Scientists have restored and preserved some cellular activities and structures in the brains of pigs that had been decapitated for food production four hours before. The researchers saw circulation in major arteries and small blood vessels, metabolism and responsiveness to drugs at the cellular level and even spontaneous synaptic activity in neurons, among other things. The team formulated a unique solution and circulated it through the isolated brains using a network of pumps and filters called BrainEx. The solution was cell-free, did not coagulate and contained a haemoglobin-based oxygen carrier and a wide range of pharmacological agents.
The remarkable study, published in this week’s Nature1, offers the promise of an animal or even human whole-brain model in which many cellular functions are intact. At present, cells from animal and human brains can be sustained in culture for weeks, but only so much can be gleaned from isolated cells. Tissue slices can provide snapshots of local structural organization, yet they are woefully inadequate for questions about function and global connectivity, because much of the 3D structure is lost during tissue preparation2.
The work also raises a host of ethical issues. There was no evidence of any global electrical activity — the kind of higher-order brain functioning associated with consciousness. Nor was there any sign of the capacity to perceive the environment and experience sensations. Even so, because of the possibilities it opens up, the BrainEx study highlights potential limitations in the current regulations for animals used in research.
Most fundamentally, in our view, it throws into question long-standing assumptions about what makes an animal — or a human — alive.
Signs of what?
The pig brains used in the study, which was conducted by a team based largely at Yale School of Medicine in New Haven, Connecticut, produced a flat line on an electroencephalogram (EEG) of brain activity. Had any degree of sentience been recovered, let alone consciousness, one would expect to see low-amplitude waves in the alpha (8–12 Hz) and beta (13–30 Hz) range, at the very least3,4. In consultations with the Neuroethics Working Group of the US National Institutes of Health (NIH) BRAIN Initiative and in discussions with us, the researchers have stated that if they had detected such activity, they would have administered anaesthetic agents to prevent any experience similar to pain or distress, and would have reduced the brain temperature to swiftly quell the activity.
Researchers already study whole organs, and maintain cellular activity for a few seconds to minutes in slices of animal and human brains. Thus, on the face of it, in the absence of EEG activity, the BrainEx study does not raise fundamentally different issues from those encountered in the use of animal or human brain tissue after death.
Yet, until now, neuroscientists and others have assumed two things. First, that neural activity and consciousness are irretrievably lost within seconds to minutes of interrupting blood flow in mammalian brains. Second, that, unless circulation is quickly restored, there is a largely irreversible progression towards cell death and the death of the organism1.
The BrainEx study used pig brains that had received no oxygen, glucose or other nutrients for four hours. As such, it opens up possibilities that were previously unthinkable.
Take the lack of EEG activity. This activity could have been lost irreversibly when the pigs were slaughtered. Another possibility, however, is that the lack of EEG activity was a function of the study design. The researchers used several chemical agents in their solution that inhibit neural activity, hypothesizing that the tissues would be more likely to show some recovery if cellular activity were reduced. Had these blockers been removed at some point, perhaps the team would have detected EEG activity.
Another possibility needing investigation is that something similar to shock treatment for the heart is required to reset the firing of neurons in the brain to a level that is detectable. Or maybe it takes longer than six hours (the length of the BrainEx perfusion, following the four hours after death) for the cells to recover sufficiently for this kind of brain activity to emerge5. Physicians sometimes lower the core body temperatures of people who have had a heart attack, to induce a hypothermic coma. This can limit damage caused by swelling in the brain, for instance, and aid cellular recovery. In these cases, patients seem to need at least 24 hours of ‘cooling treatment’.
Obviously, more data are needed, including the replication of the BrainEx findings in other laboratories by other groups. But we’re reminded of a line from the 1987 film The Princess Bride: “There’s a big difference between mostly dead and all dead. Mostly dead is slightly alive.” Even with all the unknowns, the discovery that mammalian brains can be made to seem ‘slightly alive’, hours after the animals had been killed, has implications that ethicists, regulators and society more broadly must now think through.
To be clear, the BrainEx study did not breach any ethical guidelines for research. The team sought guidance from Yale University’s Institutional Animal Care and Use Committee (IACUC), which exists to ensure that the use of animals aligns with what is required by US law for federally funded research. The committee decided that oversight was unnecessary. The pigs, having been raised as livestock, were exempt from animal welfare laws and were killed before the study started. In the United States, the 1966 Animal Welfare Act is the only federal law that regulates how animals are treated in research, and applies to either living or dead animals. It explicitly excludes animals raised for food. Meanwhile, the policies and regulations of the US Public Health Service, which funds most US research involving animals — mainly through the NIH — do not specify any protections for animals after their death.
Had the research been conducted outside the United States, the response from ethics or regulatory bodies would almost certainly have been the same. The European Union’s Directive on the Protection of Animals Used for Scientific Purposes largely aims to prevent (or minimize) any pain, suffering or distress experienced by live animals. It, too, specifically excludes animals raised for agriculture (see go.nature.com/2cpdgjr). In China, both the Ministry of Science and Technology and the provincial bureaus of science and technology ensure that researchers follow local regulations and that they abide by the National Standard on Laboratory Animal Welfare in China6. Here, too, the protections exclude animals raised for food, and the main focus is on eliminating or reducing live animals’ potential pain and distress.
In our view, new guidelines are needed for studies involving the preservation or restoration of whole brains, because animals used for such research could end up in a grey area — not alive, but not completely dead. Five issues in particular need addressing.
First, how should researchers try to detect signs of consciousness or sentience? On its own, EEG activity would not reliably signal a conscious brain; such activity is nearly always detected in people who are under general anaesthesia7. EEG activity might provide an appropriate measure should it be detected along with responsiveness to transcranial magnetic stimulation (TMS) — a non-invasive way of stimulating brain activity, using a magnetic coil held near the head. Together with other measures, this would determine the brain’s perturbational complexity index, a way of identifying the level of consciousness8. Furthermore, recent research in humans using functional magnetic resonance imaging indicates that certain patterns of neuronal activity may provide a correlate for consciousness9.
Second, which species make appropriate models for this type of research on brain perfusion? And what kinds of research and results would be needed to justify the use of other models? (In our view, investigators should proceed cautiously with testing in other mammals, particularly in pigs, dogs or primates, at this time10.)
Third, until more is known, is the use of neuronal activity blockers sufficient to safeguard against the emergence of capabilities associated with sentience, such as the capacity to feel pain? It might be necessary to apply BrainEx or similar systems to mice or rats, both with and without neuronal activity blockers, to better understand the blockers’ role.
Fourth, under which scenarios should anaesthetics be used in follow-on studies, to safeguard against the possibility of inducing any experience similar to pain or distress? And under what scenarios might it be permissible not to use them? (We think that the use of anaesthetics in follow-on studies should be mandatory at this time, given all of the unknowns.)
Finally, for how long should BrainEx or similar artificial circulatory systems be run? Such systems might be effective for only a certain period of time, or there could be a limit as to how much recovery can be achieved. This knowledge will inform analyses of risks and benefits.
Although it is a long way off, researchers might one day consider using a system similar to BrainEx to treat humans for brain damage caused by a lack of oxygen. Until now, neuroscientists and physicians have assumed that the cell death caused by this is irreversible. Treatment generally involves working with a person’s remaining healthy brain tissue to help rehabilitate mobility, motor and other skills.
Before developing whole human-brain models outside the body — and certainly before the use of brain perfusion in the clinic — investigators need to arm people with enough information for them to make informed decisions. Most fundamentally, patients or donors will need to understand what kinds of brain activity could result and what that activity could mean. They will also need to know the chances of recovery being only partial, and the implications that will have.
Another question is what information, if any, could plausibly be retrieved from the brain. Various groups are developing ways to decode the neural activity of living people, for instance to probe their memories or the images they have seen in their dreams11,12. Could such approaches one day be applied to brains after death?
Such possibilities (if they come to pass at all) are far in the future. Yet we need to think through at least some of them now. Hundreds of people worldwide have already paid to have their brains frozen and stored, in the hope that scientists will one day be able to revive them. It’s easy to imagine misapplications of brain perfusion following the publication of the BrainEx study alone.
It might not be easy for others to replicate the study, despite the BrainEx team providing detailed information on the device, perfusate and methods. As a first step, the investigators, their home institutions and the NIH should facilitate the transfer of the technology and know-how to other researchers and institutions. Any follow-up and independent studies should be just as transparent as this one.
Crucially, future researchers will need guidance through the potential scientific, ethical and political questions opened up by this research.
Precedents exist. Internationally, research involving stem cells derived from human embryos has successfully been steered by the 2005 Guidelines for Human Embryonic Stem Cell Research released by the US Institute of Medicine and US National Research Council — the substance of which was almost entirely adopted by the International Society for Stem Cell Research. Ongoing efforts to set guidelines for human genome-editing research hold lessons, too. Key actors here are the US National Academy of Sciences, the US National Academy of Medicine, the UK Royal Society, the Hong Kong Academy of Sciences, the Chinese Academy of Sciences and the Nuffield Council on Bioethics.
In other contexts, such as in biomedical engineering (see, for example, go.nature.com/2t6kon5), artificial intelligence13 and debates around the definition of death14, international conferences are being held to help find common ground across countries and to develop frameworks that enable responsible scientific progress.
We think that the latest research on brain resuscitation demands the same kind of international attention. A starting point could be the guiding principles issued last December by the Neuroethics Working Group of the NIH BRAIN Initiative, which held a 2018 workshop on research with human neural tissue15.
Citizens must be part of the process. Engaging non-scientists in delineating the ethical boundaries of this research doesn’t guarantee its public acceptance in the future; and nor should it, necessarily. But not engaging other stakeholders could help to precipitate its rejection.
In our view, discussion about the appropriate path for this research should not wait for follow-up studies. The Yale group was conscientious and consulted the local institutional IACUC, Yale bioethicists, NIH programme officers and even the NIH Neuroethics Working Group. The researchers did what they could, and probably more than many would have done, to ensure that they were acting appropriately in a void of ethical analysis on the issue.
Now is the time to fill that void.
Nature 568, 299-302 (2019)