Predictive and translational models for renal drug safety evaluation

Piyush Bajaj, Rosalinde Masereeuw, J. Eric McDuffie

Research output: Chapter in Book/Report/Conference proceedingChapterAcademicpeer-review

Abstract

Drug-induced kidney injury (DIKI) is a common cause of morbidity in patients and of attrition in drug development pipelines. Preclinical detection of candidate drugs at risk of causing clinical DIKI is hampered by the poor sensitivity of established in vitro and in vivo models of kidney injury, leading to failures occurring at disproportionately later stages of drug development. While rat and canine models show high specificity in repeated-dose toxicity and safety pharmacology studies for DIKI, their sensitivity remains low, necessitating additional, humanized in silico, in vitro, or in vivo endpoints to avoid late stage clinical attrition. Among the myriad of cell types of the kidney that can be potential targets for xenobiotic insult, the proximal tubule epithelial cells (PTECs) are most commonly associated with DIKI. Drug-induced injury to the renal tubular epithelium is often a function of xenobiotic transport and metabolism, both of which are quickly lost in traditional two-dimensional (2D) static cultures. Here, we highlight how cellular models of PTEC injury have improved in recent years, evolving from simple transporter-overexpressing cell lines cultured in static 2D environments to complex flow-based microphysiological systems in three dimensions. This has enabled better recapitulation of PTEC function in vitro corroborated by the presence of physiological levels of renal transporters and metabolizing enzymes, thus leading to a higher sensitivity to nephrotoxicants as well as the ability to measure qualified acute kidney injury biomarkers. However, for each of these new advanced model systems, thorough validation is critical to establish their relevance/translation to clinical DIKI and understand their appropriate context of use. Machine learning and modeling as supplemental tools for these advanced in vitro models can enhance predictivity further. Drug discovery of the future might thus entail a combination of these in vitro and in silico approaches for improved preclinical prediction of DIKI.
Original languageEnglish
Title of host publicationIdentification and quantification of drugs, metabolites, drug metabolizing enzymes, and transporters
Subtitle of host publicationConcepts, Methods, and Translational Sciences
PublisherElsevier Science
Chapter18
Pages507-534
Number of pages28
Edition2
DOIs
Publication statusPublished - 2020

Keywords

  • Biomarkers
  • Drug-induced kidney injury (DIKI)
  • Microphysiological systems (MPS)
  • Nephrotoxicity
  • Organ-on-a-chip
  • Proximal tubule toxicity

Fingerprint

Dive into the research topics of 'Predictive and translational models for renal drug safety evaluation'. Together they form a unique fingerprint.

Cite this