The longevity “bug”
Early in his medical career, Morten Scheibye-Knudsen spent six months working as a physician in Nuuk, the capital of Greenland. It is a small town of fewer than 20,000 people, but he was enchanted, and almost didn’t return to his native Denmark. We should be grateful that he did listen to the call of anti-aging research, and head back to Copenhagen: he went on to become one of the leading pioneers applying modern artificial intelligence (AI) to the field of longevity medicine.
Scheibye-Knudsen caught the longevity “bug” as a teenager, appalled by watching his grandparents grow frail and die. But it took a couple of decades before he could dedicate himself to the fight against aging. After a brief stint as a firefighter, he did his medical training at Copenhagen University. Then he moved to Baltimore in the USA, spending eight years researching at the National Institute on Aging, a division of the National Institutes of Health (NIA-NIH), and teaching at the Johns Hopkins University. Returning to Copenhagen, he obtained funding to set up his own lab in 2016, investigating how DNA damage causes aging, and how to stop it.
He set up his lab at a fairly young age, and is now a prominent member of the longevity medicine community, the editor of a prestigious journal, and on the boards of substantial investment organisations. He attributes this success to his ability to bridge the gap between clinicians and scientists – groups which are further apart in attitudes and approaches than you might think.
Early use of computational statistics
His credibility on the pure science side rests on his early combination of computational statistics with laboratory experimentation. He was researching diseases which cause accelerated aging, such as progeria, and Werner syndrome, and he had an idea that it could be instructive to use advanced statistical techniques to compare these rare diseases with the symptoms of normal aging. With help from his physicist brother, he successfully applied machine learning algorithms like support vector machines to the data.
This was before the 2012 Big Bang in AI, when the noted computer scientist Geoff Hinton got an algorithm called backpropagation to work, and launched the field of deep learning – a re-branding of the old idea of neural nets. Deep learning has enabled machines to recognise faces, process natural language, and beat humans at Go. It is what has provoked the last decade’s explosion of interest in AI.
More recently, Scheibye-Knudsen has been collaborating closely with Alex Zhavoronkov, the founder of Insilico Labs, who is pioneering the use of deep learning in longevity research. The two hold weekly meetings, and they have screened 15,000 compounds for their ability to induce the repair response to DNA damage. Scheibye-Knudsen’s lab is one of the few around the world which is conducting clinical trials on drugs that are explicitly targeting aging as a disease.
NAD
The NAD molecule (nicotinamide adenine dinucleotide) is an essential catalyst in our bodies’ healthy metabolism, and its availability declines with age. The clinical trials conducted by Scheibye-Knudsen and his team of twenty researchers are designed to assess whether a “precursor” (component) of NAD called NR (nicotinamide riboside) could form the basis of an effective therapy. The lab’s work on this project is assisted by the fact that Scheibye-Knudsen founded and runs Denmark’s largest recruitment organisation for clinical trials.
Clinical trials are the great hope for the longevity community, as they can provide compelling demonstrations of the efficacy of therapies. These demonstrations are essential to persuade clinicians to take seriously the idea that aging is a disease in its own right – a disease which we can eventually cure – and not simply an unfortunate aspect of the human condition which we must grin and bear. Research scientists have been investigating aging since at least the 1930s, and they have some pretty good ideas about the kinds of therapies that could be effective: around 500 molecules have proven effective in animal models. But clinical trials are expensive and hard to initiate, so none of these molecules have yet passed the gold standard test which could persuade clinicians to take them seriously.
TAME
The biggest clinical trial under way today is the six-year TAME (Targeting Aging with Metformin) trial, run by the veteran medic and researcher, Nir Barzilai. TAME is tracking 3,000 people between the ages of 65 and 79, while the Scheibye-Knudsen lab’s trial involves only 80 people. If you’re using morbidity and mortality as the outcome metrics, 80 people is not enough, but if you’re using biomarkers it’s fine. Biomarkers are biological “clocks”, which indicate a person’s biological age, as opposed to merely their chronological age.
Just one or two successful clinical trials, proving definitively that aging can be tackled as a disease in its own right, could be seminal in changing attitudes within the medical profession. Alex Zhavoronkov of Insilico says that longevity science can only proceed as fast as the medics allow it to, so the scientists must convert them.
The metformin trial is studying the impact on longevity of a drug which has been used for decades to treat diabetes. Its side-effects are minimal, and well understood. For this reason, many people in the longevity community use it – including Scheibye-Knudsen. As a doctor, he writes his own prescriptions for it, and he always writes, “treatment against aging” on the paper he hands to the pharmacists. Disappointingly, they rarely even notice.
Europe’s advantage
Given the right data, modern AI could probably identify several other effective treatments against aging among the drugs which have been widely used to treat other conditions. This could present an interesting opportunity for European governments, which are rightly concerned that the development and deployment of sophisticated AI at scale is a virtual duopoly between China and the USA.
European healthcare systems are mostly run by governments, who have a powerful incentive to reduce their cost. This does not apply to the same extent in the US, where insurance companies cover the costs. Drugs which have been widely used for decades are cheap, and healthcare systems are expensive. Reducing the burden on taxpayers of the cost of healthcare for elderly people is not just a desirable outcome; it is increasingly going to become an economic necessity. Europe produces excellent AI researchers, but they are often hired away by the tech giants from China and the US. If European governments prodded their medical establishments into recognising aging as a target disease, and acknowledged AI as a vital part of the solution, they could stimulate the growth of indigenous AI powerhouses.
European healthcare systems have another potential advantage. They are integrated systems – at least in theory – because they are generally owned by the state, even if parts are operated by private operators. This means – again, in theory – that they can obtain and collate more and better data than the more fragmented system in the US. The main bottleneck in healthcare generally, and in longevity in particular, says Scheibye-Knudsen, is the lack of consistent, high-quality data.
This is another area where progress depends on the medical profession getting on board with longevity. In order to gather the necessary data, and sanction its release, a critical mass of doctors must be persuaded that aging is a disease in its own right which can be tackled.
