Multi-Disciplinary Reading – Book Reviews

The Age of Scientific Wellness: Why the Future of Medicine Is Personalized, Predictive, Data-Rich, and in Your Hands by Leroy Hood and Nathan Price (2023)

In this book the authors argue that the contemporary medical paradigm is one of “sickcare” rather than that of “healthcare.” To elaborate, in today’s world, the medical interventions happen long after a disease has taken hold, following a centuries-old medical playbook that goes like this: Wait for something to go wrong → Try to identify what caused the problem → Try to fix it → If it works, try the same approach on the next person → If it doesn’t work, treat the complications—or write a death certificate and move along.

But, it doesn’t have to be this way…and it shall not remain this way for too long.

The authors say that we are in the first stage of the largest paradigm shift in healthcare since the beginning of modern medicine. In the times to come the way we approach health will change so profoundly that we will struggle to understand why we ever did things any other way. In a not-too-distant future the healthcare will focus on the optimization of individual wellness and early detection of wellness-to-disease transitions with a potential for reversal before the emergence of clinically diagnosable disease.

This paradigm shift will be brought about by (1) systems-based approached for modeling human physiology a.k.a. “systems biology” (2) collection & mining of vast troves genomic, phenomic and digital data for all individuals (3) AI (4) having better understanding of biological aging and the on-going research of reversing biological aging (5) better understanding of brain and neurological disorders (6) immunotherapy for cancer.

My notes below don’t do justice to the depth and to the ‘interconnectedness of chapters’ of this book. If you are interested in the future of healthcare, then do give it a read.

The Rise of Systems Biology

• Currently, we study biological systems one gene at a time, one protein at a time, one disease symptom at a time, ignoring the ever-mounting evidence that there are very few processes in our bodies in which A causes B, and B causes C, and so on in a simple, linear manner. Biology generally operates as complex systems rather than linear pathways.

• Despite the tremendous advances that gene and protein sequencing offered us, most of the research world is still locked in classically reductionist research.

• Scientists engaged in systems biology conduct experiments to determine how each part of the human body is connected to other parts and to the whole.

• Academy of Sciences report described the biology of the future as systems biology.

• These days, when scientists are asked to speak about the future of biological research, they generally point to a future that is holistic, data-intense, integrative, dynamic, and network-based—all systems approaches.

• There are four major types of biological information: DNA, RNA, proteins, and biological networks.

Mining vast troves of information

• There are three categories of information we need for every person to implement a data-driven vision of wellness and prevention. The first is the genome, the source code of life, which is virtually invariant throughout a person’s lifetime (except for the mutations in cancer cells).

• The second is the phenome, an assessment of your body’s status at any point in time during your life as a result of the interaction of your genome, your lifestyle, and your environment. Your phenome changes continuously, and it can be sampled at any time through certain measurements such as the gut microbiome and blood analytes.

• The third category is digital measurements of health. Health data from smartphones, watches, smart rings, and other wearables that can track their heart rate, body temperature, respiration, activity level, calories consumed and burned, sleep, menstrual cycles, blood sugar, hormone balances, and more.

• With these three categories of data in hand, we can assess the optimal physiology of one’s body and brain and detect early transitory phases many years—even decades—before a disease becomes clinically apparent.

• Actionable possibilities that have arisen in the recent years: The American Society of Human Genetics has identified seventy-six disease-causing variants for which there are possible actions to prevent disease.

• The new field of clinical pharmacogenomics promises to teach us how our individual genetic variants and their associated drug toxicities can affect our choices for appropriate drug treatment.

Understanding Genetic Risk

• Genetics that affect the risk of disease fall into three categories: dominant genes, recessive genes, and polygenic risk scores.

• Dominant genes are responsible for diseases that can be caused by a single copy of the “bad” gene. Most dominant genes have lower degrees of penetrance, meaning they elevate the risk, but disease is not a certainty. This is the case with BRCA1 and BRCA2, the breast and ovarian cancer genes.

• An individual needs two copies of a defective gene, however, to inherit a so-called Mendelian recessive genetic disease like in sickle cell anemia, cystic fibrosis, Tay-Sachs disease, and hemochromatosis.

• The third category of genetic diseases occurs as the result of a polygenic combination of many—up to thousands—of variant genes, each contributing minutely to the total genetic risk.

Reversing biological aging

• “Hallmarks of aging” are: (1) changes in the ways genes are expressed over time (2) a growing inability to get rid of dysfunctional proteins through autophagy (3) exhaustion of stem cells (4) accumulation of zombie-like senescent cells that inflame healthy cells

• Nicotinamide riboside (NR), which boosts NAD+ levels in humans and has been tied to an increased health span in a number of animal models. NR is an alternative form of vitamin B3 that acts as a precursor for the production of nicotinamide adenine dinucleotide, or NAD+, a chemical compound that is vital for the process of converting food into energy, repairing damaged DNA, and maintaining circadian rhythms.

• There is also a lot of current interest in a similar molecule, nicotinamide mononucleotide or NMN. NMN works through similar mechanisms to increase production capacity for NAD+

• Another intriguing candidate molecule is rapamycin, which was initially used to treat renal transplant patients as an immunosuppressant

• Epigenetics measures the extent of methylation of one’s DNA and hence the extent to which gene expression is blocked (thus slowing aging).

Alzheimer’s and Dementia

• Alzheimer’s patients have an estimated annual cost of $500 billion a year in the United States alone.

• At its heart the key disease mechanism is associated with metabolic energy deficits in the brain. PET scans show metabolic deficits as many as ten to fifteen years before cognitive decline sets in for those diagnosed with Alzheimer’s.

• The APOE4 variant of gene APOE is a major genetic risk factor for Alzheimer’s. One bad copy of this gene increases risk 2x; having two bad copies makes it 3x.

• Efforts to find a “magic pill” for Alzheimer’s are ongoing—and most of these still target amyloid, the protein whose aggregation in the brain was one of the first features noted originally by doctors as a cause for the disease.

• The prevailing hypothesis about Alzheimer’s has for decades focused on the buildup of amyloid plaque, which authors feel is woefully incomplete. Following hundreds of failed clinical trials, the tide has been turning against the conventional amyloid thesis of Alzheimer’s

• As late as 2021, the FDA approved a new monoclonal antibody that targets amyloid called aducanumab, created by Biogen. This approval was a controversial decision, to say the least. None of the eleven scientists on the advisory panel voted to approve the drug, citing a lack of evidence for efficacy. The FDA took the highly unusual action of overriding the recommendations of the scientists and moving ahead with approval anyway, reasoning that, although it showed no significant effects on cognition, at least it reduced amyloid plaques.

Cancer – The Promise of Immunotherapy

• Cancer is one of the fields that is increasingly moving away from a “one size fits all” approach to what is known as precision treatment. It has probably gone further than any other area of medicine toward personalization of treatments based on a patient’s genetic profile.

• Before the advent of immunotherapy, the entrenched view that “slash-poison-burn” (surgery-chemo-radiation) was the best way to tackle cancer. But in the past decade, there has been an explosion of immunotherapy research, prompting tremendous changes in clinical practice. The prediction that there would one day be a fourth kind of cancer therapy—immunotherapy, along with surgery, chemotherapy, and radiotherapy—has finally come true.

• Immunotherapies can be thought of as “training” the sentry T and B cells, helping them to become even better at identifying and fighting invading cells.

• Immune checkpoint therapy is a new type of immunotherapy in which drugs are used to block proteins that keep one’s immune responses from effectively identifying and attacking cancer cells. This strategy does not work with every cancer, but when it works, it really works—and it can have transformative effects for patients with hard-to-treat cancers such as melanoma, multiple myeloma, and renal cell carcinoma.

• The promise of immunotherapies has rejuvenated cancer research. It has also driven the development of new cancer technologies, strategies, and companies. Between 2017 and 2021, the number of immuno-oncology drugs in development grew by a staggering 233 percent to over 4,700.

• A report from the Cancer Research Institute identified a 60 percent increase in the number of academic and research groups actively conducting immunotherapies between 2017 and 2019, leading to the approval of thirty-one new drugs by the FDA.

• Immunotherapy holds enormous promise, but it cures only a few cancers (mostly tumors in the blood or liquid tumors, such as leukemias or lymphomas) and has not yet been effective for most organ or tissue tumors.

• Cancer diagnostic biomarkers are molecules whose levels (or fluctuations) reflect a transition in health status. These can include proteins, metabolites, lipids, fragments of “cell-free DNA” released into the bloodstream when tumor cells die, and blood exosomes.

• The company Grail recently made a remarkable advance in cancer biomarker discovery. By analyzing the epigenetic profiles of white blood cells it was able to diagnosis about fifty different cancers. Grail was recently purchased by the Illumina for $6 billion.

• Immunotherapy, new approaches to identifying blood biomarkers, and genome / phenome clinical trials will move us aggressively toward effective cancer treatments in the next ten years.

AI imperative

• AI systems are already transforming healthcare. Those changes will accelerate in the coming years to such a degree that AI will soon be as much a part of our healthcare experience.

• Protein folding: There is a long-standing “grand challenge” for computational biology: being able to predict the shape of a folded protein given just its gene (amino acid) sequence. This is an ideal problem for data-driven AI in that it is well defined, offers concrete ways of measuring “better” or “worse” predictions.

• It’s also a question of great importance, because the shape of a protein determines its function—what chemical reactions it may catalyze, what molecular machines it will help build as multiprotein complexes, and how it interacts with other molecules to assemble cells, tissues, and organs. So if we want to understand the chain of events that moves us from DNA to RNA to amino acids—the building blocks of proteins—and onward to the full complexity of human life, we have to know how proteins fold.

• This challenge of predicting protein structures is so important to the scientific community that a competition (CASP – Critical Assessment of Protein Structure Prediction) is held every two years.

• DeepMind’s AlphaFold entered the CASP contest in 2018, it beat the competition by nearly 50 percent, coming in at an accuracy of nearly 60 percent. In 2020, AlphaFold 2.0 blew the original AlphaFold away, making an even bigger leap forward and jumping to nearly 85 percent.

• Protein folding problem is far from solved. Proteins interact dynamically and have multiple configurations as they carry out their functions in living systems. And while AI is now good at predicting the crystallized structure, none of the current approaches can predict proteins in all of their potential structures, and validating such predictions is a significant challenge.

• Another major domain of medical AI is interpreting imaging data. Images make up a large fraction of medical data and generally take significant amounts of time from trained experts to interpret. AI technologies are now providing help in extracting, visualizing, and interpreting imaging data.

• One of the most exciting technologies under development today enables us to simulate each individual’s unique physiology (top down) and biochemistry (bottom up) in a computer, creating a “digital twin”. These programs essentially seek to create a computational version of all that is known about your unique physiology and biochemistry so that, when the time comes, medical interventions intended to help save or extend your life could first be simulated on your computational twin.

Genomics: Leading the Way

• Of all the -omics technologies (proteomics, metabolomics, transcriptomics, epigenomics, microbiomics, genomics), genomics is the first that is being integrated at scale into healthcare systems.

• The widest adoption of genomics today is for sequencing tumors and comparing these sequences to the genome a patient was born with. This information is then used to target therapies with a higher likelihood of success.

• When the first genome was sequenced, it was true that genetic codes didn’t provide much actionable information. As of 2022, there were at least seventy-six genetic variants classified as clinically actionable, including for colon cancer, breast cancer, ovarian cancer, heart failure, and sudden cardiac death.

• The combination of genomics with other data types—including clinical, wearable metrics, and multiomic deep phenotyping data—considerably increases actionability.

• The costs for whole genome sequencing are coming down; some of these predict that the cost of whole genomes will drop to $10 in the not-too-distant future.

• One of the very big advantages of whole genome sequencing is its ability to help you decide what biomarkers to monitor because for most people, right now, it would be both prohibitively expensive and overwhelmingly complicated to test for every potentially meaningful marker.

Random

• The human tragedy and economic costs of diabetes are staggering—some $327 billion is spent annually on its treatment in the United States alone.

• False-positive cancer screening test results are so common that many doctors have concluded that some of these tests are worthless.

• Few people realize it, but the ten most popular drugs in the United States today—from esomeprazole and rosuvastatin to fluticasone and pegfilgrastim—work, collectively, for only about 10 percent of treated patients. Too many people are being subjected to their known side effects without benefit to their underlying condition.

• Freeman Dyson writing in Imagined Worlds: “New directions in science are launched by new tools much more often than by new concepts. The effect of a concept-driven revolution is to explain old things in new ways. The effect of a tool-driven revolution is to discover new things that have to be explained.”