• 07.28.14

The Coming Human Body On A Chip That Will Change How We Make Drugs

No more animal testing and no more guesswork about whether drugs that work on animals might also work on humans. Scientists are making an entire electronic set of organs that can test our drugs quickly and easily.

Over the last century, animal welfare advocates have protested the rising use of animals in biomedical research and, later, in drug, chemical, and cosmetic testing. One 2008 study estimated 115 million animals are used a year for scientific research alone. Agree with the cause or not, their reasons are well understood to anyone who has felt a little sad about that frog dissection in middle school.


But lately the movement to end animal testing has drawn some more surprising bedfellows. Scientific researchers, regulators like the FDA, and even the military are realizing the current paradigm of using animals as proxies for human beings is broken. Instead, they’re working to develop technologies that could eliminate or drastically reduce the use of animals in favor of more accurate, efficient, and (by default) kinder alternatives.

Humans who want to live healthier and longer lives–i.e. everyone–may eventually be the biggest beneficiaries of the shift in thinking. “If our goal is to create better drugs, in a way that is much more efficient, time and cost-wise, I think it’s almost inevitable that we will have to either minimize or do away with animal testing,” says Dan Tagle, associate director of the NIH’s National Center for Advancing Translational Sciences.

The Startup That Is Creating A ‘Human On A Chip’

Tagle is referring to what’s recognized as a crisis in drug development today: Drug companies have hit a wall in developing new drugs. Some 90% of new drugs fail in human clinical trials based on safety and effectiveness, and it now takes an average of 14 years and often billions of dollars to actually deliver a new drug to the market.

One reason for the high failure rate is that drugs that first seem promising in rodents often don’t have the same response in people. In fact, so-called “animal models” are only typically 30% to 60% predictive of human responses, says Tagle. Then there are the potentially life-saving drug therapies that never make it to human clinical trials because they’re toxic to mice. In these cases, there’s no way to measure the lost opportunity when animals predict the wrong response.

Along with the FDA and the military’s R&D wing, DARPA, Tagle leads an NIH program that’s funding one major effort to develop alternatives to animals in drug testing development. “Organs-on-a-chip” don’t look like much: They are small, flexible pieces of plastic, but when hollow micro-fluidic channels inside them are lined with human cells, they can mimic human systems far more effectively than simple petri dish cell cultures.

“The first program began about five years, ago, with lung-on-a-chip,” says Geraldine Hamilton, a senior staff scientist at Harvard University’s Wyss Institute for Biologically Inspired Engineering, which has led the technology’s development and is launching a new startup company this week to bring it to the commercial market. “We’ve also got the lung, gut, liver and kidney. We’re working on skin. The goal is really to do the whole human body, and then we can fluidically link multiple chips to capture interactions between different organs and eventually recreate a body on a chip.”

Emulate, the new startup that will license Wyss’s technology, isn’t looking to literally create a human body but rather to represent its “essential functions” and develop a platform that’s easy for all scientists and doctors to use, says Hamilton, who will become Emulate’s president and chief scientific officer.


Borrowing microfabrication techniques from the semiconductor industry, each organ-on-a-chip is built with small features, such as channels, vessels, and flexible membranes, designed to recreate the flow and forces that cells experience inside a human body. The structure can mimic the inhalation of, say, an asthma medication into the lungs and, later, how it’s broken down in the liver. It might one day help the military test treatments for biological or chemical weapons (the Defense Department would prefer to see if their treatments work before disaster strikes); hospitals to use a patient’s own stem cells to develop and test “personalized” treatments for their disease (one of Emulate’s investors is the Cedars-Sinai Medical Center in Los Angeles); and, of course, drug companies to more quickly screen promising new drugs.

Mapping The Human Toxome

Advocates like Pascaline Clerc, a senior director at the Humane Society of the United States and a former cellular biologist whose work once involved lab work on animals, view the organ-on-a-chip work at Wyss as among the most promising of several advances aimed at reducing or eventually replacing animal testing.

Yet while quite impressed with the progress so far, Thomas Hartung, director of the Center for Alternatives to Animal Testing at the Johns Hopkins Bloomberg School of Public Health, notes there will be many challenges in getting a “reasonably complete model of a human system” right.

“There’s a certain hype at the moment because a lot of money was made available for this,” he says. “The body is incredibly efficient. It’s very difficult to reproduce this.”

Hartung is leading a project that takes the opposite approach. Rather than trying to reproduce the complexity of the human system, The Human Toxome Project is working to break it down into its simplest elements. Starting with endocrine disruption, the chemicals that interfere with the body’s hormonal system, his team is beginning to model the ways in which the body’s cells process the tens of thousands of chemicals that consumers are exposed to in their daily lives, many of which have never undergone health or toxicity testing. (EPA rarely requires companies to conduct toxicological tests for new chemicals before marketing their products. Out of some 28,000 “pre-marketing notifications,” the agency has requested testing about 200 times, says Hartung. “The are overwhelmed by the task.”)

While it would cost millions of dollars per substance to test each chemical on animals one by one (and the tests wouldn’t necessarily be accurate), Hartung is betting that it will be a more manageable task to understand each and every way a chemical can cause harm in the human body. If scientists can precisely understand each potential “pathway” of action at a molecular level, at some point they might be able to develop simple, quick tests for whether any substance might, for example, be an endocrine disruptor just like BPA, the now infamous chemical used in food containers and packaging.

Can Technologies End Animal Testing?

Neither the human toxome nor the human-on-a-chip project are going to end animal testing any time soon. Before that can happen, they could help change the way drugs and products are designed, enabling R&D projects to fail faster or move more quickly through the product development pipeline. Hamilton believes the economics of tissue chip technology will improve as they are used in industries to design better and safer products and medicines.


Clerc, of the Humane Society, says that to change the entire system, there will have to a much larger shift away from investments in animal models. For example, the FDA would have to stop requiring animal test data before a drug enters human trials–and for it to do that, it would have to be extremely confident that other technologies provide equal or better data. That type of comparison will take time. For example, the NIH’s Tissue Chip program is just beginning to fund a study that will look at a list of about 100 drugs that seemed fine in rodents but eventually failed in human trials. It will see if the new technology would have made a better prediction. Hamilton says that her new startup startup, Emulate, will work with different companies that have good data sets to make this kind of side-by-side testing possible.

Overall, Hartung is “pretty optimistic” that industries can dramatically reduce animal testing in the future. He notes that cell culture only became a standard tool in the 1980s, and now it’s a multi-billion dollar industry serving these markets. Pluripotent stem cell technologies, developed in the last decade, are still pretty new. Eventually, the idea that we should use animal approximations instead of real knowledge about human systems will seem antiquated. “We are learning more and more that mice and rats don’t predict humans. The shortcomings of animal testing are becoming clear,” he says.

About the author

Jessica Leber is a staff editor and writer for Fast Company's Co.Exist. Previously, she was a business reporter for MIT’s Technology Review and an environmental reporter at ClimateWire.