong S, et al. Effect of CYP3A4 inducer dexamethasone on hepatotoxicity of lapatinib: clinical and in vitro evidence. Breast Cancer Res Treat. 2012;133(2):703-711.
24. Zenke Y, Umemura S, Sugiyama E, et al. Successful treatment with afatinib after grade 3 hepatotoxicity induced by both gefitinib and erlotinib in EGFR mutation-positive non-small cell lung cancer. Lung Cancer. 2016;99:1-3.
25. Li X, Kamenecka TM, Cameron MD. Bioactivation of the epidermal growth factor receptor inhibitor gefitinib: implications for pulmonary and hepatic toxicities. Chem Res Toxicol. 2009;22(10):1736-1742.
26. Lucena MI, Garcia-Martin E, Andrade RJ, et al. Mitochondrial superoxide dismutase and glutathione peroxidase in idiosyncratic drug-induced liver injury. Hepatology. 2010;52(1):303-312.