Kidney-On-A-Chip Validated For Drug Toxicity Research
The kidneys are vital organs that filter waste products and drugs from the blood into the urine, help regulate blood pressure, and serve other critical functions. In the U.S., kidney damage from medically prescribed drugs is estimated to affect about 1 in 5 adults, and as many as 2 in 3 elderly adults. In some cases, kidney damage accumulates well before the therapeutic goal of a drug is achieved, which limits treatment regimens.
Historically, animal models have been used to evaluate drug-induced kidney damage but these approaches have limitations due to physiological differences between humans and animals and the inability to track real-time molecular-level drug effects.
Over the past decade or so, organoids (miniature versions of organs grown from cells isolated from humans or animals), and spheroids (a simpler version of these), have piqued interest as promising alternatives to animal experimentation in drug research. Using human organoids or spheroids instead of animal models is advantageous in studying drug toxicity in the kidney as the basic biological mechanisms of damage differ significantly between rodents and humans.
In this study, the researchers took the organoid/spheroid model one step further by embedding human kidney-derived spheroids with microsensors capable of detecting early signs of kidney damage at the molecular level in real-time. They tested their model with the immunosuppressant Cyclosporine A and cisplatin, an anti-cancer drug — both classified as essential medicines by the World Health Organization with known, significant risks for kidney toxicity. They found that both drugs elicited early signs of injury at drug concentrations about 20 times lower than previously recognized as toxic.
Second, the authors grew human-derived kidney spheroids that incorporated endothelial cells and embedded them with microsensors. Here again, it was observed that cellular damage accumulated early, sometimes within a few hours, and at drug concentrations at least 100 times lower than maximum blood concentrations of drug that have been previously measured in human patients, which suggests a reason for the high rate of kidney toxicity in treatment regimens with either of these two widely prescribed drugs. The microsensors also revealed telling changes in metabolism within the drug-treated spheroids, even at nominally non-toxic concentrations: an abnormal accumulation of glucose and increased fat synthesis. Reversing this fat accumulation by treating the spheroids with a drug that prevents the uptake of glucose by the kidney reduced levels of toxicity.
The authors suggest that, while fat accumulation as a consequence of kidney damage has long been recognized and repeated in animal models, their unique findings position glucose accumulation as a key culprit in drug-induced kidney damage. Furthermore, they claim that increased glucose accumulation is a process that would not be seen in rodent models, considering that direct fat uptake (rather than glucose uptake) by the rodent kidney is what underlies fat deposition in the context of those animals. Importantly, if a phenomenon observed within humans cannot be seen in an animal model, therapeutic interventions relevant to it can’t be studied.
To validate these findings, the authors examined archival clinical data of almost 250 human patients who had been diagnosed with kidney damage. Patients who took either drug along with a glucose uptake inhibitor showed a significant decrease in four different markers of kidney damage. These retrospective findings suggest that reevaluation of cyclosporine or cisplatin treatment protocols to include co-treatment with a glucose uptake inhibitor could significantly improve the toxicity profile of these two drugs in human patients.
These findings highlight the potential of human organoid models to revolutionize drug toxicity testing, providing insights that animal models simply can’t. By demonstrating real-time human-specific responses, this study reveals the limitations of animal testing for drug-induced kidney damage. Use of these models not only improves the reliability of toxicity assessments but can also pave the way to reducing animal testing — ultimately promoting a more accurate and ethical approach to drug development and safety.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8897043/