Beating the odds by sticking to the serial strategy — Vertex’s Altshuler on The BioCentury Show

How Vertex is translating its success from CF into new areas, picking indication by indication, while leveraging AI

Vertex’s approach to research, defined over a decade ago to beat the dismal odds of success in biotech, remains core to its strategy, even as the industry evolves and technologies expand. While that strategy, which EVP and CSO David Altshuler calls “serial innovation,” has seen the company launch eight drugs since 2011 and jump market cap tiers on the success of its cystic fibrosis portfolio, it’s now being tested in other areas as Vertex expands to pain, renal disease, Type I diabetes, and more.

On The BioCentury Show, Altshuler discussed the rising interest in causal biology and the role of AI, noting that the fundamentals of picking targets and programs have not changed.

He said that when he was recruited to Vertex Pharmaceuticals Inc. (NASDAQ:VRTX) by then-CEO Jeff Leiden, who is now the chair, the goal was “not to just discover one medicine and market it, and not just one disease, but a second, and a third,” and that required beating the regular odds in R&D. “If you only work on the causes of human biology as documented in people, you’re more likely to succeed,” said Altshuler.

While that’s become a mantra across much of R&D these days, many of the companies he collaborated with prior to Vertex, while he was an academic and physician, were working on projects that were “based on mice or cells in a dish or a model of some kind.”

“You won’t succeed 100% of the time,” Altshuler added. “For many of those cases where we understand the cause of disease, but there’s not yet a treatment, it’s because the technology to solve it isn’t obvious.”

Altshuler believes that AI will have a big impact, but is one more tool in the toolbox and won’t change the fundamental strategy.

“Clearly AI is a game changer,” said Altshuler. The issue is when and where to implement it. 

“One thing I’ve learned, having been now CSO for 10 years, is it’s not one thing. It’s 500 things you have to do right, and AI can help with multiple of them, but it can’t help with others. And so if all you think about is the tool, you might accelerate one piece. But you’re actually not going to accelerate the end point because you haven’t solved the other problems,” said Altshuler.

He added that Vertex is using AI in multiple places across the R&D continuum to build on rather than replace what its researchers can do.

For example, in the pain and cystic fibrosis programs the team knows a lot about the proteins and how they work, and has thousands of molecules with structure-activity relationships. “We use AI to try and say, what can we learn from those, and how can we design new molecules? But if we didn’t have all that expertise and if our people didn’t have the right assays and knowledge of where to use it, I don’t think it’d be that helpful.”

In clinical trials, AI can be useful for identifying patients, as well as for streamlining paperwork. “When you have a protocol for your trial, and then you have, let’s say 200 sites, there’s a lot of forms you have to fill out with the protocol. And up until recently, we would have human beings do it and then human beings would check it,” Altshuler said. Now, human beings still check the final product, but there’s a lot less manual work. 

“It doesn’t replace the human being; it’s not going from eight hours to five minutes, but it makes those individuals more efficient,” he said.

Altshuler also discussed the importance of the academia/industry interface, and how Vertex picks its programs beyond CF in Type I diabetes, Duchenne muscular dystrophy and more.