Using artificial intelligence, BioVista Inc. created a drug profiling platform that can identify new uses for existing compounds, providing the start-up with dozens of options for internal development as well as partnerships and financing.
Charlottesville, Va.-based BioVista began as a research project run by brothers Aris and Andreas Persidis. Aris boasted a biotech background that included co-founding Anadys Pharmaceuticals Inc. and Cellzome Inc., while Andreas brought artificial intelligence expertise gained at software developer Delcam International.
About 15 years ago, the brothers set out to apply AI and engineering concepts to the life sciences. Andreas Persidis explained that merging ideas from two disparate fields often provides out-of-the-box solutions, pointing to how the fix for a leaking heart valve came from submarine design concepts.
But it wasn't until early 2008 that BioVista vaulted from research project into corporate mode. That's when the brothers and their team discovered how to endow their data management platform with an awareness of concepts like drugs, diseases, side effects, organs, genes, biological pathways and more. They then applied the technology to 8,000 diseases, 12,000 adverse events and 15,000 drugs from the world's formularies, using every scientific publication, conference presentation, FDA report and patent filing available.
It would be easy to dismiss BioVista's technology as data mining, but Andreas Persidis explained that it is "not a case of reshuffling existing information, but creating new things."
Biotech and pharma companies apparently understand the distinction. BioVista has eight ongoing projects that involve using its technology to provide discovery, repositioning, adverse event profiling, patient stratification and other services for partners. All of the company's deals involve not just fees but milestone payments and future royalties, Aris Persidis said.
Yet BioVista is more than just a service provider. The company wanted to "practice what we preached" and develop an internal pipeline, Andreas Persidis said. So the BioVista team selected unmet medical needs and applied its platform to identify generic drugs that might find a new use meeting those needs.
The strategy resulted in an internal pipeline comprising more than 30 repurposed generic compounds. None of the molecules or original indications has been disclosed.
Lead drugs BVA-101 and BVA-201 are in preclinical studies for multiple sclerosis. BVA-201 has shown positive results in mouse models, including a statistically significant reduction of experimental allergic encephalomyelitis severity comparable to that of dexamethasone. BVA-201 also protected neural axons and myelin against degeneration.
Another star of BioVista's pipeline is epilepsy drug BVA-601, which has demonstrated anti-epileptic and neuroprotective activity in mouse models. The company also has drugs in various stages of research and preclinical development for eye disorders, diabetes and obesity, oncology, autoimmune and infectious diseases and other central nervous system disorders.
BioVista said partnering discussions are ongoing for BVA-201 and BVA-601, but the company has not yet determined how far into preclinical and possibly clinical studies it is willing to take its compounds before out-licensing them. Since the drugs in its pipeline are generic, BioVista anticipates reformulating and redosing the compounds as appropriate to fit their new indications and provide differentiation.
Beyond generics, BioVista also is evaluating branded and even experimental drugs with its platform and partnering with the innovator companies to explore the new indications it identifies.
"Our strength is a prolific discovery capability," Andreas Persidis said, likening the technology to the "goose that lays golden eggs."
All of that partnering has provided BioVista with a stream of income, but the company also recently completed its first outside funding round. In mid-2008, BioVista began approaching money managers about angel funding, and the company closed a seed round in eight days. The amount of funding was not disclosed, but is expected to last about two years.
BioVista started out predicting side effects, and such services are still a part of its business model. But Aris Persidis noted that "a side effect can also be a mirror image of a new indication," and the company's discovery business is where it expects to both provide and tap into the greatest value.
Between 30 percent and 45 percent of drug approvals over the last few years have been expansions of existing products into new indications, but most new uses are discovered by accident or are obvious extensions, such as related tumor types, Aris Persidis said. He added that BioVista is the first company to "promote the art and science of repositioning systematically."