Predicting Covid-19 vaccine approvals with Opyl

SARS-CoV-2 and the disease it causes, Covid-19, has severely disrupted the world for nine months now. Although public health measures, such as social distancing, wearing masks and rigorous hand washing, have undoubtedly tempered the impact of the viral disease, these are not long-term solutions to the pandemic. Instead, they have been employed as short-term ways to control the spread of the disease while the world waits for a vaccine. 

Although technically two vaccines have been approved - Russia’s Sputnik V and Chinese CanSino Biologics’ candidate – this is only for emergency use and there remain significant safety concerns due to limited clinical trial data. 

As the pandemic wages on, with limited signs of easing, the big questions being asked are when will we have a safe and effective vaccine for Covid-19? And which of the candidates in development are most likely to approved? 

Launched in 2019 and named after Australian opals, Melbourne-based health data analytics company Opyl believes it has the answer. 

By leveraging publicly available data from global registries on 31 Covid-19 clinical trials, Opyl CEO Michelle Gallaher explains that her company has conducted a study that applies its artificial intelligence (AI)-powered technology to determine the probability of success for each Covid-19 vaccine candidate and then rank them in order from highest to lowest. 

To support the detailed and insightful protocol data from Covid-19 trials and because all respiratory vaccine trials have consistent and similar protocols, Gallaher notes that Opyl also integrated protocol and outcome data from previous respiratory and viral vaccine trials to its predictor engine.

In carrying out these AI predictive analytics, Opyl found that the probability of success for vaccines ranged from 5% to 17%. However, this increased to an average of 20% for those candidates in Phase III, of which there are currently 11, according to the New York TimesTracker. 

Those at the higher end of the scale were RNA vaccines. These vaccines were the only options to have any candidates with a greater than 10% probability of achieving one primary endpoint with statistical significance in the trial. 

Moderna’s mRNA-1273 and BioNTech and Pfizer’s BNT162 are two RNA vaccine candidates that look particularly promising to Opyl. Both are mRNA vaccines, currently in Phase III and II/III trials respectively, and their executive teams are hoping to advance them towards US approval by the end of this year. 

It is particularly interesting that two mRNA vaccines appear to be the most promising given that this is a new vaccine technology that has never been approved before. However, this novel and innovative approach has “the promise to revolutionise the field by addressing current manufacturing challenges and offering novel vaccine compositions,” as stated in a Nature article.

A piece in the Journal of the American Medical Association notes that mRNA vaccines are much quicker and cheaper to develop than traditional approaches – with a possible scalability benefit – as well as potentially being better at conferring the necessary, longer-term T cell immune response to the virus. 

The next most promising vaccine candidates were non-replicating viral vector approaches with AstraZeneca and the University of Oxford’s ChAdOx1/AZD1222 leading the pack. AstraZeneca’s vaccine relies upon a chimpanzee adenovirus vector to introduce a portion of the SARS-CoV-2 virus into the patient. By using a chimpanzee adenovirus, rather than one that affects human, AstraZeneca hopes to avoid the pre-existing immunity issue, where your immune system clears the vector too early, that other non-replicating viral vector vaccines have faced. 

Interestingly, Gallaher notes that these frontrunners have not changed in the time since Opyl first ran its model back in August. 

Opyl’s findings are in-keeping with other analyses of the Covid-19 vaccine race. For instance, all three are included in both the Operation Warp Speed initiative that aims to accelerate the development, production and delivery of millions of doses of Covid-19 vaccines. The UK Government has also bet on the success of AstraZeneca and Pfizer’s vaccines by including them in a similar Vaccines Taskforce.

In addition to looking at Covid-19 vaccines, Opyl has evaluated 444 Covid-19 therapeutics and devices trials. In fact, what is particularly interesting to Gallaher is that, according to Opyl’s analysis, “therapeutics have a much higher chance of success and they're far more likely to appear on market”. 

They had between 20% and 30% probability of success, which is very favourable compared to the maximum of 17% for vaccines. 

Gallaher notes that she is concerned that “much social and economic policy hinges on a vaccine being available”, when the most likely outcome is that a combination of therapeutics will be available to treat Covid-19. These results, therefore, explain why lots of expert scientific policy advice throughout the pandemic has hinged around therapeutics. 

The leading candidates identified by Opyl were Gilead’s remdesivir and PTC Therapeutics’ PTC299. Other candidates that Opyl picks out as promising are anti-inflammatory drugs Roche’s Actemra, Eli Lilly’s baricitinib and Novartis’s canakinumab, as well as GlaxoSmithKline’s antibody drug otilimab.

Remdesivir has received significant attention since the pandemic began. Originally developed to treat Ebola, Gilead is studying this anti-viral drug for patients hospitalised with Covid-19. The drug has received emergency authorisation by US and European authorities. 

However, the efficacy of remdesivir has recently been called into question by the large, international SOLIDARITY trial carried out by the World Health Organization. It found that remdesivir, as well as other experimental drugs for Covid-19 including hydroxychloroquine, lopinavir/ritonavir and interferon, had little or no effects on mortality, initiation of ventilation and duration of hospital stay in hospitalised Covid-19 patients.

There have been significant efforts to combine remdesivir with other agents to improve efficacy in hospitalised Covid-19 patients. The drug has been combined with many of the drugs noted by Opyl, including baricitinib and Actemra. 

Repurposed from oncology indications, PTC299 is currently in Phase II/III clinical trials of adults hospitalised with Covid-19. This drug has a differentiated dual mechanism of action against SARS-CoV-2; it blocks both viral replication and mitigates the over-reaction of the immune system caused by the virus, known as the cytokine storm.

Controlling the cytokine storm is also the element of Covid-19 that Actemra, baricitinib, canakinumab and otilimab are aiming to combat. This is the focus of most drugs in development for Covid-19 because this hyper-inflammation is the primary reasons that SARS-CoV-2 is so deadly. The virus has killed more than one million people globally to date. 

Although Opyl has used Covid-19 trials as proof-of-concept for its newly launched predictive analytics platform, Gallaher has big plans for the company’s future. 

Importantly, Opyl isn’t solely focused on the probability of success, but also on the variables themselves causing that probability. In the long-run, Gallaher hopes Opyl will work inside pharma companies and use these insights to support decision-making by clinical trial designers. 

Poor trial design is a major cause of study failure, so Gallaher hopes Opyl’s technology can “give some comfort to pharma companies that they are designing the trial and putting all the resources behind a trial that has the best chance of success”. 

In the short-term, Gallaher is focused on Opyl’s next project, which will be looking at cell therapy trials, particularly in the oncology space. The reason why she chose cell therapies is because it is both where her personal expertise lies and an area of strength for Australia.