Illustration: Aïda Amer/Axios
The goal of reaching an era of individualized precision medicine will first require a closer look at the broader population.
The big picture: Large clinical trials and massive databases of de-identified genetic and other health information — sometimes from generations of populations — are offering scientists and doctors data to decipher why certain individuals have a higher risk of disease or different responses to treatments.
- Until the roles of genetics, ancestry, environment, diet, age and gender are better understood, precision medicine will remain an elusive goal.
- By running large clinical trials of treatments that incorporate people from all communities and backgrounds, there's a better chance of knowing if certain groups will respond differently.
- And by collecting and tracking genetic and other information from people from diverse backgrounds and circumstances, researchers — often using machine-learning or AI — can examine what happens to a particular group's health over longer periods of time.
What's happening: There are many institutions gathering this data, including...
- Large, multigenerational studies like the Framingham Study that looks at cardiovascular and neurological diseases or the Nurses' Health Study examining risk factors behind women's chronic diseases.
- First conceived 20 years ago, the U.K. Biobank continues to collect genetic and health information from 500,000 U.K. participants, leading to insights on topics like how COVID might damage the brain and why there are "never COVID" cohorts.
- The U.S. Million Veteran Program, which started about a decade ago, has collected some genetic and health data from roughly 870,000 veterans, with research starting to be published.
What's new: The COVID-19 pandemic led various groups to collectively create large-scale studies to seek safe and effective COVID treatments as rapidly as possible, such as the U.K.'s Recovery trial on more than 47,000 participants and the WHO's Solidarity Therapeutics Trial on 14,200 randomized hospitalized patients globally.
Growing awareness of the problems caused by a lack of diversity in clinical trials and in most genetic databases has led to other changes.
- The U.S. research program All of Us, which has more than 480,000 participants, recently released whole genome sequences from almost 100,000 U.S. participants, almost half of whom are from underrepresented groups.
- That All of Us data is expected to lead to discoveries that could reduce health disparities and help us "move away from a one-size-fits-all approach to treatment," says Karriem Watson, chief engagement officer of All of Us.
- Smaller public-private partnerships are also tackling the issue. For instance, Sema4, a patient-centered AI health care company, launched the REPRESENT trial to enroll 5,000 advanced-stage cancer patients from diverse populations in comprehensive genetic and genomic testing, says Sema4 chief medical officer and oncologist William Oh.
Reality check: Personalized medicine continues to face serious challenges, and has sometimes resulted in deadly missed targets. But many hope accumulating data from large, more diverse trials will help alleviate those issues.
Between the lines: Large cohort studies are one of the key "strategies to be able to understand the risk factors associated with cancer and with other diseases," says Marcia Cruz-Correa, physician-scientist at the University of Puerto Rico Comprehensive Cancer Center.
- For example, the Framingham study found that high levels of cholesterol were a risk factor for coronary artery disease — "and now we don't even think twice about checking cholesterol levels as a standard of care," says Cruz-Correa.
- "We've been basically missing the boat on a very large proportion of patients who are underrepresented in clinical cancer treatment and research," Oh says.
The bottom line: These massive datasets are expected to help tease out the biological and socioeconomic factors of disease, Oh says. "They're all tied together."
Go deeper: