Right before the Thanksgiving break, while simultaneously attending the SfRBM annual conference (an event featuring lots of hard core wet-bench science grounded in reality), I had the dubious honor of also attending the Longevity Summit. The latter is a new online event featuring talks from leaders in the burgeoning field of longevity research, centered on the new crop of biotechnology firms in this area. For those who want to watch the talks, they’re available here.
Anyone following the aging research field over the past decade or more is probably familiar with the bold claims – human lifespan extension is within our grasp, within some arbitrary time-frame such as 20 years. Famously, such claims have been made by colorful characters such as Aubrey DeGrey (yes, that guy). On the scientific side of things, claims have been repeatedly made for the existence of “longevity genes”, most famously the sirtuins, with Glaxo-Smithkline eventually abandoning their $700m investment in David Sinclair’s company Sirtris once they realized the underlying science was unsound (the exit may have been accelerated by the minor issue of senior personnel selling resveratrol out the back door). I also had fun-and-games uncovering fraud by a senior post-doc’ in the lab of Leonard Guarente, whose lab the sirtuins were discovered in. Throw in a long-standing trend for anti-aging interventions being hawked as dietary supplements, with all manner of polyphenols and other plant-based nutri-ceuticals (resveratrol, quercetin, curcumin, etc.) neatly side-stepping regulation by the FDA, and it’s easy to see how the field of longevity medicine has a reputation for selling “snake oil” based on not very rigorous science. Even such foundational principles as the free radical theory of aging have been largely debunked, and the entire concept that macromolecular damage is an underlying cause of aging has also been criticized. The fact that many aging studies are hugely influenced by survivorship bias is often overlooked, and this leads to an argument that oxidative stress may even be beneficial for aging, because the longest lived organisms have the most of it!
As if the field wasn’t enough of a mess already, things are about to go off the deep end, thanks to the intersection of longevity biotech’ with three other decidedly sketchy things… Artificial Intelligence, Cryptocurrencies, and a Libertarian attitude to regulation…
Before going into detail, I should clarify this blog post is not meant as a complete rip on everyone who came within 100 feet of this event, or the entire field of
aging research geroscience. It’s also not meant as an individual critique on any one company, and specifically is not a direct critique of the company named “Longevity Biotech“, or any specific technology or scientist. Rather, it is a lament about the entire ecosystem of longevity biotech, and how it appears to be a bit “flaky”.
For sure there were a few good talks at the Longevity Summit, and most notably the opening lecture from Charles Brenner had a great take-down on all of the reasons why longevity genes are unlikely to exist (TL/DR – genes only propagate if selected for, and there’s no selective pressure for longevity after reproductive age). There was also a thought-provoking talk from Antonio Tataranni of PepsiCo, about the role the food and beverage industry has to play in making lifespan-extending interventions (if such things exist) more accessible by putting them in food. This idea has some historical precedent (iodine in table salt, vitamin D in milk, fortified breakfast cereals etc.), but would have sounded better coming from a representative of the USDA, rather than someone working for the second largest food & beverage corporation on the planet, which is partly responsible for our current obesity epidemic.
So, there was some good stuff, but the inflection point for me came with the realization that are an absolute shit-ton of new biotech’ start-ups in the longevity field, and as of today (Dec’ 2021) not a single one of them has bought a drug to market! We’re talking billions of dollars of investment, largely based on hype, and so far it’s all just pre-clinical testing or Phase I trials at the very most, with a lot of dietary supplements and other FDA end-runs mixed in. Put simply, despite the bold claims, there really isn’t much to show for all that hype and money. Heck, even down in the trenches of basic model-organism research, there simply isn’t a whole lot of consistency or robustness in the data, with simple things such as mouse strain having huge impacts on the effectiveness of anti-aging candidate drugs.
Despite these problems in the underlying basic science of anti-aging therapy, there has been massive recent growth in the biotechnology sector in this area. Here are just few of the players in this field… Shift Bioscience, Samumed, Gerostate Alpha, Human Longevity Inc., Spring Discovery, Centaura, Fauna Bio, Juvena Therapeutics, Gordian Biotech, GlycanAge, Deep Longevity, ArriveBio, Loyal, BioViva, Calico, Cambrian Biopharma, Vita Therapeutics, Senolytic Thereapeutics, MetroBiotech, Unity Biotech, Senisca, Oisin Biotech, Gray Matter Bio, Siwa Therapeutics, Turn.Bio, Rejuveron, Juvenescence, Rejuvenate Bio, AgeX, Elevian, BioAge, Retro.Bio, Cleara Biotech, Booster Therapeutics, PonceDeLeon Health, Fountain Therapeutics, Juvenon, InsideTracker, and of course Longevity Biotech‘, etc. This is by no means an exhaustive list, and my best guess is there are more than 100 such companies in existence, all selling the idea of extending lifespan or healthy aging.
One company in particular, ResTORbio, is targeting mTOR signaling, but appears to have discovered they can make more money selling
vibrators personal massagers! (Internet Archive link in-case the site disappears).
Many of these companies are also “meta” (no, not FaceBook but the older meaning of the word). Their intent is not to bring a drug to market. Rather, they’re offering services to the other longevity biotech’ companies who might. As an example, one of the big challenges in the aging field is how to measure aging. Clinical trials cannot simply wait until people die to see if an intervention works, so we need measures of “biological age”. There are some good candidates out there such as the epigenetic methylation clock, or various proteomics based clocks, but the usefulness of these clocks in actual clinical trials is yet to be proven (i.e., there has yet to be a demonstration that altering a human aging clock or biomarker actually equates to extending human lifespan). Some recent data has suggested that such clocks simply do not work, but that hasn’t stopped companies from selling such clocks to the public, to track their “real” age, even though there’s no indication of whether that’s actually a useful number. A potential exit path for many of these clock-based companies is to license their product to a biotech company that actually has a drug candidate, but that company may still have a hard time convincing regulators that measurement X is actually meaningful for human lifespan, rather than just an epiphenomenon. The FDA is pretty rigorous about biomarkers used in trials having a connection to outcomes.
It’s also not surprising that there’s a lot of churn in this area of biotech. Many companies I was planning to list here don’t exist any more, or have been bought out or dissolved. Others simply pivot to a specific disease indication as soon as they get some good data, and abandon aging as an indication. Many of them have websites that make it utterly impossible to figure out exactly what they do, or are designed to give the reader epilepsy. There are a lot of dead links out there in aging biotech, if you read news or blog articles from just a couple of years ago. Many of the companies are focused on very rare diseases, another partial end-run around the FDA by seeking orphan drug designation… get a drug to market for something (anything), then use that foot-hold with off-label prescription to get it into a wider population.
Many of these companies are using machine learning (ML) or artificial intelligence (AI) to do such wonderful things as “analyze millions of data points from every level of biological organization, to create an ever-evolving model that captures the full complexity of aging”. That’s great, but AI has a massive Achilles’ heel known as GIGO – garbage in, garbage out. Put simply, there’s a possibility that most of the data being fed in to these models is of low quality. Many of these AI models rely on -omics data obtained from methods such as RNA-Seq. These methods are very expensive, such that most published RNA-Seq data from academia is usually based on a small number of biological replicates from each experimental group (typically N= 3 or 4). It’s widely acknowledged that much of published academic science is complete crap and not reproducible. The pharmaceutical industry has wasted huge amounts of money failing to reproduce basic findings, and large scale reproducibility studies have been undertaken without very good outcomes. Now take that and dial up the risk with low-N ‘omics data.
There’s a lot of focus in aging research on the Yamanaka factors – four transcription factors that can reprogram cells to pluripotency. Many of the companies are applying “deep learning” to interrogate this, and if you can understand why a model with only 4 variables requires AI or ML to decipher, well I guess good for you. And of course, there are companies using AI/ML to build ever more complex aging clocks which is meta on meta.
Computer modeling in biology can be useful in some areas where the “rules” are well understood, but elsewhere is fraught with problems. For example in metabolic modeling, many of the constants fed into models (kM, Vmax etc) are from decades old literature spread across multiple cell types and tissues. You can’t build a model of the Krebs’ cycle by mixing values from pigeon heart mitochondria and mouse skeletal muscle mitochondria, then test it in HeLa cells. AI-assisted Drug Design does not have a particularly stellar track record, and applications of AI in the life sciences in general are not very reproducible. AI also has huge problems of bias – most famously racism. Train a facial recognition AI model to spot criminals by feeding it pictures of incarcerated persons, who because of systemic racism are overwhelmingly black, and you don’t get a criminal-spotting algorithm, you get a black-spotting algorithm. The American Society of Mathematicians has even gone so far as to call for a boycott on their members working with law enforcement agencies to develop crime prediction tools. As such, I am skeptical that many of the AI models being built in longevity science, based on mice or cell data, will be useful in humans.
Put simply, it is hard to decipher exactly what AI is being used for in longevity biotech, because the usefulness of AI for anything is still a bit of a mystery, so when you apply it to a field with fairly weak foundational science, the problems multiply. Despite AI being a good way to attract money, biology is not digital, so a pure bioinformatics driven approach to “solving” it does not seem deliverable.
There’s a robust and growing group of venture capital and angel investors willing to fund the “healthy aging revolution”, and just last month a new Longevity Biotech Association was incorporated to promote investment (with many existing Longevity Biotech CEOs on the board). Key players are the Longevity Vision Fund, Apollo Health Ventures, Kizoo, Forever-Healthy, R42 Group, and others. Looking through the websites of these funds reveals hundreds of start-ups, all pulling down tens of millions in funding each, and it’s not hard to estimate this is multi-billion dollar enterprise overall. There are even bizarre longevity biotech online communities, where for the bargain price of $3k a year you can get access to influencers and leading minds in the field (I think I puked a bit in my mouth when I read that site).
A rather scary development in this area is the emergence of cryptocurrency as a source of funding (Bitcoin, Ethereum, Doge, etc). Some of the companies on the list above have come into existence due to the availability of large amounts of money that originated in trading cryptocurrencies. For example, Gordian Biotech has an “Impetus” longevity grant program that is funded by a group of crypto investors. One of the Longevity Summit talks focused on crypto funding for longevity biotech and research. The longevity biotech field also has a lot of overlap with key players in the LifeBoat Foundation, which appears to be a catch-all conspiracy theory website (alien invasions, asteroids, bunkers), with money coming from BitCoin. There are several other strange websites in this area such as Longevity Technology, which seems to be a curated website to bring investors and biotech founders together, with a mix of blog articles and bizarre product placement reviews.
I won’t go into the multitude of problems with Cryptocurrencies, from pump-and-dump schemes to their horrendous carbon footprint. I will simply note that the IRS seems to have taken an interest recently. Once the IRS starts taxing crypto the same way as real money, the bubble will likely contract and many investments may be worthless. I would also hazard a guess that at least some of the companies listed above are actually using some of their seed money to trade crypto as a source of revenue. Hey, if it’s good enough for Tesla then why not? Wouldn’t it be ironic if the people living to 150 years of age were the same ones whose bitcoin investments overheated the very planet they have to spend their extended lifespans on?
As evidenced by the final talk of the Longevity Summit, featuring Matt Kaeberlin (the only speaker who doesn’t have a company!) and Elizabeth Parrish (CEO of BioViva), there’s a strong libertarian “government needs to get out of the way” thread running through the longevity industry. While not calling for the outright abolition of the FDA, it was scary that Parrish essentially argued “people are going overseas for these therapeutics, so because of medical tourism we need to fix the regulatory process so they can get those therapies here”. I was impressed that Kaeberlin kept a straight face! Arguing that China (where prisoners are used for clinical trials) or places where governments just don’t care about safety are approving therapies, is not the home run you think it is. Most people are familiar with the reasons why large pharmaceutical companies run trials in poor countries – it’s because they can get away with shit that wouldn’t fly at home (interesting side note – my father worked for a pharma company in Ghana in the 1960s).
We’ve already seen the impact of weakening regulations with the right to try movement, and the disastrous approval of Alzheimers drugs that simply do not work. Some companies are taking an interesting approach – making an end-run around the FDA by trying to get interventions approved for pets such as dogs – with Parrish drawing parallels between the strength of regulations in veterinary and human medicine. As most recently demonstrated by the ivermectin horse-paste debacle, I’m less enthusiastic about dog medicine being the fountain of human youth. As a simple example, the product made by Juvenon is toxic to cats.
The notion that somehow aging is “special” and therefore shouldn’t be regulated like “normal” diseases is not convincing. The same argument could be made about any number of other conditions. For example cancer – for years we’ve been told it’s not one disease but hundreds of diseases, and therefore we need to think about it differently. And yet, cancer medications including personalized therapies such as CAR-T seem perfectly capable of getting approved within the existing framework of the FDA. Reminding people that there’s an entire NIH institute devoted to aging, brings responses such as “the A in NIA just stands for Alzheimers”, which of course immediately overlooks that AD (and Parkinsons and Huntingtons) are all leading causes of age related morbidity.
Overall, it’s just very hard to divorce the whole notion of “aging is different” from “we don’t want to deal with the same laws as other medicine makers”. The case would be a lot more convincing if the medicines actually existed, which they currently don’t. This is known as jumping the gun. If anti-aging therapies would come close to approaching their clinical trial end points, then maybe we would have the basis to discuss an accelerated approval process. Until then, we should not be making special dispensations. We need robust regulation, especially for treatments that people are likely to ingest for several decades of their lives.
Other red flags
(2) A LOT of the science behind the longevity biotech industry is coming from a small number of laboratories concentrated in the San Francisco Bay area, with several biotech founders being affiliated with a few of labs in just a couple of institutions, and these same labs also birthing many of those in the longevity funding VC space. The conflicts of interest created by the blurring of corporate/academic boundaries are troubling. The heads of institutes and the boards of the 100 or so companies, have a lot of overlap. An inordinate number of people in the field are white and young and very good at TeD style presentations.There are a lot of tech-bro’s in this area, with some undoubtedly jacked-up on nootropics. The lack of adequate external oversight created by such an in-bred ecosystem (where the CEOs and lab-scientists and VC funders and influencers and coders are all drawn the same small group of people and often wear multiple hats) again points toward a bubble.
In summary, aging science was already in a bit of a messy state with not a great reputation before biotech’ came along. Now the triple bubble of crypto-currencies, AI-hype and lax regulation, are threatening to make everything a whole lot more sketchy. While extremely expensive and niche options such as plasma therapy (quite literally feasting on the blood of the young) will be available to the ultra-wealthy, personally I don’t believe that – significant human lifespan/healthspan extension with cheap small molecule drugs – will be achieved within a reasonable timeframe, before most of these companies run out of VC money. I remain open to being surprised.