All Articles

Hype in Science

There has been some hype recently in Silicon Valley circles about scientific progress in regenerative technologies. I’m specifically talking about from people like Jack Kirchman (nasjaq), Balaji Srinivasan, and followers.

Example 1. Example 2. Example 3 in the bio

Today, I wanted to actually read the reports that are referenced, and do a fact check. Essentially what I’m going to do it go point by point through some TikTok/YouTube videos that Jack made, and this blog post that Balaji wrote titled The Purpose of Technology, in which he argues that “the ultimate purpose of technology is to eliminate mortality”.

Exhibit 1: “It may be possible to reverse aging”

This is a TikTok video by nasjaq, link here.

In it, he essentially tells the story of Greg Fahy, and his research with growth hormones and thymus regeneration. The basic premise (or really just the entire video script), is pasted below for reference:

“In 1986, Greg Fahy read a paper about rats that were injected with cells that secreted growth hormones. And the researchers did this because as rates age, their thymus shrinks, and since the thymus is involved in the immune system, when it shrinks, rats can’t fight off diseases and they get sick. But injecting them with growth hormones reverses the shrinking and now the rats can fight off diseases. This is important because in humans, the thymus also shrinks with age which decreases our immune system function and this is partially why elderly people get the flu. So years after reading this rat study, in 2003 Greg decided to inject himself with human growth hormone and DHEA (a steroid hormone), and within one month, his thymus increased in size. After this success, he started a company called Intervene Immune and injected 9 volunteers with GH, DHEA, and metformin, and also found that their thymus increased. But the coolest part is that the volunteers epigenetic age decreased by up to 2 years meaning that these people became genetically younger. And now they’re doing a large trial that should be done soon…And I’m really excited because since these drugs are so cheap, this could be the first commercial drug to start reversing aging.”

Ok, so lots to unpack here but in general sounds pretty cool right? Essentially Jack is retelling the story laid out in this Nature news article. Here are links to the larger clinical trial he mentioned, the initial case report in 2003, in which Dr. Fahy injected medicine into himself, the paper where the 9 person trial was published (from which the majority of the claims made in the video are based), and then finally, the website of Intervene Immune which is the company that is supposedly working on commercializing this stuff.

Everything here is true, but the devil is in the details. I’m not going to say much about there not being a control group because this is a very early study, even though that is a pretty big concern (who knows if just the act of injecting medications or taking pills encourages people to engage in habits that eventuate the same results!). To be clear, I’m glad they are doing a second larger trial and I’m not faulting the researchers for trying some high potential experiments. The main issue I have with the video is the total lack of consideration of the limitations and risks of the study, of which there are many. It’s TikTok click bait. “Doctors hate him!” for preteens and Twitter simps.

  1. The video goes from a 9 trial study (10 if we include the investigator’s 2003 self study), to hyping it as the first commercial drug to start reversing aging. Not only does the FDA not recognize aging as an indication (making it impossible to be FDA approved as a therapy), the whole premise of reversing aging is poorly defined. Are we increasing healthspan? lifespan? both? Aging is incredibly difficult to study because by nature it must be longitudinal. In cancer research, we can see if a tumor reappears, but really how are we supposed to know if we are aging less? The study utilizes an epigenetic clock, essentially measuring aging using methylation signatures in key cell types. In all trial participants, their epigenetic age was significantly lower than their chronological age regardless, and even though there was an effect size of 2 years shown, the error of even the best epigenetic aging clock is 3.6 years so there is some degree of uncertainty with the results here. The cells used in this study to do the epigenetic clock stuff were PBMCs, most of which mature or directly interact with the thymus. Thymus regeneration might improve immune function, and the immune system is certainly involved in various other age related diseases, but it seems like a leap to measure the epigenetic age of someone only using PBMCs and claim that you reversed their age overall.

Here is the actual data (epigenetic age over time from a variety of different models) if you are curious. The medicine was halted at month 12, so that is why (probably) there is somewhat of a rebound at month 18. Honestly, I think the data looks pretty compelling, but again: there are caveats!

Epigenetic Age

  1. 30% of trials don’t pass Phase II trials, and even after Phase III confirmatory trials, the success rate is only 50%. These statistics are typically representative of efforts of billions of dollars and millions of hours by seasoned biopharmas in indications with better characterized biology. Aging is similar to ‘cancer’ in that it is a huge scientific landscape that is heterogeneous, and unlikely to be driven by single agent activity (like the Thymus). The trial only included white men age 51-65, which is a pretty sad (as in not diverse) volunteer population.

  2. Side effects. There are reasons why medications are generally restricted or prescribed for use instead of just sold off the shelf. The cocktail of drugs in this study (recombinant human growth hormone (rhGH), DHEA, and metformin) all have side effects and can be actively harmful. In this trial at least, the investigators noted arthralgias (2 cases), anxiety (1 case), carpal tunnel syndrome (1 case), fluid retention (1 case), mild gynecomastia (1 case), and muscle soreness (1 case). The participants also had to do rhGH injections 3-4 times weekly and take pills for an entire year. rhGH can also cause diabetes, which may be why DHEA and metformin were also included (they are anti-diabetic medications).

So while Thymus regeneration could definitely have some effects, it likely isn’t a one stop shop for stopping aging, much less reversing it.

Exhibit 2: Bioelectricity and Limb Regeneration

I actually think this one is pretty cool. It comes from a well regarded researcher (Mike Levin)who didn’t skip the line and try his science in humans, but actually tried to validate the biology from simple model organisms. Here, Jack actually made a 16 minute Youtube video (which I appreciate). Many things are covered in the video, including manipulation of Planarian worms to grow multiple heads, differently shaped morphology resembling other species, regeneration of frog legs, grafting of limbs and eyes to various places on tadpole bodies, stimulation of limb regeneration using an electrical ‘biodome’, and ‘xenobots’. The gist of the video is that bioelectricity (essentially changing the membrane potential of cells), can be utilized to solve many problems, most presciently limb regeneration. By applying electrical potential in a ‘biodome’, regenerative pathways are stimulated, causing an epigenetic change that allows regeneration similar to childhood growth to occur.

Off the bat, huge respect should go out to Mike Levin, who learned about bioelectricity from a used book he bought at a bookstore in 1986 while visiting the world fair in Vancouver when he was a Sophomore in high school. The man went on to study CS and bio at Tufts, where he published 2 papers, founded a software company, and graduated on to work on a PhD at Harvard, where he later joined as an assistant professor at age 30. His work in bioelectricity stemmed from an early fascination, likely before he could have developed the scientific intuition to reject it. Work from his lab has ‘rediscovered’ (for lack of a better word) this field and his success has allowed it to reach pop culture and receive substantial funding. See this PopSci article for more details.

Again, my gripe with Jack isn’t that what he presented is factually untrue, it’s just spun in a way that I think is unproductive. There is no discussion into how any of the experiments are done (relating to feasibility in humans), the ethics of any of it, or anything relating to any sort of limitations.

“If we’ve solved limb regeneration, we’ve solved cancer therapy, we’ve solved brain defects…how soon until we’ve solved aging?”

Really? Ok so clearly it’s dramatized for effect but without any discussion of the reasons why this wouldn’t work or any of the potential limitations of this approach in humans, I don’t like where this is going.

Science has a long history of hyped and tragically failed experiments. These have not only wasted people’s money, but also people’s careers, time, effort, and trust into our scientific processes. Early failures in gene therapy set the field back decades. Sirtris Pharma’s failures have made longevity startups rely on public website fundraisers for cash. Cell therapies almost had their own moment with toxic CRS in small children. Lastly, I will mention but not talk about Theranos. Vinay Prasad and his team found that in the past 15 years of ‘highly promising’ cancer research findings, only 19% ever delivered efficacious therapies, and most delivered marginal benefits with a median benefit in survival of just 2.8 months. The reality is that parachutes are hard to come by in medicine.

“Trust me, I know it seems farfetched, but I’m just extrapolating based on prior research”

I appreciate the acknowledgement that this is very clearly farfetched, but I don’t think we should be trusting people in science. We should be trusting data. Extrapolations are an important part of the scientific process, but should be supported by reason, not just blinding referring to prior research as if the onus to verify the claims should be on the viewer. Otherwise, we’re spreading misinformation.

Strong science should be robust and rigorous and to get there, we should not be rushing or hyping therapy. We should be trying our hardest not to develop emotional connections to any type of idea, especially without any clinical data. Bioelectricity is interesting and deserves further study, but without a well understood mechanism, it is far far away from reaching clinical value.

“I hope that this inspires people to go research this”

One last point about rhetoric. An irritating thing about influencers is the disconnect between their reality and the steps that actually need to get done in order to solve problems. I don’t think it is necessarily bad, but it just has the energy of a ‘that’s your job not mine’ kind of attitude. Scientists work long and hard hours often doing very menial tasks at ridiculously low wages for their intellectual pay grade. I’d hope that along with such goals stated at the end of videos, influencers actually do achieve some good for the people whose work they are piggy backing off of for views. Tim Ferriss does a good job of this.

Exhibit 3: The Purpose of Technology

On to Balaji Srinivasan’s article from July 2020. I’ll preface this by saying that Balaji is definitely a smart guy. He’s cofounded and exited multiple successful companies, graduated with a BS, 2 MS degrees, and a PhD from Stanford, and has been a pretty prolific angel investor. That said, it has been several years of detachment from benchwork, and I disagree on several points in the article.

First, the basic premises of his argument is as follows: The purpose of technology is to reduce scarcity, thus the purpose of technology is to reduce mortality. It is possible to reduce mortality. You were not aware that it was possible to reduce mortality. This is because people with scientific backgrounds have not written about technological progress. We won’t get life extension unless we build a media ecosystem online. We need to start evangelizing technological progress with every word and action.

Essentially what the guy is saying is that scientists need to be hype men/women for their work and that science in general needs more marketing.

On some level, I agree. I think I wrote on one of my college apps that the one question I would ask Elon Musk is how to evangelize technology and get people excited about it. Excitement around technology is important because it encourages favorable public policy and spurs interest in youth, supposedly increasing talent pipelines. These are macro trends that are without a doubt important. Take for example electric cars, space ships, and solar panels, which Elon hyped to great effect with Tesla, SpaceX, and Solar City. These worked because it was possible to rapidly prototype and build. Cars, as much as Tesla was able to improve the underlying battery technology, were already a known commodity. Solar panels and even rocket ships had been built before. The challenge was doing things at scale. In other words, we knew how to build rocket ships, solar panels, and electric cars. We don’t know how to cure ‘cancer’ yet unfortunately.

Elon

For disciplines like the life sciences where there is such an intense focus on developing medicines and therapies, I worry that these macro ‘pushes’ aren’t as useful. What are you realistically hoping to accomplish? People who are interested in science will stay in science and people who aren’t will do other things. Making it sound cool might convince a few more people to give it a shot, but science is hard. How many of those people will stay in the industry? It isn’t like there is a shortage of scientists and engineers given the constraints of funding just yet. Perhaps if you are trying to increase the flow of cash into the life sciences, we may see more investment into budding companies. This may be useful but I have a hard time believing Twitter advocacy convinces LPs to buy bio funds.

Another avenue might be convincing voters to make science funding a priority. The NIH receives a $42 billion budget, which admittedly has seen recent increases (7% in 2020), but is still too small (in my opinion) relative to the number of smart, hardworking scientists and hard problems that need to be solved. I know you really can’t be comparing entire government agencies so simply, but just as a relative sense of scale, the Department of Homeland Security which was established in 2002 as a response to 9/11 (also does important work!), will receive a budget of $54.9 billion in 2021. Again, I think audience matters here, and the reality is that Twitter/TikTok isn’t getting you the most bang for your buck.

Even while funding would be great, I certainly don’t think science needs more marketing. John Mandrola and colleagues make the argument in The Case for Being a Medical Conservative that medical marketing allows propagation of low value care that is more expensive and does not benefit clinical outcomes. I agree, and think that we should be ditching low value scientific endeavors that are conducted to appease investors or further careers. Science doesn’t need guys on Twitter telling them to move faster, it needs more focused and targeted study of reproducible and robust results. If we allow hype to infiltrate science literature and don’t study limitations, we run the risk of lies, bubbles, and wasted resources.

Ok but dude relax they’re TikTok videos

Ok true but these people are completely serious -_-

I think the reason why this bothers me is that I think as scientists, we should be looking for killer experiments (those that would shutdown a project ), and not trying to build up ideas that are destined eventually to fail. Platforms like Twitter and TikTok easily sensationalized and by design, meant for only short term consumption. They’re built upon hype, and I think they do have second order effects even if they are meant for light hearted consumption. There was some discussion of this during Nobel Prize week as well, but in general we just don’t want to be training a generation of youth to do science because of the prestige.

Overall, I think the moral of the story is to stop hyping things. It isn’t necessary and isn’t productive. If you genuinely want to learn more about or encourage discussion of the science, don’t just link the paper, but feature it. Leave it to the scientists who did the experiments to explain what they did and if you do want to extrapolate, do it with a cautious tone with careful discussion with domain experts. And that’s all folks!

Published Dec 26, 2020

Harvard-MIT PhD Student