What went wrong?

The University of Queensland’s vaccine will not proceed to further clinical trials because every single participant injected with it tested weakly positive on certain HIV screening tests.

Because it borrowed a protein from HIV,that was always a risk for this particular vaccine.

Given that,and the fact several other promising vaccines are close to approval,the university and manufacturer CSL made the heartbreaking decision this week to essentially shut down the project. The current phase 1 trial will be allowed to continue,but the vaccine will not progress to phase 2 trials.

“I think there is a single word that sums it all up:devastated,” said Professor Paul Young,co-lead of the university’s vaccine team.

“The last 24 hours have been particularly difficult for the team. It’s challenging times,but that’s science. It does not always work.”

The clamps

The University of Queensland vaccine is built on a novel technology invented in Professor Young’s lab,which they call themolecular clamp.

Subunit vaccines,like this one,contain a small fragment of virus. The immune system recognises the virus and learns to neutralise it. In the case of COVID-19,that fragment was the virus’s spike protein.

These vaccines face a key challenge.

Spike proteins change shape when they bind with human cells,opening and fusing with the cell wall to inject viral genetic code.

The immune system will recognise this foreign protein and generate neutralising antibodies.

But,crucially,these antibodies will be made for the opened spike protein.

That means they will not be effective at blocking the closed spike protein on the virus - which they need to do to stop infection.

The molecular clamp was designed to solve this issue. It uses a wrapper of protein,wrapped around the base of the spike,to hold it in its closed shape. That allows the immune system to generate a strong response.

Unfortunately,the best wrapper the University of Queensland team found happened to come from the human immunodeficiency virus – HIV.

“The reality was,we tested several different constructs in the earliest days of development. The HIV fragments were providing the highest levels of stability,” said Professor Young.

This wrapper protein,known as GP41,wraps up the proteins HIV uses to infect cells into a neat bundle. Professor Young's team found they could extract it and use it to do exactly the same to COVID-19's spike protein.

A weak false positive

Human immune systems are built to generate antibodies to any foreign virus they encounter.

Designing a vaccine that included a fragment of HIV without making it trigger HIV antibodies was always going to be difficult.

To do it,the University of Queensland team carefully removed the known sites on the HIV protein that antibodies bind to. The hope was this would mean the vaccine would generate no HIV antibodies – just antibodies to COVID-19.

“They have tried to stop that by getting rid of the bits that are known to be the most likely to trigger that immune response,” said University of Queensland virologist Associate Professor Ian Mackay.

“It just turns out that we don’t know everything."

In early testing,it seemed to work. But the challenge was always going to be humans.

Standard HIV tests do not work on animals,meaning the scientists could not check if their vaccine was generating HIV antibodies before human clinical trials started.

The 216 phase 1 trial participants were told HIV false positives were a risk when they signed on,and their blood was closely monitored for signs of HIV antibodies.

Every single participant who had been given the vaccine tested weakly positive on certain standard HIV tests that are built to look specifically for antibodies to the HIV wrapper protein – known as glycoprotein 41.

Many of the positives were weak. “We’re right on the cusp” of detection,said Professor Trent Munro,director of the university's vaccine program. But they were there.

It’s not yet clear just why these antibodies showed up,given the team painstakingly removed likely antibody binding sites from the HIV protein wrapper.

Professor Heidi Drummer,group head of viral entry and vaccines at the Burnet Institute,described GP41 as"particularly immunogenic"- meaning it tends to create antibodies.

Hiding it from the immune system while ensuring it still held the spike protein in the right shape was a serious scientific challenge,she said.

“Sometimes when you do this,you hide one region and another becomes immunodominant. It’s like plugging holes in a leaking ship – another leak springs up somewhere else."

Everyone involved in the project is at pains to stress this is no way poses a risk to the health of the participants.

They do not have HIV.

Prompted by the vaccine,their immune system has generated a natural defenceagainst HIV. It is just unfortunate that this natural defence is exactly what common HIV tests look for.

The team hoped the levels would drop quickly,but they did not. They explored other options with HIV specialists. But in the end it was decided that having the entire population of Australia test positive on a standard HIV test was not an acceptable outcome.

A second point weighed heavily in the decision to cancel:the success of other vaccines. Pfizer’s mRNA vaccine has already won approval in Britain and Canada,and there are several other promising candidates approaching the finish line.

“If none of the other vaccines had been as successful ... we might have recommended to government that this vaccine proceed,” said Dr Russell Basser,senior vice-president of Seqirus,CSL’s vaccine arm.

Are the results of this phase-1 clinical trial a surprise?

Yes and no.

It is a shame this project – on which millions of federal and state government dollars were spent – has not worked out.

But scientific research projects often fail,and they often fail at different steps along the way. There are no guarantees. Really,the amazing thing is we have not seen more COVID-19 vaccines fail clinical trials.

New vaccines and drugs have to go through three phases of clinical testing before they can be approved,with a sharp drop-off rate at each stage. Last year, US researchers looked at 406,038 clinical trials between 2000 and 2015,and found that only a little over half of vaccines studied ever made it to phase 2 – and only 33.4 per cent ever reached approval.

A closer look at the University of Queensland’s vaccine shows us why there was always a lot of uncertainty around this project.

The vaccine is built onmolecular clamp technology,a new idea that tries to address an old problem with subunit vaccines – vaccines that contain a small sliver of a virus that you want to train the immune system to neutralise. In the case of COVID-19,that’s the spike protein.

Once injected,these proteins tend to lose their shape,meaning the immune response is not perfect. The University of Queensland’s big idea was a set of clamps to hold the protein in just the right configuration to generate a strong immune response.

Molecular clamp vaccines are entirely novel. Such a vaccine has never been made before. Indeed,before COVID-19 reared its head,the UQ team expected to spend the next few years working on their prototype. Instead,what should have been a proof of concept became seen by the public as a pandemic-buster. Compare that to mRNA vaccines,which scientists have been working on for decades.

For such an early-stage project to not work perfectly on the first try does not come as too much of a surprise.

How important was the University of Queensland-CSL deal? Will this leave us unprotected?

It was important without being crucial. The UQ-CSL deal was a major investment,and if it had worked would have delivered 50 million doses of vaccine in 2021.

The biggest advantage UQ’s vaccine enjoyed was it could be made here by CSL. Unfortunately,the company does not have the technology to make the cutting-edge mRNA vaccines that currently lead the race,meaning they will all have to be manufactured offshore and shipped in – when we can get them.

But there was always a good chance this vaccine would fall over. Australia’s government,to its credit,has built a strategy that was not reliant on it succeeding. And UQ’s vaccine was not a global front-runner;indeed,because it was such new technology it was running a little behind the pack.

That’s the key takeaway here:the UQ vaccine would have been nice to have but its failure does not leave us unprotected.

What else is in the pipeline?

In response to the cancellation of the contract,the Australian government announced it would increase orders for the AstraZeneca and Novavax vaccine. The government claimed this was likely to mean Australia would have earlier access to a vaccine.

Prime Minister Scott Morrison said this:"We are increasing our production and purchase of AstraZeneca vaccines from 33.8 million to 53.8 million,and we're increasing our access to the Novavax vacine from 40 million to 51 million – so that's an extra 20 million doses of AstraZeneca,and an extra 11 million doses of Novavax. The AstraZeneca vaccines,of course,are manufactured here in Melbourne by CSL."

Apart from the UQ-CSL option,Australia has signed four other separate agreements for COVID-19 vaccine supply.

We already have 10 million doses – probably enough for 5 million people – of Pfizer’s mRNA vaccine on order,due to arrive in 2021. That vaccine is the clear frontrunner and has already achieved regulatory approval in Britain and Canada;Britain administered its first dose on Tuesday.

Pfizer’s vaccine is likely to need two shots,and must be transported at minus-70 degrees. The company has designed sophisticated dry-ice-filled “Eskies” to ship the jabs around the world,but logistics are going to be a major hurdle for this vaccine.

Then we already have 3.8 million doses of AstraZeneca’s vaccine due in early 2021,with CSL holding a contract to make another 30 million doses in Australia over the course of 2021.

The trial data on that vaccine isspotty. The developers initially claimed a 90 per cent protection rate,but later admitted that was only when they accidentally gave the wrong doses;people given the correct dose achieved about 62 per cent protection. More studies need to be done – but the vaccine did seem to be wholly protective against severe life-threatening COVID-19,data published inthe Lancet on December 8 show.

Our third vaccine purchase agreement is with American developer Novavax for 40 million doses of vaccine over the course of 2021,with an agreement for an extra 10 million doses if needed.

Its clinical trials are running a little behind AstraZeneca and Pfizer. Two of its three phase-3 trials are running but the company does not expect them to report data before early 2021,while a third trial is yet to commence, the company said on November 30.

Compared to these vaccines,UQ’s was already some way back in the race.

We also have access to an international agreement called COVAX,which lets us purchase a broad range of vaccines if they succeed. But it’s not clear how useful that agreement actually would be to us – it really exists to ensure less-wealthy countries have fair access to a vaccine.