It’s no secret that drugs don’t work equally well for everyone. Doug Cole, chairman and co-founder of Vesalius Therapeutics, saw these limitations while he was a practicing physician in the 1990s. Even if a doctor diagnoses patients with the same illness, that condition leads to differences in symptoms and disease progression from one individual to the next, he said.
Generations of physicians have known about these variations, but they haven’t been able to do much about them because drugs are largely one size fits all, Cole said. Vesalius wants to change that. The startup is using artificial intelligence to analyze data, finding patterns in disease in order to identify groups of people who share disease biology. With that information, Vesalius aims to develop drugs that treat particular groups of patients. The Cambridge, Massachusetts-based startup launched on Wednesday backed by $75 million from Flagship Pioneering, the venture capital firm that has been incubating the young company for the past two years.
Applying AI to understand disease and guide drug development is becoming more common these days, but much of the AI efforts are for rare diseases, conditions where a genetic mutation can be identified, enabling scientists to focus on a particular target. That approach applies to about 10% of diseases, Cole said. Vesalius is developing drugs to address the other 90%. Drilling down into disease to find groups of patients has already been tried with success. Cole pointed to cancer, which decades ago was widely treated with radiation. Research over the years has developed more targeted therapies, treatments that work on patients who have a particular genetic makeup.
“That’s in some ways, the precedent for [Vesalius’s approach],” he said. “How this gets applied to diseases where you can’t find a single genetic cause is something that people haven’t really figured out yet.”
Vesalius is trying to figure that out. The differences a common disease has from one person to the next does have a genetic basis, but rather than being rooted in a single genetic mutation, those differences come from groups of genes that Vesalius calls genetic circuits. In many common diseases, these circuits aren’t functioning properly, said Christopher Austin, the former founding director of the National Center for Advanced Translational Sciences at the National Institutes of Health and now Vesalius’s CEO. Vesalius analyzes data from a wide range of sources. Clinical data show what disease a patient had. Billing codes offer insight into the symptoms. Genetic data reveal the biology of the disease.
A Deep-dive Into Specialty Pharma
A specialty drug is a class of prescription medications used to treat complex, chronic or rare medical conditions. Although this classification was originally intended to define the treatment of rare, also termed “orphan” diseases, affecting fewer than 200,000 people in the US, more recently, specialty drugs have emerged as the cornerstone of treatment for chronic and complex diseases such as cancer, autoimmune conditions, diabetes, hepatitis C, and HIV/AIDS.
All of the data are crunched in Vesalius’s AI-driven technology platform, called Diamond. The analysis separates patients into different groups based on their disease biology. With that information, Vesalius can then run experiments to screen for molecules that have the potential to restore the function of genetic circuits. Vesalius has not specified which diseases it aims to treat, but Cole said examples of common diseases include diabetes, Alzheimer’s disease, and various cardiovascular disorders. Accordingly, the drugs the startup develops could be various types, such as small molecules or antibodies, depending on the best way to intervene and restore the function of a genetic circuit in a particular group of patients.
Identifying the patient population for a drug tends to come later in the drug development process, Austin said. Typically, drug research starts with a protein or a gene that scientists want to target. Years later, when a therapeutic candidate is being readied for the human testing, scientists then think about which patients to enroll for the clinical trial.
“In our view, that’s really backward from what you ought to do,” Austin said. “What you ought to do is define up front who your patient population is going to be. As someone who has done a lot of recruiting for clinical trials, I think we’re going to have a lot more success recruiting for these disorders.”
The launch of Vesalius comes at a time of a proliferation of data along with the emergence of new technologies to analyze them. It’s not simply a matter of analyzing the data, Austin said. The experiments that Vesalius runs produce additional data that must be analyzed. It’s only fairly recently that the technological tools to perform these analyses have been up to the task, he said.
Vesalius is the latest company to emerge from Flagship Labs, the venture capital firm’s startup incubator. Senda Biosciences, a Flagship startup developing drugs based on AI analysis of how humans interact with bacteria, launched last summer. Other Flagship-founded companies bringing AI to drug discovery include Generate Biomedicines and Valo Health.
The main focus of Vesalius is developing its own drugs. But the startup’s technology also offers the opportunity to help other drug companies. For some currently available drugs, as well as many failed ones, large groups of people do not respond to treatment. Austin said that Diamond’s ability to identify groups of patients that are likely responders may be able to revive or resurrect drug programs. The company has not yet used Diamond for such applications, but Austin said that the company is now looking to partner with drug companies for that purpose.
“It’s on our to-do list, maybe as early as later this year,” he said.
Photo: metamorworks, Getty Images
