Prognosia’s fast track to commercialization relied on St. Louis-based expertise

This month’s news that Seoul-based Lunit completed an acquisition of St. Louis-based Prognosia Inc. points to an important, yet sometimes overlooked, part of the local innovation landscape: a successful exit for a homegrown startup. 

Prognosia was established last September to commercialize research breakthroughs by Joy Jiang and Graham Colditz, professors at WashU Medicine who developed software that uses artificial intelligence to analyze mammograms and better predict a person’s risk of developing breast cancer within five years. The rapid-fire sale defies what’s normally a long road of commercialization for devices that need Food and Drug Administration approval—and shows how a community working together can get big ideas into the world more quickly. 

Colditz, who’s spent roughly three decades of his public health career on predicting and reducing breast cancer risk, explains the idea behind Prognosia came after the pair realized there was an opportunity to better analyze the data mammograms capture. He credits Jiang, who had experience in developing better ways to analyze and interpret data, with crafting an initial project that looked at just one element of a mammogram, breast density, and showed that it decreased more slowly in women who were developing breast cancer. 

“[A] screening mammogram has four images, literally 10s of millions of bits of data, and so we said, ‘We can do better than a one-dimensional summary,’” Colditz explains. “Joy developed a range of statistical methods and then ultimately processed mammograms to bring the whole mammogram in, not just a simple summary, and use that whole image to predict risk.”

The results their technology produced were also calibrated to established guidelines for the risk management of breast cancer, Colditz notes.

The software immediately struck a chord with radiologists at WashU and members of the university’s Office of Technology Management, who saw the immense value of bringing the tech into clinics, he says. Jiang even shadowed WashU radiologist Debbie Bennett to see how mammograms were used in a clinical setting.

“We had a dream that this technology really should be available to women anywhere they have a mammogram, [to] be able to get a copy and the risk [prediction] while they’re in a clinical setting,” Colditz says.

But the path through the Food and Drug Administration to get clearance for software as a medical device is challenging, and on top of that, medical systems generally need time to test and implement new software. Colditz recalls BJC outlining a six-month process to bring something like it into their own system. On top of that, as Colditz acknowledges, the journey was “way outside [the] training as a public health physician or statistician.”

To help with commercialization, Colditz and Jiang began working with David Smoller, an executive in residence with St. Louis-based BioGenerator, a nonprofit that helps develop and invest in early stage bioscience and agriculture companies. He stepped in to help with business planning, utilizing his experience growing multiple startups that were later acquired by public companies.

“We helped them get incorporated, build their pitch deck, figure out what their solution is, what their go to market [strategy] is,” he says of Prognosia. “We started to talk to venture capital groups about financing. BioGenerator came in early and gave them some small startup money.”

Prognosia formed last September and licensed the software from WashU. BioGenerator then helped with submissions to the FDA; WashU aided the founders in their search for a company CEO. While taking these steps Colditz says they also made a connection with Volpara, a software company with AI detection tools already deployed in thousands of clinical settings in the U.S..

“Initially, we had no idea whether they’d want to even talk to us,” he recalls. But Volpara, which Lunit acquired last year, did take an interest.

“We were talking to Volpara and Lunit, but because we didn’t know where that was going, we were working in parallel to make sure we had the potential to build a realistic pathway to clinic,” Colditz says. 

Eventually, Lunit decided it wanted to further develop Prognosia’s software as well, and issued a term sheet to acquire the company, though Smoller remembers pushing Lunit on its initial offer.

Smoller didn’t disclose the final acquisition terms, but described the purchase as “life changing” for Colditz and Jiang: “You know, we made millionaires out of these people,” he says. “Those two are great scientists and their product should get rewarded.”

Equally important was Lunit’s commitment to developing the technology and continuing to work with Jiang and Colditz, Smoller says.

“Lunit almost required that they become consultants, that together they would build this,” he says. “Because Joy and Graham are really key to this. It’s not just this algorithm.”

While Prognosia’s acquisition shows how St. Louis-based expertise can fuel a commercial breakthrough, it also points to something that Smoller says the local ecosystem needs to demonstrate more frequently if it wants to play in the big leagues.

“One thing we need in St. Louis that’s missing is more exits,” he says. “We talk about a lot of companies being started. Count the number that have exited.” 

And Smoller views this ability to return equity a key next step for St. Louis’ innovation landscape to become established like Silicon Valley or Kendall Square in Cambridge, Massachusetts. 

The acquisitions for companies that Smoller started all came within five years of their founding, he says. “That’s what investors want, a return on their investment in a period of time.”

Jim McCarter, senior managing director of BioGenerator Ventures, agrees. Even as a nonprofit, he says BioGenerator models potential exits and won’t invest in something if it doesn’t suggest favorable returns.

“You have to be thinking about the end at the very beginning,” he says. “How are you going to actually build a product that a consumer or user wants and then how does that eventually get you a return for any money that you take in investments?” 

A cycle generating consistent returns for local, national, or international investors can breed confidence among them to invest in St. Louis-based startups again, McCarter explains. It can also help keep talent from leaving.

“Just because a company is acquired by a company elsewhere, oftentimes the workforce stays here,” McCarter says. “It can [also] result in the refurbishment of the entrepreneurial talent, so somebody may exit a company and then they’ll start another one,”

It’s what Smoller hopes for Colditz and Jiang since Prognosia was acquired. Colditz, for his part, says he doesn’t view the acquisition by Lunit as a “one-time opportunity,” and that he and Jiang are already discussing what problem that clinicians face that they could tackle next.

“Can we find those sweet spots where what we develop will really translate to solving problems tomorrow?” he says. “It’s not just a really cool academic idea and a nice challenge, but it’s actually got a path to clinic to improve the lives of women.”