“A doctor can save maybe a few hundred lives in a lifetime. A researcher can save the whole world.” – Craig Venter, PhD
Contributors: Mark Dittman, Jacob Beran, Conor Good, Tyler Morton, and Ben Sheehey
Large data sets and machine learning seem to be changing every industry (note: machine learning is a branch of computer science where low-level algorithms are applied to discover patterns in the data). Healthcare tends to be a slow adopter of technology. According to Dr. Eric Topol, director of the Scripps Translational Science Institute and renowned cardiologist, we overestimate clinical advances in the short term (1-2 years) and underestimate the advances in the long-term (10+ years). While healthcare in clinical practice may not be changing much at the moment, it will very likely be dramatically different in a decade. Machine learning and the massive amount of medical data collected may not be much of a force at the moment, but what will it look like in ten years?
For a long time, ancestry companies like Ancestry.com have been collecting DNA, mainly in the interest of determining family trees and family history. While most customers focus on the ancestry aspect of their DNA, companies like Calico (California Life Company, a subsidiary of Alphabet) are focused on the large DNA dataset. According to Ancestry’s vice president Ken Chahine, now that Ancestry has acquired 1 million DNA samples they can begin to elucidate DNA factors that lead to longevity in order to see how heritable longevity is2.
Similar DNA companies focused on ancestry are doing similar things with their large datasets, with a lesser focus on longevity. 23andme is a DNA company that is using its large dataset to partner with drug companies to realize new drug targets. 23andme also uses surveys of its consumers to crowdsource research and discover new gene functions
Image Source: University of Iowa
Due to the drastic drop in price of genomic sequencing, it is an interesting time to be in this developing field. Questions abound as to how actionable this data is. Is it even beneficial to be aware of genetic risk factors? Craig Venter, a scientist who played a lead role in the sequencing of the first complete human genome, believes that people should know their genetic information, along with individual benefits and risks. Venter is approaching 70 and openly states that he has no intention of dying soon, which likely explains his motivation to start this company. Venter’s company, Human Longevity Inc., recently launched the somewhat controversial Health Nucleus in San Diego6.
Image Source: Human Longevity Inc.
A day-long visit to Human Longevity Inc.’s Health Nucleus means getting numerous tests done: full genome sequence, full microbiome sequence, metabalome sequence, full-body MRI, EKG, regular biomarkers such as testosterone, etc. And all of this comes at the price of $25,000. The value proposition is potentially catching diseases in their early stage, such as a tumor in its early stage of growth or the beginnings of the onset of diabetes. The field of genomics is relatively nascent, so Human Longevity Inc. proposes that through the Health Nucleus, patients will be the first to know about rapidly developing genomic findings and how it applies to them. The idea is to harness all of the genetic advancements and the massive amount of bioinformatics that they are gathering to allow the consumers to live a longer, and more importantly, healthier life.
Just how actionable is the massive amount of information that you would get from the Health Nucleus? Will you be better off for knowing it? I recently spoke to the lead practicing doctor in the program, Dr. Helen Messier. I asked Dr. Messier how actionable all of the data is and if it amounts to immediate benefits. Dr. Messier explained that certain datasets are more actionable than others. The metabolome is currently better understood and she noted that they were able to see the beginnings of several disease onsets and could reverse them with therapy. Other datasets, such as the microbiome, are not very actionable at this stage. However, even having the baseline data could be beneficial down the road. Say you develop cancer 10 years down the road, when these fields are better developed. Chemotherapy is increasingly known to damage the microbiome. Having had your baseline from when you are healthy, perhaps you could use that data to restore your previous microbiome. Dr. Messier also stated that the clinical service will improve on a yearly basis as the data grows and is better understood. So right now, this testing amounts to a very expensive baseline. Which, if you have the money, may be worth it.
Dr. Messier also noted that while $25,000 out of pocket is expensive by any measure, healthcare costs tend to run high. A simple emergency room procedure could easily add up to a few thousand dollars. There is also the issue of price invisibility—consumers are often unaware of costs when insurance foots the bill. Dr. Messier stated that the business doesn’t profit from clinical care, but in selling the aggregated, anonymized data on the back-end. Were it not for the data, the clinical care costs at the Health Nucleus would be much higher. In essence, your trip to the Health Nucleus is subsidized by the selling of your anonymized data to a variety of companies.
Venter and the Health Nucleus have some strong critics in the field. Most medical doctors and researchers seem to think that while the Health Nucleus has tons of potential with its future dataset, however there isn’t much benefit that patients will derive from the information. According to Robert Green, a clinical geneticist at Harvard University, “what appears at first blush to be medically relevant information could be helpful, could be confusing, or could actually be harmful.”
Human Longevity isn’t alone in its pursuits to utilize big data to conquer aging and disease. Arivale is a Seattle-based company that sequences the microbiome, genome, and uses blood markers to help users get well and stay that way for as long as possible. Arivale is currently in beta stage with openings only for those in the Seattle area1, but they differ from Human Longevity on costs and range of tests. Arivale’s testing is far less comprehensive and comes at a much more affordable price. At the end of the day however, the business plan is still the same.
While there are certainly privacy concerns with all of this medical information, there is a growing need to fix healthcare in our aging and diseased population. GE Healthcare predicts that the global spending on healthcare will be north of 20 billion dollars per year by 20203. The aging population necessitates a delay of disease if the population is going to continue to reach a higher mean lifespan5. It is very likely that the retirement age will increase as populations are barely hitting or are under replacement birth rates. Social security benefits and pension plans may run dry and seniors will be forced to work more years. Hence Health Nucleus’s tagline: “it is not just a long life we are striving for, but a life that is worth living.” Venter is aware that the world may not be able to handle such an old population with a rapidly increasing mean age, hence the strategy to allow them to work longer.
As represented in examples from this article, big data is going to massively change healthcare and our understanding of biology. That being said, it is difficult to discern the risks or benefits at this time. Right now, the majority of healthcare data is siloed in archives anyway4, and it is too early to tell how patients will react to knowing their genetic risks or low levels of a biomarker that is poorly understood. Will they change their lifestyle for the better? Will it lead to better drugs? Or will the best advice continue to be exercise, sleep well, and eat a Mediterranean diet? The game changer will be when machine learning is able to analyze millions of individual datasets and then translate it into actionable, individualized advice. Until then, they are just numbers on a spreadsheet.
- Bishop, Todd. “Genomics Pioneer Lee Hood to Unveil New Startup Arivale, Breaking New Ground in ‘scientific Wellness'” GeekWire. GeekWire, 19 June 2015. Web. 09 Nov. 2015.
- Chen, Caroline. “Google Is Scouring Ancestry.com to Find Out What’s in Your Genes.” com. Bloomberg, 21 July 2015. Web. 09 Nov. 2015.
- “An Economist’s View – How Innovation Can Really Help Healthcare.” GE Healthcare The Pulse. General Electric Company, 03 June 2014. Web. 09 Nov. 2015.
- Marr, Bernard. “How Big Data Is Changing Healthcare.” Forbes. Forbes Magazine, 21 Apr. 2015. Web. 09 Nov. 2015.
- Yew, Lee. “Warning Bell for Developed Countries: Declining Birth Rates.” Forbes. Forbes Magazine, 16 Oct. 2012. Web. 09 Nov. 2015.
- Zimmer, Carl. “Craig Venter’s Health Nucleus Tries to Reshape Medicine.” STAT. N.p., 05 Nov. 2015. Web. 09 Nov. 2015.