Stop it

It’s not often that world-class scientists band together to investigate disease with no intention of curing it. Yet upward of 55 scientists at Stanford’s Precision Health and Integrated Diagnostics Center are doing just that in a push to get researchers and physicians off their heels and onto their toes in the battle against disease.

At the center, the goal is not to find a fix for the world’s most pressing ailments; it’s to detect them at their earliest stage, if not prevent them entirely.

“We want to be proactive, not reactive,” said the center’s director, Sanjiv “Sam” Gambhir, MD, PhD, professor and chair of radiology. “My thinking here was, ‘What can we do so that the whole diagnostic field better aligns with precision health?’ I think the way to get the biggest gain — although it will take several decades to play out — is to lead the charge on proactive research across multiple diseases in a broad-picture kind of way.”

Established last year, the center, abbreviated PHIND and pronounced “find,” backs dozens of scientists eager to test out some pretty nontraditional health-research ideas, such as nanosensor-equipped toilets that extract data from daily, um, deposits; bras that image breast tissue in search of abnormal changes; and a menstrual pad that can detect biomarkers of disease.

But the central concept serves to ground the very philosophy of PHIND: using repetitive, precise measurements of individuals’ health to make diagnoses earlier and ultimately stop disease before it causes real damage.

“I think the way to get the biggest gain — although it will take several decades to play out — is to lead the charge on proactive research across multiple diseases in a broad-picture kind of way.”

“PHIND is the only center to use precision health in such an enormous scope and scale,” said Ryan Spitler, PhD, deputy director of the center. “We’re looking at healthy and at-risk individuals to understand cardiovascular disease, cancer, neurological and mental health, and diabetes. In conjunction with those disease areas are the different ways in which you can measure transitions from health to disease: wearables, implantables, data analytics and molecular mechanisms.”

The effort exemplifies Stanford Medicine’s focus on precision health under the leadership of Lloyd Minor, MD, dean of the School of Medicine.

The reality upon which all the PHIND projects hinge is often overlooked: Almost every person to ever fall ill was, at one point, healthy. So while other big research entities gun for the next breakthrough therapy, PHIND scientists put the transition to disease, rather than disease itself, under the microscope to prevent healthy people from becoming patients. 

Stopping the ‘spikes’

One such scientist is Michael Snyder, PhD, professor and chair of genetics. More casually, he’s known as “the omics guy.”

Omics refers to the study of all the molecules in a particular biological sub-category in an organism — like all the genes, or all the proteins. Snyder and his collaborators are putting various omics profiles to work in a clinical trial that aims to prevent Type 2 diabetes.

They’re studying 100 people, all considered healthy but potentially prediabetic. Each participant receives a device to track and report glucose levels in their blood, around the clock. But the researchers also take samples of every participant’s microbiome (the mass of bacteria that mobs our gut) and metabolome (the collection of molecules produced during metabolism). They store these samples to examine later.

Snyder is looking for “spikers” — people whose glucose levels skyrocket after eating carbohydrates. A sharp uptick in glucose indicates that something is askew with either a person’s insulin — a hormone that helps the body turn carbohydrates into usable energy — or how that person’s body takes up glucose, and it’s a telltale sign of diabetes.

But insulin and carbohydrates are not the sole culprits of blood glucose booms. “The microbiome plays a big role in people’s glucose levels spiking,” said Snyder, the Stanford W. Ascherman, MD, FACS, Professor in Genetics.

“So what we plan to do with this trial is monitor each person’s glucose levels and the composition of their microbiome and their metabolome, then use this information to ideally piece together a diet plan that keeps their glucose levels under control.”

“It’s not just ‘Don’t eat carbs.’ Different people spike to different things. I spike to bananas; you might be fine with bananas, but you might spike to rice. We need to pinpoint the dietary needs for each person.”

The key, Snyder said, is precision. “It’s not just, ‘Don’t eat carbs.’ Different people spike to different things,” Snyder said. “I spike to bananas; you might be fine with bananas, but you might spike to rice. We need to pinpoint the dietary needs for each person.”

Snyder’s goal is to compile all of this information — glucose readings, microbiome and metabolome profiles — and use machine learning to prevent the onset of diabetes by not only predicting who’s at high risk for diabetes, but also by prescribing diets that harmonize with each participant’s metabolic tendencies.

“About 70 percent of prediabetic people become diabetic, and that’s why it’s so crucial to catch and manage these conditions before people even show symptoms,” Snyder said. “We think the PHIND center will be very powerful for understanding basic metabolic control. It’s one of the biggest problems out there, and we hope this project will help us better understand and control people’s metabolic function, especially in glucose control and diabetes.”

Preventing teen depression and suicide

The beauty of precision health is that it can apply to nearly any field of biology, even the harder-to-pin-down ones, such as mental health.

In partnership with PHIND, Ian Gotlib, PhD, professor of psychology, is applying a rigorous approach to understand the many spokes that support psychological well-being, particularly as it relates to depression and suicidal behavior in teens. Gotlib’s goal is to compile neurobiological, molecular and experience-based information and use machine learning to predict which mixes of factors could predispose someone to dangerous, even life-threatening, behaviors.

“We know that adolescence is a peak period for the rise of depression, but we cannot predict these increases in depressive or suicidal behavior,” said Gotlib, who is the David Starr Jordan Professor. “We don’t yet have a sense of how to do that, which makes prevention difficult. So for the past five years my group has been conducting a comprehensive assessment of mental health in children and adolescents, and now with PHIND we’re empowered to go even deeper and consider new mental health factors for longer periods of time.”

Gotlib’s study follows 220 boys and girls from late childhood (ages 8 to 11) into their early teenage years (ages 13 to 16). In the first leg of the study, scientists interviewed children and their parents about the children’s stressful early life events — moving homes, parents’ divorces, witnessing violence, things of that nature. They also measured other aspects of stress, including cortisol levels, and assessed pubertal hormone levels and functional and structural brain connectivity.

“These things — depression, anxiety, suicidal behavior — develop through adolescence, and the only way to understand that development is to start with a sample of healthy people and study what the risk factors and biological signs are.”

Gotlib and his team are continuing to follow these adolescents into their teen years. With PHIND, they’re not only able to keep measuring parameters from the first part of the study, but also able to buttress the data with new measures of inflammatory markers, high levels of which are often seen in people with depression.

They’re even able to use data from watches that monitor activity and rest, and from the teens’ cellphones to tell if something is amiss. There’s no spying going on here; they’re not looking at the content of the phone, but rather the physical movements of the phone — how often it’s used to send messages, how quickly the user responds to messages. (When things start to slow down, it can flag depression.) The teens’ phones were also equipped with an app that prompts them to record information about their mood and activities in real time.

“We tend to neglect healthy people, but people aren’t born with the disease state we’re looking at,” Gotlib said. “These things — depression, anxiety, suicidal behavior — develop through adolescence, and the only way to understand that development is to start with a sample of healthy people and study what the risk factors and biological signs are. This early detection and understanding the basics of the disease are central to PHIND’s mission.”

Blood-based cancer clues

The need for quicker cancer diagnostics has researchers combing the genome for molecular hiccups indicative of the disease — whether it’s how tumors start, spread or, ideally, how they can be stopped.

In a PHIND-funded project, Christina Curtis, PhD, assistant professor of medicine and of genetics, and Anshul Kundaje, PhD, assistant professor of computer science and of genetics, have turned to a simple blood draw, or liquid biopsy, to reveal cancer’s most complex secrets. Curtis’ goal: Use blood analysis as a tell-all source that not only flags the presence of cancer, but reveals where it came from in the body and if it’s poised to infiltrate other organ systems. And she wants to do it all on an earlier timeline, with heightened precision.

It’s a lot to ask of a run-of-the-mill blood draw, but Curtis and her group are devising a search tactic that goes beyond identifying rare mutations in DNA. She’s looking at epigenomic footprints, types of markers typically embedded in DNA. While these markers are often found enclosed in the cell, Curtis takes advantage of cell-free DNA, which floats openly in the bloodstream after shedding from a tumor or from healthy tissues.

“We’re taking a really different approach. We leverage epigenomic profiles, which contain information about which tissue the cell-free DNA is derived from and if it’s cancerous,” Curtis said. “And while certain mutations are important hallmarks of cancer, there’s a unique profile that the epigenome provides, including clues about the cell’s activity or state.”

“A big part of the challenge is that cell-free DNA in and of itself hasn’t been deeply studied yet. What we’re looking for here are really needles in haystacks — rare molecules that have been shed from somewhere in the body.”

Research from Curtis’ lab shows that some cancers are born to be bad from day one; mutations and epigenomic factors render the cancerous cells aggressive, malignant and more lethal. Curtis hopes a blood-based analysis could detect that kind of cancer and its origination before the patient even shows symptoms. In that sense, it would work as a screen, she posits, that everyone could incorporate into their routine annual checkup.

That’s still a long time away, Curtis said. But in her research, she’s beginning to apply the blood-based technique to consenting cancer patients, working backward to test her technology’s ability to pinpoint cancer types and aberrant signaling from cell-free DNA. So far, her preliminary research has yielded robust results.

“A big part of the challenge is that cell-free DNA in and of itself hasn’t been deeply studied yet. What we’re looking for here are really needles in haystacks — rare molecules that have been shed from somewhere in the body,” Curtis said. “There’s still a question surrounding what the makeup of a healthy individual looks like, so we’re working on understanding that, too, because without that, we have no meaningful reference.”

The future of PHIND

Technology, however, can advance only as far as researchers’ understanding of biology enables it. “The smart toilet, which is being developed in my lab, can’t work miracles if it doesn’t know what to look for in the urine. It’s not a crystal ball; it has to know what biomarkers to detect,” said Gambhir, the Virginia and D.K. Ludwig Professor for Clinical Investigation in Cancer Research. “That’s why we need more people on the basic biology side to understand the early changes as cells transition from normal to ill cells.”

“Science isn’t often discovery out of nowhere. It comes out of fortuitous collisions in which different fields that don’t typically communicate come together.”

To this end, PHIND has doled out $2.75 million to help catalyze basic, prevention-focused research at Stanford based on a competitive formal process. Some 20 projects were funded in the initial round. In May, the center announced the availability of an additional $1.5 million in seed funding. The goal is to launch as many as 12 new research projects by the end of this year.

“Science isn’t often discovery out of nowhere. It comes out of fortuitous collisions in which different fields that don’t typically communicate come together,” Gambhir said. “And that’s what we want to facilitate with PHIND to empower the science behind precision health and earlier diagnostics of multiple types.”