BA.4 and BA.5 are responsible for a new wave of Covid-19 cases in South Africa. At least one of them has arrived on our shores. So what does the science tell us about these new omicron variants?

While most countries are winding down their testing and sequencing efforts, South Africa has been doing an absolutely stellar job of detecting new Covid-19 virus variants. It was the country that first identified the three original omicron lineages, BA.1, BA.2, and BA.3, back in November last year. Once other countries started to look, they found omicron everywhere.

BA.1 started the initial global omicron wave, followed by the more infectious BA.2. In South Africa they only had one large BA.1 wave. Here in New Zealand, BA.1 and BA.2 arrived and seeded into the community very close together, so we had both at the same time – though BA.2 became the dominant lineage. BA.3 never really took off anywhere. 

Now South Africa has identified two new omicron lineages, BA.4 and BA.5. The data suggests BA.4 originated in mid-December and BA.5 in early January. At the moment, they are responsible for a new rise in cases in South Africa, the country’s a fifth wave of Covid-19. South Africa’s positivity rate – the number of tests that are coming back positive – has jumped from 4% to 20% in the last few weeks. In that time, daily cases have risen from the hundreds to the thousands. 

It’s not clear at the moment whether what’s driving the rise in cases is 1) BA.4 and BA.5 being more infectious, 2) BA.4 and BA.5 having mutations that help them evade immunity even more, or 3) whether everyone’s immunity from the last wave is now waning, making them susceptible to infection again. Given the timing since the last wave, it seems almost certain waning immunity is at play. It’ll be a little while before we know more about the other two options. Because it takes a few weeks for cases to progress to hospitalisations, it’ll also be a while before we know whether these new lineages are more dangerous than the virus already is, at least for people without access to the new antivirals.  

So what do we know about BA.4 and BA.5 at the moment?

The picture below shows a family tree for the Covid-19 virus, going right back to the early version that kicked off the pandemic – referred to as 19A. The omicron lineages are shown orangey-red. BA.1 is referred to as 21K, BA.2 as 21L, BA.4 as 22A and BA.5 as 22B. 22C is another lineage of omicron called BA.2.12.1 which has become dominant in New York State. As you can see from the picture, 22A and 22B both come from 21L, meaning BA.4 and BA.5 have evolved from BA.2.

What made omicron so unusual was the sheer number of mutations it has compared to previous variants of concern like delta, and the ability many of those mutations give the virus to infect people who have already had Covid and/or been vaccinated. They also changed what we call the tissue tropism of the virus. Delta seemed to prefer to infect cells deeper in the lungs, where as omicron is more up in the throat and nose. You can find out more about what mutations BA.1 has here and BA.2 has here. 

BA.4 and BA.5 are different from BA.2 by about eight to 10 mutations. Some of these we’ve seen before in other variants of concern, like epsilon, and others look like they’ll help the virus bind to its receptor on human cells a little better. One of the most concerning changes in both BA.4 and BA.5 is the L452R mutation. This is a mutation that delta has but BA.1 and BA.2 omicron don’t have. In March, Yahan Zhang and colleagues published the results of their experiments making a version of omicron containing the L452R mutation. Worryingly, they found that L452R enhanced the ability of omicron to infect the lung tissues of mice engineered to have the human version of ACE2 – the receptor the virus uses to infect our cells.

Remember how before I said omicron doesn’t normally infect lung cells so well? Well, the fact BA.4 and BA.5 now have this mutation suggests they’ll be able to infect those deeper lung tissues. Does that mean these new lineages will be able to cause more serious disease than BA.2? Possibly. What we don’t know is how the L452R mutation will behave in combination with the other mutations BA.4 and BA.5 have. Will they cancel each other out, or will their effects be additive?   

How worried should we be about these new variants?

Frankly, I’m worried about all versions of omicron around. All have the capacity to cause serious illness and death, in at least some people. And all have the capacity to cause lingering and debilitating symptoms, aka long Covid. Yes, even the BA.1 and BA.2 lineages of omicron. And that’s before we even factor in what the long-term impact of the damage the virus can do to our various organs even after a mild infection.

But I’m also worried about the variants of the virus we don’t yet know about but are almost certainly out there evolving away. Remember that omicron came out of nowhere. At the time BA.1, BA.2 and BA.3 emerged, the world was reeling from the delta wave and wondering how it might evolve in the future. Then omicron appeared and caused the biggest wave of the pandemic so far. Look at the virus family tree again. Omicron is from a completely different branch to delta. Where did it come from, and what’s the chance of another completely different variant emerging? Pretty high, it turns out.

One big problem with the “getting back to normal” phase that most countries have adopted is that it is putting immunocompromised people at high risk of infection. While this is obviously really dangerous for their health, it also has the potential to be really bad for everyone else’s too. Most people who catch Covid-19 will clear the virus in a few days or weeks. But some immunocompromised people are at increased risk of the virus replicating in their cells for many months. This is called a chronic or persistent infection and is different to long Covid where symptoms persist after the virus is cleared. 

Researchers recently reported the case of a 48-year-old woman who had the virus for 335 days. She’s diabetic and in complete remission after having previously had a form of blood cancer. She must have been one of the United States’ early cases as she was first admitted to hospital in April 2020. She was discharged a month later despite still having a fever, cough, and needing extra oxygen. She was PCR tested every three months, sometimes testing positive and sometimes negative. Her symptoms came and went, which sounds an awful lot like long Covid. Then in March 2021, her condition worsened and she was readmitted to hospital. She tested positive, again. 

But here’s the thing. Genome sequencing of the virus from swabs taken during her first hospitalisation in May 2020 and then her second in March 2021 showed the virus was the same variant. But that variant wasn’t the one circulating any more. In other words, she hadn’t been reinfected but had remained infected the whole time. And during that time, the virus infecting her had evolved, developing numerous mutations, including ones affecting the effectiveness of antibody therapies. 

Treatment of Covid-19 has come a long way since the woman was first admitted to hospital, so this time she received antivirals as well as serum from people who’d recovered from infection, which contained their protective antibodies. This time the treatment worked and three months later she was testing negative. We don’t know how rare this is, but another group of researchers, this time from the UK, just put out a press release saying they’d identified nine immunocompromised patients who were chronically infected, the longest for 505 days.

What’s clear is that the next variants of concern are currently evolving all around the world. Because they are evolving in both healthy and immunocompromised people, we’ve no idea what the next variant of concern might actually be. It may be one of these new omicron lineages, or it may be a form of delta, or it may be a version of the virus that was circulating a year or more ago.

We need to keep public health protections in place and strengthen our Covid surveillance

Researchers from the US have recently shown that large super-spreader events also contribute to the emergence of new variants. They found that most new highly infectious variants will fizzle out if they only infect a few people. But if these variants meet a super-spreading event, they are much more likely to gain a foothold in the community. That means as well as doing our best to reduce community transmission and stop the virus from infecting immunocompromised people, we need to reduce the chances a large gathering will turn into a super-spreading event. That should help delay or even prevent the emergence of at least some new variants of concern. 

What worries me is that we likely won’t even know the next new variant has emerged unless or until there are cases in a country like South Africa which is still testing widely and routinely doing genome sequencing. The reality is that PCR testing and genome sequencing are expensive, so most countries are dismantling their testing systems as part of “getting back to normal”, or making people pay for previously free tests, further reducing uptake. The Guardian reports the head of global diagnostics alliance FIND, William Rodriguez, as saying that despite the world experiencing the omicron wave over the last four months, “testing rates have plummeted by 70% to 90% worldwide”. The widespread use of rapid antigen tests means fewer samples suitable for sequencing too. 

As the director-general of the World Health Organisation, Dr Tedros Adhanom Ghebreyesus, puts it: “When it comes to a deadly virus, ignorance is not bliss.”