BEATing COVID-19 from home
FDA, biopharmas are using real-world data from online initiative BEAT19 to inform clinical trial design
As knowledge regarding the natural history of COVID-19 continues to evolve, biopharmas are looking to real-world data generated from the BEAT19 initiative to inform clinical trial design for COVID-19 therapies.
Behavior, Environment and Treatments for COVID-19 (BEAT19), run by AI company xCures Inc., aims to collect demographic and health data from 100,000 non-hospitalized individuals in the U.S. with the goal of distinguishing between early and late onset symptoms, measuring the severity and duration of each symptom, and identifying contrasting symptoms between COVID-positive and COVID-negative cases.
While most COVID-19 natural history data collected so far comes from hospitalized patients, BEAT19 is filling a gap in knowledge regarding symptoms and disease progression in non-hospitalized individuals.
Mark Shapiro, VP of clinical development at xCures, told BioCentury the idea for BEAT19 grew out of the company’s platform, which continuously collects real-world data from cancer patients on all types of treatments to inform treatment decisions.
A big difference is scientists are starting from scratch with COVID-19. “We thought it would be important to come up with a more definitive understanding of what exactly the symptoms are,” Shapiro said.
“The fatigue here is severe and it seems to affect almost everybody who has this illness.”
BEAT19 went from an initial concept to participant enrollment in only nine days. The process included IRB approval, clinical trial registry and building a website.
It collects real-world data through daily online surveys from voluntary participants ages 15 or older with or without COVID-19.
As of June 18, 3,387 participants across all 50 states had completed 53,905 surveys. So far, the self-reported race and ethnicity data mirror the 2010 U.S. census, and 70% of respondents identify as female.
Shapiro said companies designing COVID-19 trials had reached out to xCures early in the initiative about accessing real-world data from BEAT19 because they needed a model of what the timeline and severity of symptoms look like in a COVID-19 patient to measure the efficacy of their treatment or vaccine.
According to Shapiro, FDA is using real-world data from BEAT19 to define symptomatic infection and provide guidance to drug and vaccine developers about what symptoms should be measured in clinical trials, and Takeda Pharmaceutical Co. Ltd. (Tokyo:4502; NYSE:TAK) and other undisclosed companies are using data from BEAT19 to inform clinical trial designs.
Trends so far
One aim of the initiative is to create a symptom severity model to understand when symptoms start and end relative to one another and how severe the symptoms are over time.
For example, BEAT19 has revealed that while fever is common, it is usually a low-grade fever. But early in the pandemic, high fever was a prerequisite for getting tested for COVID-19, which likely led to symptomatic patients being denied testing.
BEAT19 data also show that headache is a severe symptom but resolves quickly whereas fatigue is severe yet persistent. Many participants experience fatigue for over a month.
Figure: A theoretical COVID-19 symptom timeline and severity model
Symptom A is an early onset symptom and is moderate and short-lived. Symptom B starts to appear right before the peak of symptom A and is mild yet persistent. Symptom C emerges further into the infection timeline when symptom A is almost gone. Its peak coincides with symptom B’s peak and is severe and persistent. Source: Interview with xCures Inc.
“It’s not like the flu. The fatigue here is severe and it seems to affect almost everybody who has this illness. It’s as severe as what we see in lung cancer patients on chemotherapy,” Shapiro noted.
Participants reported coughing, aches and headache as the most burdensome symptoms following fatigue.
Despite the widely reported news that anosmia, or the loss of the sense of smell, may be a reliable early symptom of COVID-19, BEAT19 data show it is a minor symptom and isn’t necessarily only an early symptom.
Furthermore, anosmia is not a definitive COVID-19 diagnosis -- some COVID-positive participants never experience anosmia and some participants who reported anosmia tested positive for a different viral infection, such as influenza.
Shapiro said BEAT19 has collected but not yet analyzed outcomes data on individuals with vs. without cancer and other chronic illnesses.
xCures plans to launch the initiative soon in Brazil, where the BEAT19 protocol has been approved; the website is in the testing phase. Shapiro noted that information about COVID-19 symptoms from winter in the Southern Hemisphere could be informative for the upcoming winter in the Northern Hemisphere.
BEAT19 anonymized data is available through GitHub and the group is preparing a submission for publication.