‘I wouldn’t bet against the virus’: Why omicron won’t be the last variant

The coronavirus’s evolution has been a wild ride. It’s not over yet.

The long omicron winter is over, but SARS-CoV-2 isn’t out of surprises.

Even as many countries lift pandemic restrictions, virus hunters remain on high alert — warning that it’s not a question of if another dangerous variant emerges, but when. The sheer number of infections worldwide, at least 450 million and counting, has given the virus ample opportunity to evolve. And SARS-CoV-2 has thrown curveballs before. The pace at which variants have emerged and how different they’ve been from one another has startled scientists.

“I wouldn’t bet against the virus,” said Thomas Friedrich, a virologist at the University of Wisconsin-Madison. “Omicron won’t be the last variant.”

Researchers trying to get ahead of the virus are tracking its evolution in real time through genomic sequencing. They’re using powerful computational tools to anticipate future changes. And they’re hunting for animal species that could be harboring SARS-CoV-2, giving the virus a place to lurk and evolve. That has made the virus one of the most studied in history.

“I think we’ve learned at this point to not underestimate what kind of sort of evolutionary leaps this virus can make,” said Stephen Goldstein, a virologist at the University of Utah. “We should be pretty humble about making predictions about what’s next.”

Eventually, as humanity builds up immunity to SARS-CoV-2 through infection and vaccination, virologists expect disease to become milder on average. But there’s no telling how long that will take, or how many variants the world will encounter along the way. Whether pi (the next letter in the Greek alphabet) proves a greater or lesser challenge than its predecessor is anyone’s guess.

The making of a variant

Evolution is born of error.

Viruses, like living things, introduce random mutations into their genetic codes when they copy themselves. RNA viruses, a broad group that includes SARS-CoV-2, have a naturally higher mutation rate than animals. But coronaviruses have a built-in proofreading mechanism that lowers that rate compared with other RNA viruses like HIV or influenza.

That hasn’t stopped new covid variants from emerging faster than expected.

“A single mutation doesn’t make a new variant,” said Siobain Duffy, a virologist at Rutgers University. A variant gets that status “when it starts to behave a little differently in human populations, and that can be a few changes or a lot of changes.”

Usually, mutations are benign or even actively harm a virus. But every so often, a change makes a virus better at infiltrating cells or evading the defenses of its host. Viruses with such mutations should outcompete their less-fit relatives over time. Whether that happens, and how quickly, depends on a range of factors.

The first is luck. Whether a mutation helps or hurts a virus is purely random. Each time SARS-CoV-2 copies itself is like a turn at a giant slot machine with 30,000 panels — equal to the number of letters in the coronavirus’ genetic code. The chance that any instance of copying will yield a beneficial error in one of those windows is infinitesimally small. But we’ve given the virus trillions of chances.

“If you let the virus spread uncontrollably, and it’s spreading in millions of people, then of course you’re giving an excellent chance to accumulate mutations,” said Theodora Hatziioannou, a virologist at Rockefeller University.

Major variants sequenced in U.S. versus India

Within a single infected person, viruses copy themselves billions of times. Add this up over hundreds of millions of infections, and the possible combinations of mutations are mind boggling. Vaccination cuts into this opportunity by preventing infection outright or shortening its duration, but it doesn’t eliminate it.

Still, just the emergence of a new mutation by itself isn’t enough to create a new variant.

“Even if a beneficial mutation is generated in one individual, the chances that it actually gets out of that individual [and into another] is pretty small,” for an infection that lasts a couple weeks, said Friedrich. But when enough people get infected, “something highly improbable may happen.” Indeed, many experts think the delta variant likely evolved this way, as mutations spawned from rampant spread gradually cohered into a distinct lineage.

Distinctly different variants can also emerge from protracted infections in people with a weakened immune system, where covid can simmer for months. That’s because these people often mount an immune response that’s not strong enough to fully clear the virus.

Sparring with an immunocompromised adversary can “teach” the virus to evade the immune system, by exposing the virus to weakened versions of our infection-fighting moves. For instance, antibodies that might normally overwhelm the virus can only mount a faint attack in an immunocompromised person, which can favor antibody-dodging mutations while also giving them the chance to proliferate. In such individuals, viruses can accumulate many mutations that persist for long enough to become dominant in the individual. When the virus then jumps to other hosts, it can look markedly different from related strains. It’s akin to how stopping a course of antibiotics early can breed antibiotic resistant bacteria.

Researchers have seen this evolution play out in immunocompromised patients hospitalized with covid. Within one patient, the virus population accumulated more than 30 mutations over about 150 days, including seven that have been linked to immune evasion. In another, the virus remained relatively stable for about two months until a course of convalescent plasma therapy, which floods the patient with antibodies of someone who recovered from infection. After that, viruses with distinct mutations that hide the virus from antibodies quickly became dominant. Both patients eventually died from their infections.

Some of the mutations detected in immunocompromised people are shared by major variants, such as alpha, suggesting that evolution within immunocompromised people can spark big jumps in coronavirus evolution.

New drugs might help prevent immunocompromised folks from getting infected in the first place, protecting them while also reducing the chance for variant evolution. Evusheld, which combines two antibody treatments, cuts risk of infection by 77 percent and is cleared for use in immunocompromised people. But supply is limited, and only about 850,000 courses have been distributed so far. Without widely available treatments, measures like masking and social distancing remain the only protection for immunocompromised people.

A deer is seen at Grouse Mountain in Vancouver, Canada, on June 12, 2020.

New variants can also emerge outside of humans. So far, SARS-CoV-2 has been found in over two dozen animal species, including minks, house cats, dogs, hamsters and tree shrews. Some of these species are dead-ends for the virus, one-off infections that don’t spark outbreaks. But others may become reservoirs for the virus that foster divergent, potentially dangerous, evolution. In this new environment, strains can go down unexpected evolutionary paths that may thwart our defenses, were the virus to somehow jump back into humans. Such reservoirs make eradication of any virus in humans virtually impossible.

Scientists have been trying to figure out what species could serve as reservoirs by actively looking for the virus in wild animals and modeling which species’ have respiratory systems susceptible to SARS-CoV-2 infection and are likely to be in close contact with humans. White-tailed deer have emerged as an animal of special concern, since over 40 percent of some populations have been infected. In late February, Canadian researchers identified a highly divergent strain in deer that may have spread to one person in southwestern Ontario.

Virologists are still sorting out the relative likelihood of these different variant-spawning scenarios. But some combination of them may have been responsible for the variants we’ve seen so far, which have kind of been all over the place, genetically speaking. Each new strain has sprouted from an unexpected spot on the family tree.

Scattershot evolution

In general, the sorts of viruses humans get every year tend to evolve in a ladderlike pattern. “It’s actually more of a stair-step,” said Emma Hodcroft, a virologist at the University of Bern and co-developer of Nextstrain, a site that tracks pathogen spread. Next year’s influenza strain tends to be a direct descendant from last year’s, she said, creating a stair-like family tree.

“So far, that’s not been the case for SARS-CoV-2,” she said. “None of the variants of concern so far have come from a previous variant of concern. They’ve all come from very different parts of the tree.”

That scattershot evolution is probably not due to SARS-CoV-2 being an especially weird virus, said Goldstein, but simply that we’ve never before been able to track a new virus blaze across the globe with such granularity. HIV burst into the public consciousness in the late 1980s, but scientists didn’t have the sequencing capacity to follow its evolution in real time as it spread around the world until relatively recently.

“We expect when the virus enters a new host that there can kind of be a burst of evolutionary change as the virus adapts to that new host,” he said. “We’re seeing that process with SARS-CoV-2, which we haven’t seen with other coronaviruses, which entered the human population before modern virology, let alone sequencing technology.”

Consequently, the virus has thrown us some curveballs, with omicron being the curviest.

“Omicron kind of came out of nowhere,” Duffy said. “We hadn’t expected something to be so different without seeing intermediate things that had some, but not all of the mutations that omicron had.”

The variant boasts about 50 mutations, about twice as many distinctive changes as alpha or delta. Many of omicron’s mutations were new or exceptionally rare across the millions of coronavirus samples scientists had already sequenced. That suggests omicron may have emerged from an immunocompromised person’s protracted battle with the virus, though definitive data is lacking.

The variant is exceptionally transmissible. It also gets around the substantial immunity built up around the world better than any previous version. Omicron causes milder illness than prior variants because it doesn’t quite get as deep into the lungs and because it emerged when many people had some immunity from prior infection or vaccination. Yet it still wreaked havoc across the globe — and the even more transmissible omicron offshoot BA.2 is now gaining ground.

Countries that recently experienced an omicron surge, and especially those with high vaccination rates, have fairly robust defenses against BA.2, suggesting it won’t drive enormous surges around the globe. Cases and hospitalizations are slightly rising in some places where BA.2 has become dominant, including the U.K., though a mix of relaxed mask and distancing mandates and the variant could explain the upturn.

Hatziioannou is willing to make one prediction. “The only thing I can say with certainty is that omicron will not be the last variant.”

An uncertain endgame

When omicron’s successor emerges, or what it looks like, is anyone’s guess.

Covid transmission should wane in the coming months, especially as temperatures warm across the northern hemisphere. That combo of lower transmission and built-up immunity from winter’s omicron surge could buy time, though the abrupt end of protections like mask mandates across the U.S. might keep cases higher. But there’s plenty of opportunity for the virus elsewhere, as roughly 35 percent of the global population remains unvaccinated, largely because of access issues.

“We could have a very good summer,” Hodcroft said. “Or we could have some crazy variant pop up in June. There’s no rules against it.”

There’s no guarantee future variants won’t be deadlier than omicron or the variants that preceded it. The idea that viruses always evolve to become more benign over time “is just false,” said Hodcroft. The myth stems from the idea that it’s bad for a virus to kill its host, since the host won’t be up and about spreading the virus, so over time it pays the virus to become milder. That falsely assumes that transmission and death are linked, Hodcroft said, which isn’t the case with covid, since most transmission happens before an infected person feels sick.

Other viruses also deflate this idea. “HIV is just as dangerous to get as it was in the 1980s, aside from treatment,” she said, and a more virulent strain was recently discovered in Europe. Measles has remained similarly dangerous.

In the worst-case scenario, the next covid variant could harbor a suite of mutations that render it more lethal and resistant to existing vaccines. It’s unlikely that the virus will completely evade our defenses in the near term, but a strain even slightly resistant to vaccines could drive another wave of cases, hospitalizations and deaths. It’s hard to know the odds of such a variant emerging, Hodcroft said: “We really don’t have a lot to draw on when it comes to estimating how many mutations or tricks up its sleeve it might have left.”

Given time — perhaps months, perhaps years — humanity and SARS-CoV-2 will likely reach a dynamic standoff known as endemicity. As immunity builds up, through vaccination or rounds of infection, the average severity of disease will fall, Goldstein said. An individual’s protection isn’t permanent and will wane over time, but once immunity becomes broadly distributed, we’ll reach an equilibrium akin to our relationship with influenza or the coronaviruses that cause common colds. The only truly immunologically naive people at that point will be newborns.

In such a world, the coronavirus and humans will be more evenly matched. “Variants will be less severe than they otherwise would have been in a naive population,” Goldstein said. The virus will continue changing, but surges will be less like what we’ve seen the past two years and more like winter surges of colds or the flu.

We are still awhile from reaching that equilibrium, and the road to it could still be rocky, especially as many countries drop protective measures. In the U.S. alone, more than 1,300 people are dying of covid each day, a toll roughly equal to all influenza deaths this year.

“We’ve done a terrible job at the equitable distribution of vaccines globally,” Friedrich said. “A lot of people remain susceptible and will continue to be infected, and each infection is a chance for the virus to pull the evolutionary slot machine handle.” The coronavirus may still hit the jackpot.

Jonathan Lambert

‘I wouldn’t bet against the virus’: Why omicron won’t be the last variant