AI suggests Australia may have backed a COVID-19 vaccine winner Yolanda RedrupReporter
Sep 8, 2020 – 2.24pm
Save
Share
A new analysis of the world's leading COVID-19 vaccines and therapies in development has revealed that even the small group of vaccines already in phase three trials have only a 20 per cent chance of success.
The artificial intelligence-powered analysis from local health informatics and predictive analysis company Opyl has found that mRNA-based vaccines are the most likely to succeed.
Opyl's Michelle Gallaher believes that by using AI, researchers will be able to bring COVID-19 therapies and vaccines to the market faster.
It has identified the leading candidates as those belonging to the US National Institute of Allergy and Infectious Diseases (NIAID) followed by the collaboration between BioNTech and Pfizer.
The third-most likely candidate of those already in phase three trials was Oxford University and AstraZeneca's vaccine, which uses a different approach centred on a non-replicating viral vector.
Opyl chief executive Michelle Gallaher told The Australian Financial Review this analysis was good news for the Australian government, which locked in almost 34 million doses of the AstraZeneca vaccine on Monday.
Advertisement
Some of the positive factors supporting the vaccine were the number of trial locations (approximately 20), trial participants (more than 12,000) and its randomised design. The chance of the trial succeeding could be improved by increasing the number of countries where the trial operates.
Ms Gallaher, who has spent 30 years in the biotech and pharmaceutical industry, and has also taken part in clinical trials, hoped the research would help governments and investors around the globe to pick which targets to support.
The Australian government has backed a good target here, but ... they should consider diversifying a little bit and maybe also backing an RNA vaccine as a back-up.
— Michelle Gallaher, Opyl chief executive
"There is value for governments in doing early procurement agreements ... every government wants to secure an opportunity to access vaccines or treatments at the earliest stage possible," she said.
"But it's really hard [to pick which ones will be successful]. In phase three, they're all good horses in the race, but one will win ... and candidates do still fall over.
"The Australian government has backed a good target here, but what we're also saying is they should consider diversifying a little bit and maybe also backing an RNA vaccine as a back-up."
Opyl's model, which examines 475 COVID-19 clinical trials globally, takes into account numerous factors such as the design of the trial, the number of patients enrolled, head researchers, the organisation's experience, intervention type, and the record of the hospitals and clinical trial companies in being involved in successful vaccines. It does not have access to any private data.
Opyl is updating its model each week to incorporate the latest data on vaccine trials globally and the company will continue to reassess which candidates have the greatest chance of success as more enter the phase three trials.
Across all the vaccine candidates in development, Opyl determined there was only a 5 per cent to 17 per cent chance that they would be successful.This, however, is higher than the usual 3 per cent success rates for vaccines.
As well as helping governments and investors determine what vaccine efforts to further invest in, Ms Gallaher hopes to help researchers use data to create better design protocols for their trials and recruit more patients to increase the probability of their success.
Opyl's mission gained extra importance for Ms Gallaher this year when her uncle in the UK died of COVID-19.
"He was in one of the biggest hotspots ... and they didn't even treat him. He was sent straight to palliative care. He was 73 and we were just shocked," she said. Rating the chances of success
"He was really active and went sailing and played golf, but he did have Parkinson's disease and that comorbidity put him in a vulnerable place."
While the vaccines still have relatively low odds of success, Opyl's AI algorithms have revealed a 20-30 per cent chance of success for the most progressed therapy trials. Examining all the treatment efforts under way, there is a 75 per cent likelihood of success across the group.
The therapies most likely to succeed are antivirals. Of these, Gilead and NIAID's trials of Remdesivir and PTC Therapeutics' trial of an oral drug called PTC299, which has a dual-mechanism of potentially blocking the viral replication and the inflammatory response that occurs after infection, are the leading contenders.
Antivirals are followed by antibody-based therapy trials, with the likes of Genetech/Hoffmann-La Roche's Tocilizumab trials, GlaxoSmithKline's Otilimab study, NIAID's ACTT study, Eli Lily's Baricitinib trial and Novartis Pharmaceuticals' Canakinumab having the highest odds of success, based on Opyl's current data analysis. RELATED
CSL to manufacture 81m COVID-19 vaccine doses by mid-2021
Opyl, which is listed on the ASX, was originally known as Shareroot. It went through a restructuring in February last year, which resulted in Ms Gallaher taking on the CEO role and changing the direction of the business.
She said COVID-19 had provided the perfect test case for Opyl to trial its AI algorithms and it hopes to have a platform on the market for commercial adoption within the next year.
OPL Price at posting:
19.5¢ Sentiment: None Disclosure: Not Held
(20min delay)