Can government researchers, working in collaboration with industry and academia, build a better mousetrap for preclinical testing of therapeutic compounds that will be more predictive of their response in humans, without harming the mice? That question is at the heart of an initiative by the NIH's National Center for Advancing Translational Sciences (NCATS).
Nearly at the midpoint of a five-year effort to develop technology that could greatly reduce and perhaps eventually replace the use of animals in preclinical testing, the Tissue Chip for Drug Screening program is set for a pivotal meeting this week that represents the first opportunity for the biopharma industry to offer constructive feedback. So far, the research has yielded an array of tissues-on-chips, including models of the kidney, liver, blood-brain barrier, female reproductive system and cardiovascular system. The eventual goal is to link those and other chips to form a "body-on-a-chip" that researchers can use to test the potential effects of drugs, dietary supplements and other substances across entire body systems before involving human participants.
Many drugs don't appear toxic when tested on an individual organ such as the kidney, observed Danilo Tagle, associate director for special initiatives at NCATS. When the liver breaks down some of those drugs, however, "there could be secondary metabolites that could be toxic to the kidney," he said. "You would not know that if you were just testing kidney cells."
The early years of the Tissue Chip for Drug Screening program encompassed the development of human cell sources for seeding the chip platforms, primarily using patient-specific induced pluripotent stem (iPS) cells. The iPS technology not only represented a broad swath of patients but also allowed NCATS researchers to apply a precision medicine approach to the testing of drug compounds on iPS-directed cell lines.
In the meantime, academic investigators, supported by NIH grants, began to develop the individual organs-on-chips, demonstrating they could incorporate iPS cells from patients and healthy individuals into the microengineered platforms and recreate organ function.
Testing of the individual chips developed to date with compounds known to be safe or toxic in humans has, in most cases, produced the expected results, according to Tagle. The next phase – blinded testing of compounds with known pharmacokinetic and toxicity profiles – is where the rubber will meet the road.
Tagle is optimistic that the effort to develop the three-dimensional (3D) tissue chips, each about the size of a flash drive, will overcome the inherent drawbacks in the use of existing in vitro cell culture, including their two-dimensional nature, homogeneous cell composition and lack of environmental reference in the human body.
Using the 3D bioengineered tissue chips, researchers can view and test multiple layers of heterogeneous cells that contain the proper proportions of each relevant type of cell, Tagle explained. The models could offer enormous advantages to biopharmas. In preclinical toxicology screening of drug candidates, for instance, companies would not be forced to choose between models containing immortal HepG2 cell lines or those containing primary hepatocyte cultures, both with inherent drawbacks.
Because the liver microenvironment includes a dazzling array of biology, including Kupffer, stellate and endothelial cells, the liver-on-a-chip is bioengineered to encompass the same variety, keeping in mind the appropriate cell-to-cell contact as well as liver hemodynamics. Other types of chips are designed to reflect the unique characteristics of different tissue environments, from an organ's perfusion rate to its biomechanical transduction signaling. Lung epithelial cells, for instance, are subjected to stretching to mimic the effect of human breathing.
"Data have shown that if you take lung epithelial cells and grow them in 2D, they don't look like lung epithelial cells, but if you put them in a 3D environment, they look like lung epithelial cells," Tagle said. "If you subject them to selective stretching – biomechanical stress – all of a sudden they also produce surfactants. Those biomechanical cues are important for the cells to recognize their function, so there are a lot of great advantages in terms of 3D culture that we've tried to recreate in the tissue chips."
'WE NEEDED TO MOVE AWAY FROM THE STATUS QUO'
Alternatives to animal testing have realized "enormous progress and uptake over time," said Thomas Hartung, director of the Johns Hopkins Center for Alternatives to Animal Testing and professor of environmental health sciences and molecular microbiology and immunology at Johns Hopkins Bloomberg School of Public Health. Hartung has sought for more than 25 years to improve evidence-based toxicology testing by shifting from animal testing to approaches based on human toxicity pathways.
"We see that even the big agencies are basing more and more of their work on alternative methodologies," Hartung told BioWorld Insight.
The tipping point, he said, came in 2007, when the National Academy of Sciences report, titled "Toxicity Testing in the 21st Century," postulated that the development and validation of new laboratory tools would reduce the need for animal testing, since more precise and predictive tests would be based on human cells and cell components.
"This was the first time that a committee of this caliber was saying that we needed to move away from the status quo," Hartung said. Although discussions didn't immediately ensue about how fast and in what way new techniques should be adopted, "there really was a change in the mindset," he added.
The movement to develop chip technologies that mimic human organ systems is now "very much resonating in the drug development field," Hartung said.
And for good reason. Animal models are less than ideal in predicting outcomes even in other nonhuman animal species – a correlation of only about 60 percent in models of cancer or reproductive toxicity, for example, according to Hartung.
"There's no reason to assume that animals will predict better in humans than they predict in each other," he said. "The 95 percent failure rate in drug development is telling us either that we don't predict side effects well enough to stop development or the prediction of efficacy isn't good."
Then, there's the political and environmental landscape. In 2013, the EU imposed a ban on the sale of cosmetics developed through animal testing, and pressure is building to ban animal use in drug testing as well. Bans on the use of animals in developing cosmetics also are in place in India, Israel and Norway, and multiple parties have advocated that the U.S. and Asian nations adopt a similar stance.
Although the animal rights organization People for the Ethical Treatment of Animals, or PETA, is, perhaps, the most vocal of animal testing opponents, it's hardly the only voice. The Humane Society International is on record as working to decrease and eventually end the use of animals in research and testing, as is the Animal Welfare Institute.
'WE'RE TRYING TO LINK UP DIFFERENT TISSUES ON CHIPS'
To achieve that goal, the Tissue Chip for Drug Screening program will need to replicate selective drug delivery, which is a major factor behind the use of in vivo animal models. NCATS is seeking to meet that objective, and more.
"We're trying to link up different tissues on chips, so whether a drug is orally bioavailable, absorbed through the skin or injected, we can recreate that delivery route and then trace how the drugs are metabolized and how individual organs respond to the drugs and its metabolites so we can assess toxicity in a very specific way, in terms of the cellular response," Tagle said. "Quite often, we test for toxicity in terms of a clinical response and then look for the cellular response. Here is the opportunity to look more from a pathomechanistic viewpoint."
Three-dimensional skin models have advanced the furthest, even to commercial models. The technology has been standardized and has proved its ability to test features such as skin penetration and irritation with greater reliability than animal models.
"They're not cheap," said Johns Hopkins' Hartung, "but they are giving drug companies the assurance that they have reproducible data that's comparable and accepted internationally. This is the direction these things are heading."
Tagle agreed that skin models have become "quite mature and sophisticated, with epidermal and dermal components as well as some vasculature, keratinocytes and hair follicle cells." Still, NCATS researchers are working to improve the models so they can be used not just to test topical agents but also to study intradermal applications.
The NIH also is working with industry to make other organs-on-chips technology widely available, including on a commercial scale, once validation is complete. Tagle expects that process to be complete by year-end 2017.
FDA 'AN EAGER PARTICIPANT' IN TISSUE CHIP PROGRAM
Naturally, drugmakers want to know the FDA's position on alternatives to animal testing. The agency did not respond to interview requests, but Tagle said the FDA was an early and "eager participant" in its tissue chip work, providing researchers with advice on how to create preclinical models that address relevant questions in drug development. For its part, NCATS recognized that if organs-on-chips "were going to generate data that would be used by the FDA to evaluate the toxicity profile and efficacy of a particular drug, we needed to engage the agency," he added. "Pharma will embrace the technology if the FDA blesses it."
In recent years the FDA has shown increased willingness to minimize the use of animals in drug research and development. Last month, the agency updated its own rules designed to minimize the use of animals in developing biologics while still ensuring the safety of those therapies. (See BioWorld Today, July 2, 2015.)
John McManus, CEO of Aeolus Pharmaceuticals Inc., in Mission Viejo, Calif., has some insight into the agency's philosophy based on the company's experience with the FDA's Animal Rule in developing drugs for medical countermeasures. (See BioWorld Today, July 30, 2013, and June 3, 2014.)
"At the end of the day, from our perspective, the gating factor for anything replacing the use of animals – either for preclinical testing or efficacy testing – is the FDA," McManus told BioWorld Insight. "The Animal Rule is very specific about what is required to get an approval. In medical countermeasures, everything comes back to showing a survival advantage in animals."
However, the agency has shown openness to looking at data from cancer patients receiving radiation therapy to support animal efficacy data for the company's medical countermeasure indication, "which is a little bit of a change," McManus added.
He also supports the development of preclinical models that offer more predictability than animals – especially mice – noting that Aeolus and its clinical research organization had to screen 25 to 30 strains of mice just to find a few that developed radiation-induced symptoms similar to those of humans.
"That said, they're still mice, and there are still differences," he acknowledged.
The FDA's willingness to look at non-animal data also may serve as an alternative to serial euthanasia of animals, which has been the standard practice to acquire tissue images and other organ system data across various time frames for a systemic disease like radiation sickness. That could reduce the cost of preclinical testing for biopharmas and reduce the sacrifice of lab animals.
"A lot of these non-animal model systems are very useful," McManus said. "I hope the chip technology can be used to shorten the duration of the studies, which would shorten the development time line and reduce the cost and the risk."
'WE ARE AT A CRITICAL JUNCTURE'
In the meantime, others are working to advance non-animal models. Among them, Emulate Inc., a spinout of Harvard University's Wyss Institute, is partnered with Johnson & Johnson unit Janssen Biotech Inc., which will test several therapeutic candidates in Emulate's Organs-on-Chips technology. The collaboration initially is focusing on R&D programs involving pulmonary thrombosis, liver toxicity and an undisclosed third target, but J&J has the option to extend the partnership into other organs, disease states or drugs. (See BioWorld Today, June 22, 2015.)
Another is Yissum Research Development Co. of the Hebrew University of Jerusalem, which last month said it developed a platform technology to screen and evaluate small molecules as potential drugs for human genetic disorders, including fragile X and Down syndrome. The models, also derived from human cells, can be used to validate potential drug candidates.
NCATS also wants input from drug developers. The tissue-on-a-chip workshop this week will provide industry representatives with updates on more than a dozen models in development along with discussions of functional validation, biomarkers, assays and readouts.
Among the questions the NIH is asking industry to address:
• What are the current obstacles in drug development and where could tissue chip technology play a role?
• What are major concerns or issues impeding the application of tissue chip platforms today?
• What platform features (e.g., chip materials, media, flexibility, connectivity to other platforms, cost, etc.) will be most important to enable use of organ chip systems?
• How will features of basic microanatomy and physiology be demonstrated to influence confidence in the biological relevance of chip platforms?
• What is the most significant and immediate need or gap that tissue chips could address over the next two to three years?
• Which previously used approaches to support new technology/platform/marker development could potentially be employed for tissue chips?
• What are the metrics/key indicators for successful use of tissue chips in toxicity testing in both the short term and long term?
• What FDA activities would facilitate incorporation of new technologies into the regulatory process?
• Is it feasible to submit data from new methods in parallel with existing methods?
• Overall, what outcomes would biopharma and health authority end users need to see demonstrated by these platforms to use them with confidence to support efficacy and/or safety testing?
"We are at a critical juncture," Tagle said. "We're refining the organs. We're testing drugs. Do we have the right readouts? Do we have the right assays? What's the ideal set of compounds to set for single and multiple organ toxicity? We need to engage pharma and the FDA for those answers."
But Hartung hinted at the amount of work that's still needed to prove the models.
"It's pretty naïve to believe that an animal can replace a human" in capturing useful data for drug development, he said. "But it's also naïve to believe that the cell culture of a single organ can replace an entire animal or an entire patient. We are moving at the moment into integrated testing strategies that will include a combination of computer tools and several types of cell cultures combined in a strategic way using a defined algorithm. They're not just a battery of tests."