Queright has raised an $8.5m Series A round led by Hargreave Hale AIM VCT.
The Cambridge-based biotech Qureight uses a proprietary AI platform to help speed up the drug development process by making clinical trials and data analysis more efficient.
Investors also included XTX Ventures, Guinness Ventures, Playfair Capital, Meltwind, Ascension Life Fund and Cambridge Angels.
“In really simple terms, it's about how to get drugs through the approval process,” says CEO and cofounder Dr Muhunthan Thillai.
“There are lots and lots of companies using AI to make drugs — but our question has always been once you've made a drug, whether it's through AI or in a lab, how do you get that drug from being tested in five patients right through to 5k?” he says.
Clinical trial bottlenecks
When researchers start a clinical trial for a new drug, they’re faced with a data dilemma: patient information is typically stored across multiple platforms, making it difficult and time-consuming to find and analyse the most important information.
Qureight’s platform aims to tackle that issue by collating the valuable data into one place to be analysed by AI models, which can help researchers make more informed decisions quicker than if they were analysing the results themselves. It’s hoping to tap into the lucrative clinical trials market, which is globally projected to reach over $92bn by 2030.
“Our cloud-based platform works with pharma companies, hospitals and contract research organisations to put all the data in one place. It allows us to build machine learning models, or allows other partners to build models, to interrogate that data better and answer one question: is this drug working or not?” says Thillai.
Qureight currently focuses on lung and heart diseases but hopes to branch out into pulmonary hypertension and lung cancer soon.
The analysis produced from its platform can be used to find out how patients have responded to the drugs and work out if that treatment should be tested in further trials.
“[Pharma companies need] a lot of extra information to make that crucial go/no-go decision. They might have spent $20m on a small trial, then invest $100m in the next stage — so giving them those insights is crucial to helping them make those decisions,” says Thillai.
It currently holds R&D contracts with five NHS England trusts, which grant access to anonymised patient data to improve Qureight’s ML model in return for the NHS using the platform for its own research or commercial purposes.
It’s also secured strategic partnerships with pharma companies AstraZeneca and Vicore Pharma. Sweden-based Vicore Pharma used the platform for a clinical trial last year, says Thillai, where Qureight could “effectively predict” which patients would do well on the drug, based on an imaging model it created.
The future of trials
Another possible use for Qureight’s platform is using the data it collects to develop “fake” virtual trial patients as a more affordable alternative to recruiting people.
“If you're a small pharma company and you've got only enough money to run one drug, we might be able to create a synthetic trial to see what would have happened if you'd run that trial in the real world,” says Thillai.
While the virtual patient aspect is still in beta testing, Thillai is hopeful that it could become a reality for clinical trials before too long.
“Part of our roadmap is to do a lot of work with pharma companies and go through an official FDA approvals process — so in the next 18 to 24 months, virtual trials concerning diseases might be approved,” he says.
Improving diversity within trials is also a major focus, says Thillai. The virtual patients can be customised to represent a diverse range of demographics, and he hopes that using some of the funding to branch out outside of the UK and Europe for data will help enable more representative testing.
Qureight recently signed a partnership with the King Faisal Hospital in Saudi Arabia.
“The NHS is amazing and we have amazing data partnerships here,” says Thillai — “but if we're going to work with global pharma companies, we really need different types of data.”
“Models that work in the UK might not be as good in the Middle East, Asia, or the US. So we really need as many different types of data as we can to build the strongest models,” he adds.
“It’s not just from an ethical point of view, but also from a commercial point of view: these pharma companies need to sell their drugs worldwide, not just to the West.”
