Rows of small white bungalows, once the property of workers at the local textile mill, still line the streets of Kannapolis, North Carolina. A larger-than-life bronze statue of town hero and NASCAR icon Dale Earnhardt Sr. looms over the town square, and the single-screen Gem Theater just off Main Street shows the latest movie releases on weekends. It is, on the surface, a traditional Southern town.
But a short walk from the center of Kannapolis, just down the street from the local Baptist church and across from a tractor supply store, stands a blocky, nondescript gray building with a small parking lot and manicured landscaping. Inside is a huge warehouse, the length of two football fields. The space hums with the whirr of high-powered fans in the ceiling, which cool rows of tall gray freezers, each marked with an orange biohazard sticker. A technician opens a freezer and cold white mist pours out, revealing small metal boxes full of plastic vials—vials containing human blood and urine.
These are the bodily fluids of the locals. Since 2009, scientists have been collecting and storing biological samples from Kannapolis-area residents as part of a grand experiment. The project is big and bold: collect large amounts of physical and biochemical information from 50,000 townspeople, store that information in computer databases, then sift through it for molecular signals of disease. The goal is to rewrite the textbook of medicine.
It’s something we’ve needed for a while. Even in the 21st century, with the advent of anesthesia, vaccines, antibiotics and more, medicine remains a sophisticated guessing game. For example: An individual develops pancreatic cancer. As there is no reliable test for early detection of pancreatic cancer, the tumor is almost surely large by the time it’s discovered. Our imaginary patient, newly diagnosed, is prescribed the top-of-the-line chemotherapy. The treatment is difficult; he loses his hair, can’t get out of bed and vomits daily. If that drug cocktail doesn’t work, he is prescribed the second-best therapy on the market. If that therapy doesn’t work, it’s on to the third-best—fingers crossed that this one shrinks the tumor. His medical treatment is not tailored to him; his doctors do the best with what they have, but what they have is a one-size-fits-all approach.
A growing contingent of researchers and doctors advocate overhauling contemporary medicine by creating a medical system tailored to individuals. Also called molecular medicine, personalized medicine aims to tailor treatment based on molecular markers found in a patient’s blood, spit, urine, muscle tissue or pretty much any other part of the body. These molecular signals or “biomarkers,” such as a mutated gene or an overabundant protein, might eventually be used to more accurately predict disease risk, make diagnoses and identify the best treatment options for individuals. This new paradigm of medicine has been heralded since the sequencing of the human genome, but has yet to come to fruition. In fact, no one knows if it is going to work. But the residents of Kannapolis have signed up to try to find out.
Kannapolis isn’t the first place scientists have built a biobank of human samples. The Department of Veterans Affairs’ Million Veteran Program, launched in 2011, has collected blood samples and clinical information for over 200,000 veterans. Health care provider Kaiser Permanente in California is gathering genetic and medical data on 500,000 of its members. These are just two examples, and there are many more.
But the Kannapolis study is unique and important because of its community-based approach. In an established community, scientists can compare sick patients with healthy individuals who are not only the same age and gender but also live in the same neighborhood, helping to remove environmental and socioeconomic factors from the equation. Also, by ingraining the study into the culture and families of the town, scientists can better keep track of participants and follow them over long periods of time with a low risk of dropouts—a common pitfall in long-term studies.
Pastor Andy Langford of the Central United Methodist Church was one of the first volunteers to pee in a cup and have his blood drawn. “It only took 30 minutes. It wasn’t that big a thing,” he says in a gravelly Southern drawl. Langford pauses when asked why he participated. “It has potential to improve the health of everyone. Not just in this community, but everywhere. Why wouldn’t I do it?”
Murdock and the Mill
The MURDOCK Study (Measurement to Understand the Reclassification of Disease of Cabarrus and Kannapolis) is named after its founder and funder, David H. Murdock, the billionaire owner of the Dole Food Co. In 1985, Murdock lost the love of his life and wife of 18 years, Gabriele Murdock, to ovarian cancer. After her death, he became obsessed with longevity, devouring scientific articles on nutrition and health, seeking out experts on the topics, and in 2003 founding the Dole Nutrition Institute to support nutrition research.
He couldn’t save his wife, but he wants to save the rest of us. He decided to set up the home base for this herculean mission in Kannapolis. It wasn’t chosen at random. In 1982, Murdock purchased Kannapolis’s Cannon Mills, once the world’s largest producer of sheets and towels. Three years later, he sold it to Fieldcrest Mills, which sold it to Pillowtex Corp., which eventually filed for bankruptcy, and the mill closed in 2003. According to the town records, 4,340 employees lost their jobs in a single day.
Feeling responsible for the town, Murdock bought the mill back at auction in 2004, for $6.4 million. Two years later, he demolished it and built, in its place, a lavish, 311,000-square-foot, four-story Georgian building. Today that building is the epicenter of the North Carolina Research Campus (NCRC). It houses the Core Lab, a scientific candy store stocked with all the instrumentation a scientist could dream of. Two more buildings sprang up nearby, housing even more labs and equipment. The NCRC has become home to many researchers from local universities, including North Carolina State University and the University of North Carolina, as well as the research teams of private food and agriculture companies such as Dole, Monsanto and General Mills.
As Murdock instituted his grand plan to transform Kannapolis from a textile town to a biotech hub, he met Robert Califf, a cardiologist at Duke University. Califf and his colleagues pitched an idea that appealed to Murdock’s connection to the Kannapolis community. What did he think about employing the tools of the NCRC to study the townspeople of Kannapolis?
Docs and Data
For almost 30 years, Califf has been at the forefront of a growing belief that the best medical outcomes occur when one marries a computer’s ability to collate and calculate data with the wisdom and perception of an experienced doctor.
In 2007, Califf got together with Andy Conrad—a geneticist, businessman and close confidant of Murdock, and currently director of Google X, the research lab of Google Inc.—and Duke geneticist Geoffrey Ginsburg to formulate an idea for an intensive study of personalized medicine. Their plan was to gather health information from community members and use it to bring medicine into the data age. They presented it to Murdock, who made the idea a reality with a $35 million gift.
The first phase of the MURDOCK study was to demonstrate that large datasets of human clinical and molecular information could legitimately identify disease biomarkers—that the project could eventually turn data into medicine. But since the team had not yet collected samples from Kannapolis residents, Califf and his Duke colleagues identified four test cases—in cardiovascular disease, osteoarthritis, obesity and liver disease—where they could draw on an existing biobank of patient samples and see what they could find.
The goal was to see if this biobank approach would work. For example, in the cardiovascular study, Kristin Newby and colleagues scanned the proteins, metabolites and RNA from 2,000 cardiac patients who donated a sample to a Duke biobank. Newby wanted to discover a way to ascertain a person’s risk of a heart attack. Currently, to determine an individual’s risk of attack, doctors consider factors such as one’s age, cholesterol, blood pressure and smoking history. “But those models are by no means perfect predictors,” says Newby. In fact, they’re only about 70 percent accurate. Using blood samples, Newby’s group identified groupings of proteins and metabolites—small molecules that cells produce during metabolism—that are consistently associated with heart disease risk. That work, some of it already published, with more to come, could significantly improve that 70 percent accuracy rating.
The other three initial studies—in osteoarthritis, obesity and liver disease—have also been successful in identifying biomarkers that could be useful in the clinic. These early findings successfully demonstrated the power of their model: Big data can lead to clinically important biomarkers. Now they were ready to put them to the test on a grander scale—in Kannapolis.
To date, the MURDOCK study has enrolled 10,571 individuals; the study’s goal is to push that to 50,000. The team originally hoped to enroll residents much faster. “The first year was a rough year for them in trying to get the message across,” says William “Phred” Pilkington, director of the local health department, because no one was familiar with the study or the people promoting it. To make matters worse, enrollment began in 2009, just after the Great Recession hit. “A clinical research study is very important, but when you don’t have a job, it’s not the most important thing you think about,” says Mary Lou Perry, whose parents worked at the mill and who is now employed as an administrative assistant for the MURDOCK study.
Eventually the local economy improved—thanks in part to jobs created by the NCRC, which employs many residents in administrative, maintenance and even biotechnology positions—but some residents still declined to join the study. People can be scared off by the blood draw, says Leah Bouk, the clinical research coordinator for the study, but the most common reason for declining is privacy concerns, specifically whether the government can access participants’ samples and identities (it can’t).
To gain trust and alleviate fears, the MURDOCK team members embedded themselves in the community. Organizing health fairs, they offered free blood pressure screenings and Dole bananas. They went door to door, describing the study, and distributed “I Support the MURDOCK study” yard signs. Though Murdock doesn’t live in Kannapolis, when he visits, he joins in the effort. Last year, for example, he attended an enrollment drive at a primary care practice, where he gave a short presentation, then answered questions from local doctors.
Today, enrollment proceeds at a steady clip, with about 28 enrollees per week. “The community has really begun to rally,” says Perry. Langford enrolled, then persuaded 150 members of his flock to join the effort. Marie Dockery, director of a free health clinic in Concord, a town south of Kannapolis that is included in the study, enrolled. “We have seen the effects of not being informed about health in our parents and grandparents,” she says. “I think there’s an altruistic appeal for people to do this for the betterment of our community.”
Each participant is asked to donate three tablespoons of blood and three tablespoons of urine, and signs an eight-page consent form that gives researchers access to the samples. Participants receive a $10 gift card, but most, like Dockery, cite noble motives for joining.
The MURDOCK team already has enough samples to go to work. Duke researchers are using the biobank to study aging and cognitive decline, severe acne, physical performance and multiple sclerosis.
Simon Gregory, a Duke geneticist who identified one of the first genes associated with multiple sclerosis in 2007, is searching for biomarkers that signal progression of that incurable disease. MS is an inflammatory disease in which the body’s own immune system cells attack protective sheaths around nerves, leading to irreversible nerve damage. Currently, MS is diagnosed through an MRI scan that detects scarred neural tissue. “But what we’d like to develop is a biomarker that could either tell an MS patient if they’re going to have a relapse ahead of time or be an early marker for diagnosis,” says Gregory.
Gregory’s team has collected all the data and is now in the process of analyzing it, looking for signs that might indicate disease progression.
“The tremendous thing about the MURDOCK study is that its collection protocol is so good. They collect the samples in all the same way,” says Gregory. Standardized samples, plus the NCRC’s cutting-edge instrumentation, create the “perfect storm” for biomarker research, he adds. “The MURDOCK study [unites] all of the things we need to find these answers.”
Yet for all the potential, Kannapolis locals are aware that the study is unlikely to directly affect their health in the near future. But it has already begun to affect them in other ways. “Kannapolis is once again on the map,” says Perry. “We’re proud of the fact that we’re doing our part to personalize medicine.”