Healx has raised $56 million (€51 million) to launch 40 rare disease programs while taking some of its existing assets into the clinic. The ambitious target is enabled by an AI drug discovery platform that Healx thinks gives it a scalability that has more in common with tech than biotech.
David Brown, the former global head of drug discovery at Roche, and Tim Guilliams cofounded Healx in 2014 to use AI to find opportunities to repurpose existing molecules to treat rare diseases. After quietly working on the technology in its early years, Healx raised $10 million last year to build on its early success in identifying a potential treatment for Fragile X syndrome.
That money was due to see Healx through to 2020 but Guilliams, who works as CEO, decided to pull the financing forward in light of the success rate of the company’s predictions and translations.
The upshot is Healx is equipped to begin multiple clinical programs in 2020, starting with an adaptive phase 2 trial that will assess multiple combination treatments for Fragile X syndrome. Healx is also set to start clinical trials of assets from some of the other 10 programs it has launched to date.
At 10 programs, Healx’s pipeline is broad for an early stage, 41-employee company. That is just the start, though. Using the series B funds, Healx plans to initiate another 40 programs.
"We know those are big numbers and it's ambitious. But from a preclinical perspective and a data and machine learning perspective, our approach is actually scalable,” Guilliams said. “Making the predictions, selecting the rare diseases, selecting the drugs—we can do this at a larger scale."
The scalability stems from the fact Healx takes a nontraditional approach to drug discovery. Rather than start with a particular druggable target, Healx turns its algorithms on data and leaves them to decide which rare diseases to pursue with which drugs.
“It's only afterward when we look at the output that we try to understand the mode of action and the principles of why those treatments have been predicted for that disease. By doing so, we come up with novel biology,” Guilliams said.
Healx is far from the first startup to claim to have an AI system capable of repurposing drugs, thereby slashing the cost and duration of early stage R&D. Yet, the idea is yet to lead to approved products or reshape R&D in the way its advocates predict. Healx’s belief that it can validate the concept rests in part on its experience launching its first 10 programs.
“For every program we've launched we've identified successful candidates, either in monotherapy or combinations,” Guilliams said.
The true test will come once Healx starts to generate clinical data but its performance to date has enabled it to attract significant sums from VCs that normally back technology startups. Atomico led the series B with the support of investors including Balderton Capital, which led the series A.
Healx will use the money to double its headcount, taking on more pharmacology and clinical development staff to complement its currently tech-skewed workforce. The startup lacks in-house labs, choosing instead to rely on third parties overseen by former Charles River Laboratories VP Kate Hilyard.
“When you walk into the office its more like Google than a traditional biotech,” Guilliams said.
The growing team is seeking applications from rare disease groups that want to work with Healx on its next crop of programs. All of the current programs are affiliated with one or more patient groups and that experience has convinced Healx to follow the same model as it scales up its pipeline.
"It's a real partnership,” Guilliams said. “They bring disease understanding and knowledge. A lot of them have developed patient registries. That's incredibly valuable, particularly if you start using machine learning around that. They help open doors, they help talk to the clinicians and the clinical trial centers, they help with the clinical protocol and they act as disease experts.”