But the researchers have emphasised there were no health risks to the trial participants and there was no risk of the participants developing HIV.

In a statement to the stock exchange, CSL said the trials would not move past phase one to phase two and three, after an agreement with the Australian government.

While the vaccine had been found to be safe, and had produced antibodies against the virus, CSL and the University of Queensland said it was necessary not to go further because if the vaccine was to be rolled out more widely, it would require significant changes to HIV testing procedures.

The Australian government had committed to buy more than 50 million doses of the University of Queensland vaccine, which was to be manufactured in Australia by CSL.

That deal would now be scrapped, leaving Australia with deals with three other manufacturers, and as part of the international COVAX facility.

On Friday morning Prime Minister Scott Morrison said Australia didn't expect all the vaccines it had signed deals for to be successful.

"At no stage, can I assure you, that we believed that all four of those vaccines would likely get through that process. If that had occurred, that would have been truly extraordinary, based on the process of vaccine development not only in this country, but anywhere else," he said.

"So that's why we spread our risk. That's why we backed important projects. And that's why we pre-prepared to ensure that we could deal with any issues along the way."

Now Australia has increased the number of vaccines it will buy through other developers, increasing its deal with AstraZeneca and Oxford University from 33 million doses to 53 million doses, and the deal with Novavax from 40 million doses to 51 million doses. Both of those vaccines were likely to need two doses for each person to be effective.

As part of the new deal with AstraZeneca, 20 million doses of the vaccine would be manufactured by CSL in Melbourne.

The University of Queensland vaccine candidate was using cutting-edge molecular clamp technology, which involved targeting the spikes on the outside of the coronavirus cell. The clamp comprised two fragments of a protein that was found in HIV.

Vaccine researchers from the University of Queensland and CSL said the decision was disappointing and in spite of HIV false positives the vaccine had looked promising.

"The robust neutralising immune response we saw directed against the virus exceeded expectations, looked very good, and in fact, 85 per cent of participants who had the two doses generated immune response above the mean of that of a recovered population," University of Queensland senior group leader Professor Trent Munro said.

"So, if we benchmark this against other studies out there, everything was looking very promising."

There had been 216 participants in the trial and there had been no adverse health reactions among those, CSL chief scientific officer Dr Andrew Nash said.

"There are a number of challenges in rolling out the vaccine and that these challenges might have an impact on public confidence in vaccination programs more broadly," he said.

"At this stage, and without more data, significant changes would need to be made to the well established HIV testing procedures and the healthcare setting would need to change to accommodate the rollout of this vaccine."

"This decision emphasises the rigorous assessment involved in vaccine development and the myriad considerations that must be taken into account in order to produce a vaccine that's safe and effective, logistically viable and publicly acceptable."

READ MORE:

Health Minister Greg Hunt said the issue with deciding not to go ahead was not about safety but about the risk to vaccine confidence due to the protein used in the vaccine causing a false positive in HIV testing.

"This is the scientific process working. It's the planning process working. It's an honest explanation of some of the challenges we've had," Mr Hunt said.

The university's co-lead on the vaccine, Paul Young, said re-engineering the vaccine was possible, but not in the time available.

Mr Morrison thanked Professor Young and the other researchers who had worked on the vaccine development.

Health secretary Brendan Murphy said it looked like the vaccine would have worked, it was producing antibodies, but the government couldn't risk the public's confidence in vaccines.

At a press conference held after the national cabinet meeting on Friday, Mr Morrison refuted the suggestion warnings about the risk of HIV false positives had been ignored and money wasted on backing the vaccine.

"Every cent we have invested in getting the best and most early available and save vaccines for Australians and the middle of the COVID-19 pandemic has been money well spent. Every single cent," he said.

"I mean, there are no guarantees when it comes to vaccine development and if you do not put investment in across a range of opportunities and options than you do not get one come out the other end and I think the expectation that they would be 100 per cent success rate across these naive."

Professor Murphy said no warnings were ignored.

"The possibility of false positives was raised by the University of Queensland very early on and was seen as a very, very unlikely possibility because the fragment of the HIV virus molecule was small," he said.

"It was unfortunately an unexpectedly high rate when the data came in. This was very clearly known at the time and the risks were appropriately taken and unfortunately it just became a bigger problem than anyone had anticipated."

The government has previously said it wants to start rolling out a vaccine to the Australian community early in the new year.

Three vaccine candidates have moved to provisional determination stage with the Therapeutic Goods Administration, but the University of Queensland vaccine had not reached that stage.

- with AAP