Vision of 21st Century Personalized Medicine
Lee Hood, MD, PhD
Co-founder, Chief Strategy Officer and Professor, ISB Senior Vice President and Chief Science Officer, Providence St. Joseph Health
The Beginnings and Biological Complexity
My youngest brother was born with Down Syndrome in 1944 and that initiated in me a life-long interest in human biology (we did not even know the identity of the genetic material at that time). After a BS at Caltech, an MD at Johns Hopkins and a three-year tour as a Senior Investigator at NIH, I began my academic career at Caltech in 1970 as a young assistant professor. I was convinced that the assays and strategies that we had for assessing human health did not begin to touch the complexity of an individual human – and that, if we were really going to improve human health, we would need better measurement technologies, better research strategies, and a new approach to assessing and conceptualizing human complexity. This conceptualization of complexity later came to be termed systems biology. I subsequently participated in seven paradigm changes that dealt with human complexity and I summarize them here because they framed my thinking about health care.
Seven Paradigm Changes that Framed Systems-Driven Medicine
My initial attention turned to the first paradigm change – bringing engineering and technology to biology. During my years at Caltech, my lab developed four instruments – the DNA and protein sequencers and the DNA and protein synthesizers – which became the technological foundation of modern biology. Later, at the University of Washington, we developed the ink-jet DNA synthesizer for rapid DNA array and oligonucleotide synthesis, and still later, at the Institute for Systems Biology, we pioneered the Nanostring single RNA molecule detector. These instruments brought to biology the ability to make rapid measurements on humans and other organisms, initiating the collection of data at scale. These capabilities make possible today’s biological big data and its associated analytics.
The automated DNA sequencer got me invited to the first-ever meeting on the human genome project, the second paradigm change, in the spring of 1985. Twelve scientists passed judgment on the human genome project and came to two conclusions: 1) the project was feasible, although technically challenging at that time; and 2) we were split 6-6 on whether it was a good idea, and this reflected the initial opposition to the human genome project in the biological community. Indeed, NIH was initially strongly opposed. The human genome was sequenced and these data, with improved sequencing technologies, made accessible human genetic variation and the ability to correlate it with wellness and disease phenotypes – a process that is continuing today.
The automated DNA sequencer required bringing together engineering, computer science, chemistry and molecular biology to develop this technology. I realized from this organizational experience that leading-edge biology should always be practiced in a cross-disciplinary environment – where biology can drive the development of relevant technologies and these in turn can catalyze the development of the necessary tools to analyze the data. This was the third paradigm change. Bill Gates made it possible for me to move to the University of Washington in 1992 to set up the first cross-disciplinary biology department – Molecular Biotechnology – which was very successful in pioneering tools in genomics, proteomics, cell biology and single-molecule analyses.
I resigned from the University of Washington in 2000 to initiate the fourth paradigm change by starting the Institute for Systems Biology (ISB), the first-ever research organization devoted to the discipline of systems strategies. A systems approach to biological complexity is holistic and comprehensive, and it requires thinking about the entire biological system, and not just individual genes or proteins (the then-classical reductionist approach). We quickly applied this holistic approach to a variety of biological and disease-related problems and, in the course of doing so, defined two of the central concepts for 21st century medicine – the fifth conceptual paradigm change.
Around 2001, we began applying systems approaches to disease and termed this systems medicine. Systems medicine catalyzed the concept of dense phenotyping on individual humans (e.g. making lots of measurements); pioneered the biological network view of biology and the disease-perturbed network view of disease; and catalyzed the emergence of many technologies and systems-driven strategies (putting together one or more technologies with a computational platform to enable high-throughput biological measurements – as done in dense phenotyping of humans). A few years later, systems thinking led us to the realization that health care should be predictive, preventive, personalized and participatory (from the patient’s viewpoint) – and we termed this P4 medicine or P4 health care. These conceptual paradigm changes in thinking about contemporary health care quickly led to the realization that health care should have two major domains – wellness and disease – and that the former was essentially ignored by most healthcare systems at that point in time.
From 2007-2013, we received a $100 million grant from the Grand Duchy of Luxembourg to invent the tools and strategies of systems medicine. This successful program placed us at a tipping point in 2014 where we could, for the first time, think about a pilot program in quantitative or scientific wellness – the sixth paradigm change. We persuaded 108 individuals to undergo a 9-month-long program of dense, longitudinal phenotyping and the results were so successful from the viewpoint of personal health that we started the company Arivale in mid-2015 to bring scientific wellness to consumers and continue the program for many of the original 108 “pioneers”.
The seventh paradigm change was initiated when Rod Hochman, CEO of Providence St. Joseph Health (PSJH), asked me to consider becoming the Chief Science Officer of PSJH and having ISB affiliate to become a research arm of PSJH in 2016. I accepted, as this was clearly the path for bringing scientific wellness and P4 healthcare to the U.S. health care system. The key point here is that for any paradigm change to occur in an organization, it is essential that the CEO and, indeed, the C-suite support this change.
Scientific Wellness Pilot Program and Arivale
In 2014, 108 individuals (we termed them pioneers) agreed to a nine-month program of longitudinal, dense phenotyping – whole genome sequence analyses, 1,200 blood analytes quantified every three months, quantification of bacterial species in the gut microbiome every three months, and digital health measurements with the Fitbit. All told, this resulted in billions of data points for each individual. When the individual data clouds were properly analyzed and integrated they led to a unique list of actionable possibilities for each individual that, if acted upon, would improve wellness or let one avoid disease. (The original hundred or so actionable possibilities were initially obtained from the literature by looking at the variability of genomes and the changing levels of expression of analytes and microbial species. Later, as we integrated together different types of data, the number of actionable possibilities increased strikingly and now includes multi hundreds). Coaches brought several actionable possibilities to the individuals once a month and played a vital role in education, compliance (70 percent) and longitudinal recruitment. Most of the pioneers experienced striking improvements in wellness – and their enthusiasm to continue the wellness program led to the creation of Arivale – a company bringing scientific wellness to consumers.
Arivale now has around 5,000 clients (each with their own data cloud); it practices in all U.S. states but one; it has seen more that 100 wellness-to-disease transitions for all types of common chronic diseases (to be discussed later); and ISB and Arivale are, in partnership, analyzing these data.
Analyses of the Data Clouds
The Hubble Telescope gave an unparalleled, high-resolution view of the universe that has led to many new insights. So, too, the 100 – now thousands – of individual data clouds have given an unparalleled high-resolution view of human biology and human disease. I will mention four striking observations that have come from the analyses of these data.
3,600 Statistical Correlations
When one divides all of the data from the pioneer 108 project into their six data types, one may ask whether any one data bit within a given type correlates with data bits in the five other data types – and about 3,600 such correlations have been observed. These tie together specific systems that often we did not know were correlated. These statistical correlations can be broken down into about 70 communities of related analytes – and each of these communities correlates with a physiological function or a disease pathology. Parallel observations have been made on the Arivale data as they are much more extensive in scale and they confirm and expand the earlier generalization from 108 individuals. These data enable new approaches to identifying biomarkers, drug targets and even biological mechanisms as the analyses can be done across a population at one point in time or can be analyzed dynamically by comparing the population at different points in time (e.g. before and after scientific wellness, or before and after a disease transition).
Determining Genetic Risk for Individuals for more than 100 Diseases
With the complete genome sequence and genome-wide association studies for more than 100 diseases, we can compute relative genetic risk for more than 100 different genetic diseases and genetic conditions. We have done so with data from thousands of people now, including the 108 pioneers and 1000s of participating Arivale clients. This approach enables one to use the dense, longitudinal phenotyping of scientific wellness to identify the manifestations of genetic risk in the body. In this way, we can learn what (modifiable) systems may contribute to the mechanisms of disease risk, providing clues for potentially dealing with the disease onset early and eventually even reducing genetic risks.
Healthy Aging and a Biological Age Metric
We can use the divergence in the control of the expression of blood analytes as a population ages to determine for each individual a biological age (the age your body says you are rather that the age your birthday says you are). Accordingly, if you are younger than your chronological age, you are healthier than if you are older. I, for example, am 15 years younger in biological age than my chronological age. In the Arivale population those who have diabetic and cardiovascular conditions have strikingly higher biological ages than do normal individuals (as expected). Those who are in the upper five percent of Fitbit-measured exercise have strikingly lower biological ages than those in the lower five percent. This biological age is a metric that each of us can use to optimize our ability to age in a healthy manner.
As noted, we have observed about 100 wellness-to-disease transitions for our >4,000 Arivale clients. We can use the data clouds to identify changes in concentrations of blood analytes (biomarkers) that mark the transition. We can use these perturbed analytes to identify disease-perturbed networks and, in time, we will be able to identify therapeutic strategies that can reverse the disease at its earliest transition point, well before it ever manifests itself as a disease phenotype. Primarily for research purposes, we are now routinely assessing all analytes for every blood draw for all individuals looking at outliers that might represent disease transitions. It will take a great deal of data to separate signal from noise in this regard to ultimately develop clinical tests for early disease transitions. But this type of approach will eventually be the preventive medicine of the 21st century – identifying and stopping disease at its earliest stage before it really initiates a disease phenotype.
What is 21st Century Medicine?
There have been two paradigm changes in U.S. medicine in the last century or so. In 1910, Abraham Flexner wrote a report sponsored by the Carnegie Foundation on his investigation of 155 medical trade schools then existing in the U.S. The scathing report argued that 1) science should be used to properly treat patients, 2) science should be used to educate physicians, and 3) many of the trade schools should be eliminated. A few medical schools, including Johns Hopkins, were thinking along these lines and, over the next 40 or more years, catalyzed 20th century medicine by employing the relevant life sciences that existed at that time and directing them to the identification and treatment of disease. Virtually all of the medicine now practiced in contemporary healthcare systems is 20th century medicine. 21st century medicine was born with the emergence of systems approaches to disease – and the creation of systems medicine, P4 healthcare, and scientific wellness – the latter being the very embodiment of P4 medicine. Here the focus is on a health care that is predictive, preventive, personalized and participatory. The major areas of investigation for 21st century medicine involve taking a systems approach to understanding both wellness and disease, using systems technologies and approaches for studying the wellness-to-disease transitions, and creating the biomarkers and therapies that will allow them to be reversed. The systems approach and deep phenotyping also give us powerful insights into understanding individual wellness and individual disease. I would argue that even precision medicine as currently practiced is 20th century medicine – because it generally employs very shallow phenotyping (mostly low-level genomics) and is focused almost entirely on disease – not wellness.
Bringing 21st Century Medicine to the U.S. Health Care System
ISB affiliated with PSJH in 2016 to create an entity that had several unique features for bringing 21st century medicine to a U.S. healthcare system. First, the PSJH C-suite believed in and supported 21st century medicine as defined above and thus supported our efforts to institute them gradually in PSJH. Second, ISB is one of the few institutions taking a broad systems view of healthcare and is interested in both disease and wellness. Third, we have been effective in using both philanthropy and industrial support to catalyze systems-driven changes in health care.
PSJH is the third largest non-profit health care system in the U.S. – with 51 hospitals in seven different western states. PJSH sees about 5 million patients a year and has 30 million electronic medical records. Thus, it is a large healthcare system that will be noticed as it adopts 21st century medicine.
We have created a scientific advisory committee comprised of many of the best scientists and clinicians in the PSJH to charter the future of research and bring 21st century medicine to PSJH. We are beginning to initiate a series of clinical trials that employ dense phenotyping and systems-driven strategies and technologies to reveal the mysteries of Alzheimer’s disease, breast cancer, sepsis, and scientific wellness.
For example, we have initiated a clinical trial on scientific wellness in which 1,000 PSJH employees are enrolled in the Arivale scientific wellness program for three years (and are compared against a normal population that receive ordinary treatment). The idea is to assess increased wellness in the treated population and to assess the substantial savings that will come from the practice of scientific wellness. These are the data that will bring payers around to supporting scientific wellness. We are also setting up a Scientific Wellness Clinic in PSJH and using that to bring scientific wellness to family practice physicians throughout the PSJH system.
We are also initiating a series of clinical trials for Alzheimer’s disease using three paradigm changes: 1) dense phenotyping to identify the earliest transition points in this disease, 2) a multimodal therapy to reverse early disease, and 3) cognitive learning techniques to assist in reversing early Alzheimer’s disease. Starting about 10 years ago, one scientist used a systems approach to optimizing synaptic communication, and this led to a 36-point, multimodal regimen that has had spectacular success in early observational studies. We are now establishing an Alzheimer’s Clinic in Seattle for a series of multimodal, systems-driven clinical trials for Alzheimer’s. Indeed, our first multimodal clinical trail for Alzheimer’s is now underway. From highly successful preliminary observational studies, we expect, within a five-year period, to be able to identify the earliest transitions and reverse perhaps 85 percent of early Alzheimer’s disease. This will have an enormous impact on the half trillion dollars a year we now spend on Alzheimer’s disease for 5.6 million individuals, as well as transforming the health of individuals at high risk for Alzheimer’s.
We have also initiated an N=1 study in a 65-year-old woman recently diagnosed with stage 3 bladder cancer. She (and her husband) requested that we carry out deep phenotyping across the course of her disease treatment. She was an Arivale client for four years, so we had a base line assessment of her blood chemistries before the cancer arose. The results from this study were striking. It showed us clearly which therapies failed (chemotherapy) and which were at least temporarily successful (surgical removal of the tumor). It demonstrated that her health was ideally positioned for a future course of immunotherapy. It also provided insights into many smaller details relevant to the course of this disease. It is our belief that, in the future, N=1 studies will be carried out on all of those patients with complex diseases (which includes most chronic diseases).
We are accruing patients currently for a clinical trial on breast cancer survivor patients to optimize their wellness after cancer treatment. Breast cancer patients often experience one or more of about 15 common complications – and dense phenotyping and systems-driven strategies will give us the tools to deal effectively with many of these complications.
It has been possible because of PSJH’s visionary leadership and ISB’s pioneering of 21st century medicine to begin initiating this change in the PSJH system. We have initiated clinical trials with dense phenotyping and systems-driven technologies and strategies for Alzheimer’s, scientific wellness and breast cancer survivors. We are establishing Alzheimer’s and Scientific Wellness Clinics to use as a base for extending our powerful new systems-driven approaches in these areas throughout the PSJH system. We are working on the funding to establish further clinical trials with dense phenotyping and systems-driven strategies in multiple sclerosis, diseases of pregnancy, lung cancer, Lyme disease, NASH, and type 2 diabetes.
Via our new startup company, we have developed a new type of artificial intelligence algorithm that has the ability to precisely reproduce the thinking of clinical experts for acute and chronic diseases. This system enables us to capture the best thinking for dealing with a disease in an organized manner, using virtual intelligent “Lego blocks” which can interact with one another. The result is a deep reasoning system into which one can then feed the symptoms and data of an individual and which produces insights to help inform a physician in regards to diagnosis, therapy choice, and strategies for treatment in the future. This system employs individual data from dense phenotyping and it is a continuous-learning system. As new patients go through the system, it utilizes feedback and machine learning to become more comprehensive and accurate in its predictions for diagnosis and therapy. We are now developing deep reasoning systems for Alzheimer’s disease, sepsis, and pain and addiction, and will, in the future, add scientific wellness as well as other major chronic diseases.
This marks a state change for organizing executable information on complex diseases and indeed, it will assist regular physicians, apply best medical thinking, and make them an “experts” in the diseases they are analyzing.
We also believe that education is an extremely important aspect of bringing 21st century medicine to the contemporary health care system. The essence of the educational challenge is to bring an understanding of 21st century medicine to healthcare professionals – including physicians and new medical students. ISB has an education group that is now preparing a series of modules that will form the backbone of a year-long high school biology course in 21st century medicine. This will also employ a textbook we have recently written on “Systems Biology and Systems Medicine”. We will abstract from this high school course the materials necessary to prepare a multi-day program on 21st century medicine – systems medicine, P4 health care and scientific wellness – for health care professionals and physicians. These courses will be initiated at the Seattle Science Foundation – an organization with superb audio-visual facilities.
Indeed, we are also partnering with a new medical school in Spokane, Washington, managed by Washington State University, to begin a new program for the education of physicians. The first-year class of MDs has enrolled in the Arivale scientific wellness program for the duration of its training at medical school. They will use their own data to determine how it can be translated into actionable possibilities for improving their own wellness and avoiding disease. These students will also use the new “Systems Biology and Systems Medicine” textbook to change didactic teaching. These new physicians will be leaders of 21st century medicine at both the level of practice and at the level of training and teaching.
21st century medicine will bring three important assets to the health care systems practicing it (including scientific wellness for its patients). First, it will increase the health (wellness) of individual patients. Second, it will identify individual transitions from wellness to disease for most chronic diseases, and it will in time provide the therapeutic approaches necessary for reversing most chronic diseases before they ever manifest. Finally, it will generate enormous cost savings for the healthcare system (e.g. If in five years we can reverse 85 percent of early-stage Alzheimer’s disease, then there will be tens to hundreds of billions of dollars of savings from the half trillion dollars we currently invest for this disease in the U.S. each year).
Once these benefits are realized, all healthcare systems will necessarily need to move to 21st century medicine. The question for health care systems is whether they want to be a leader or a follower in initiating 21st century medicine. Hopkins was a terrific leader in 20th century medicine and benefited enormously from this leadership. Who will be the leaders in 21st century medicine?