We Still Don’t Know How Well Covid Vaccines Stop Transmission

THIS WEEK, THE US passed a grim milestone in the ongoing coronavirus crisis: 500,000 deaths, more than the number of Americans killed in World War II, the Korean War, and the Vietnam War combined. And yet there is a growing sense of hope that the worst might now be behind us. With new cases declining and immunizations accelerating—45.2 million people have so far received at least one dose of a Covid-19 vaccine, including 20.6 million who have been fully vaccinated—many Americans are beginning to allow themselves to imagine what post-pandemic life might be like.

Achieving that is likely to take a few more months—provided vaccine makers don’t hit any production snags and worrisome variants don’t derail current progress. In the interim, an increasing number of people will find themselves in a liminal state, navigating what it means to be a vaccinated person moving through an unvaccinated world. What are its rules, and what will it take to be a good citizen of it? Answering those questions means confronting an even more fundamental unknown. A vaccinated person may be well-protected from the worst ravages of Covid-19. But it’s not clear if they can still carry the coronavirus and transmit it to susceptible people around them.

This week, two new studies—neither of which have yet gone through peer review—made splashy headlines about the extent to which vaccines slash viral spread. The first, a leaked manuscript first reported by Israeli news site Ynet before being covered by MIT Technology Review, BloombergThe Financial Times, and Vox, found that two doses of Pfizer-BioNTech’s shot drove an 89.4 percent drop in infections—both symptomatic and asymptomatic—among vaccinated people in Israel. Though they did not directly measure transmission, the study’s authors—researchers from the Israel Ministry of Health, Hebrew University, and Pfizer—stated in the abstract that the Pfizer vaccine “was highly effective in preventing SARS-CoV-2 infections.” Subsequent news coverage hailed it as the first evidence from the real world that the vaccine could strongly suppress spread of the virus. But scientists not associated with the study say that was an overstatement. (Indeed, Bloomberg later updated its story to include such criticisms, though not the headline.)

In the report, which WIRED has obtained, the research team analyzed aggregated data from Israel’s national testing and disease surveillance system, comparing infection rates in vaccinated versus unvaccinated groups of people between January 17 and February 6. However, as the study authors noted, the ministry’s testing recommendations exempt vaccinated people from requirements like getting tested after travel or being exposed to a known Covid case. Under these protocols, unvaccinated people aren’t just required to get tested more often, they might also choose to, because they’re more worried about contracting Covid-19 than people who’ve gotten the jab. And because infections—especially asymptomatic ones—are more likely to be detected in the group that’s testing more frequently, the estimated 89.4 percent transmission-blocking effect of the vaccine is almost certainly too optimistic (a caveat the authors acknowledged, saying more research is needed to confirm their findings).

“The testing rates were such a hodgepodge, I don’t know you can make any conclusions about how much the vaccine cut transmission in Israel, let alone assigning a number as concrete as 89.4 percent,” says Eric Topol, a professor of molecular medicine at the Scripps Research Institute. The only way to do a careful study of asymptomatic spread and how well the shots curtail it, he says, is to swab both groups—vaccinated and unvaccinated people—every day, ideally for months. Though perhaps prohibitively expensive, the most rigorous version of that experiment would be to also follow each positive test up with contact tracing and genomic sequencing to confirm the route of transmission. The way the Israeli team did it is not the way to get the answer, he says: “It’s a way to get a lot of interest, because everybody wants to hear this.”

Indeed, right on cue, a few hours after the Israeli study began making the rounds in American media, one of my family members sent it to me. Just a few days prior, this person—who is vaccinated with the Pfizer shot—had asked me how safe it would be to fly and immediately join a pod of unvaccinated relatives. They did not like my answer (probably still risky, but we really don’t know just how risky just yet). They liked the Israeli study’s answer much better.

Scott Halpern, an epidemiologist and critical care physician at the University of Pennsylvania, says this is a pretty classic case of optimism bias—the general proclivity of the human species to believe that our desired outcome is likely to be the correct one. (Halpern has written about how other cognitive biases have hampered a smart and effective public health response here in the US.) It’s the same neural tick that drives people who’ve been told their loved one hooked up to a ventilator in the Covid-19 ward has a 5 percent chance of surviving. Most people believe their loved one will be in that 5 percent. Halpern knows this one from first-hand experience; he’s still seeing Covid-19 patients in the ICU most weeks.

Mix that optimism bias with pandemic fatigue, and you have a recipe for some under-baked science getting slung around as rationale for people doing the things they really want to do. “Once you’ve been bombarded with bad news long enough, any glimmer of good news is something we just emotionally latch onto,” says Halpern. “That’s just human nature.”

The Israeli team did not respond to WIRED’s emailed questions. A Pfizer spokesperson declined to comment on the 22-page report, which was first described last week by Israeli journalist Nadav Eyal, who published screenshots of the text on Twitter.

The second report, a preprint posted on The Lancet Monday, blew that glimmer of good news into a bigger flame. It described a Public Health England study of health care workers in the United Kingdom who’d received the Pfizer-BioNTech vaccine, and who were tested every 14 days for Covid-19. The study found that in addition to making people less likely to get sick from the coronavirus—no surprise there—the vaccine sliced the risk of the recipient getting infected, period. By how much? Vaccinated health care workers were 72 percent less likely at 21 days after the first dose, and 86 percent 7 days after the second dose. The logical jump here is that a vaccinated person has far fewer chances to spread the virus, since the shot reduces the odds they’ll ever be carrying it around. “We provide strong evidence that vaccinating working age adults will substantially reduce asymptomatic and symptomatic SARS-CoV-2 infection and therefore reduce transmission of infection in the population,” the study authors concluded.

While this study was better controlled, Topol says testing every two weeks still isn’t frequent enough to catch new infections. “It’s really got to be daily,” he says. Those kinds of experiments are much harder and costlier to do. But both Pfizer and Moderna are reportedly working on them right now, with data rumored to drop sometime in the next few weeks. (A Pfizer spokesperson declined to confirm that timeline or provide any details until data from a study has been published. Moderna did not respond to an emailed request for more information.)

WIRED’s questions did not immediately receive a response from the research team or Public Health England.

For now, that leaves the public without a lot of firm answers. “We’re going to be sitting with incomplete knowledge on this topic for some time still,” says Topol—even though he and other scientists say they’re confident vaccination against Covid-19 will eventually be shown to reduce the chance of transmitting the virus. The question is: By how much?

“I think it is highly likely that there is some transmission-blocking effect to these vaccines,” says Kawsar Talaat, an infectious disease physician and vaccine safety researcher at the Johns Hopkins Bloomberg School of Public Health. She also led the Johns Hopkins Center for Immunization Research’s clinical trial of the Pfizer vaccine. But, she says, putting a number on the size of that effect is really tricky.

For one thing, scientists have yet to determine how many copies of the virus need to be in an infected person’s nose to make them contagious to others. “There’s not a clear threshold yet for the amount of viral load at which transmission happens,” says Talaat. Which means you can’t look at someone’s PCR result and know with certainty that the amount of virus the test is detecting is enough to make other people sick.

In general, researchers have found that less virus means milder symptoms and a lower risk of passing it on. But not always. People’s bodies and immune systems behave differently. We all breathe at different rates, exhaling particle plumes of varying size and density. Researchers also haven’t pinpointed exactly how much SARS-CoV-2 has to get into your nose to cause an infection. And that number is also subject to similar biological variables, as well as environmental ones, like which strains are circulating in your area. More transmissible variants, like B.1.1.7, which is currently spreading across much of the US, might need to take fewer cracks at the ACE2 receptor. So being able to say that X amount of virus in the nose equals Y risk of spreading to other people is just beyond the scope of the current science.

Even so, several research groups in Israel are measuring viral load in vaccinated people who later test positive for SARS-CoV-2. One team from the Israel Institute of Technology and Tel-Aviv University recently observed that people who caught the virus two to four weeks after receiving their first dose of the Pfizer vaccine had up to 4 times smaller viral loads than people who got infected in the first two weeks after getting the shot. The results suggest the vaccine reduces the risks of transmission, while not erasing them entirely.

That’s an expected result, says Talaat. No vaccine out there right now, no matter how good (and the Pfizer and Moderna shots are very good), will stop the coronavirus from causing infections in at least some people who’ve gotten the shots. Here’s why.

The two FDA-authorized vaccines in the US are first-in-their-generation genetic vaccines. What comes out the tip of that needle and into the muscles of your arm are microscopic, fat-encapsulated strands of mRNA. These strings of genetic letters carry the instructions for making bits of the coronavirus spike protein, which the live virus uses to infect human cells. Capillaries that run through the muscle whisk the mRNA molecules into the bloodstream and carry them to the nearest lymph node. Once there, the vaccine components encounter dendritic cells and macrophages—two types of immune cells that can sense when something foreign is sneaking around in the body. They grab the mRNA and use it to produce pieces of the spike protein, which they then display on their surfaces to flag down other immune cells. These begin churning out antibodies and activating T-cells to fight off what the body perceives as an infection.

There isn’t one, of course. But this fire drill prepares the immune system to quickly kick into action should the vaccinated person encounter the real coronavirus in the future. The Pfizer-BioNTech and Moderna vaccines are especially good at it. But to an antibody or a T-cell, the body is a big place, and the nose is a battlefront far-removed from the initial action in the arm. “If you give a shot in the arm, you’re likely to get immunity in the body,” says Talaat. “But it’s hard to create immunity in the mucosal surfaces where the virus colonizes.”

SARS-CoV-2 might cause its deadliest damage in the lungs, heart, and blood vessels. But its first stop in the human body is usually the nose, because that’s where inhaled viral particles first encounter cells they can invade and hijack in order to make copies of themselves. From there, the swarm of new viruses can expand into other organs—if the immune system doesn’t shut them down. And it’s from the nose that infected people can send out new clouds of contagion.

So in order for a vaccine to block transmission completely, it would have to recruit a cast of SARS-CoV-2-targeted antibodies and immune cells specifically to patrol the nasal passages, where they could glom onto any coronaviruses just after they get inhaled, and before they start their self-replication spree. This is how the nasal spray versions of flu vaccines work. But that’s just not what Pfizer or Moderna’s shots were designed to do. They were designed to create a sparser crew of wider-roaming immune defenders that can jump-start a bigger response wherever they encounter the virus, giving an infected person a better chance of beating back full-blown symptoms. “The goal of these vaccines has always been to prevent people getting hospitalized and dying, because that has the biggest public health impact,” says Talaat.

The good news about the Israeli and UK studies, even with their methodological flaws regarding transmission, Talaat says, is that they show that out in the real world, away from the controlled parameters of a clinical trial, the vaccines are working fabulously at preventing people from getting seriously ill. In the leaked Israeli report, the vaccines led to a 95 percent drop in hospitalizations and 92 percent dip in deaths. And newer, better-vetted data is already starting to back that up.

A study published Wednesday in the New England Journal of Medicine that analyzed 600,000 pairs of vaccinated and unvaccinated Israeli individuals found that two doses of the Pfizer’s shot was 92 percent protective against severe disease and 87 percent effective at preventing hospitalization. Though the study did not have data on deaths following the second dose, just a single shot lowered death rates by 72 percent. So from a public health standpoint, that makes the question of whether or not the Pfizer vaccine, or any other, stops viral spread really a secondary concern, says Talaat. “If you vaccinate enough people, then you don’t need a vaccine that stops carriage in the nose and potential transmission,” she says.

But that number does matter for answering questions like these: Is it safe to eat inside a restaurant? Or get on an airplane? Or hug your grandkids?

Say the vaccine you get is 80 percent effective at blocking viral spread. That means that should you contract the virus, you may not get seriously ill or even have a single symptom—but there’s still a 20 percent chance you’ll pass it on to someone else. And what if the vaccine you get is only 50 percent effective at blocking spread? Now it’s a coin flip.

“This is exactly the type of gray area where reasonable people might reasonably arrive at different answers,” says Halpern. “It all comes down to the fact that we don’t all have the same risk tolerance.”

Talaat’s version of this calculus involves navigating family get-togethers with her parents (vaccinated) and her siblings (unvaccinated). Since she herself is vaccinated, Talaat still wears a mask when she’s visiting her siblings. And she’ll continue to do so until they get their shots. But she feels more relaxed around her parents. “If you are in a household or pod with someone who’s not vaccinated, you should still be as careful as you can to prevent potential transmission, especially if that person is high-risk,” she says.

That doesn’t mean your post-shot life has to look exactly the same as pre-jab. But it does mean continuing to wear masks and socially distancing. Both of those further reduce the risk of spread—how much exactly, no one can say, so public health experts say better to do them all. At least for now.

That’s because whatever the “real” transmission rate for vaccinated people may turn out to be, that number doesn’t exist in a vacuum. A 50 percent reduction in risk of transmission isn’t (or shouldn’t be) particularly liberating if the virus is still running rampant in your area. But it might be if local case prevalence shrinks to the point where your odds of exposure to the coronavirus are virtually zero. The calculation can’t just be about the level of protection of the vaccine against viral spread, but also about the risks in your particular community, says Halpern.

“So where we are today is that you have to keep your mask on after getting vaccinated—not because the vaccine doesn’t work, but because there’s still too much virus in pretty much every neighborhood in America,” he says. “If everyone keeps doing it, we’re going to get to a point where the masks can come off. But we’re not there yet.”

Scientists Are Trying to Spot New Viruses Before They Cause Pandemics

Back in the summer, Dr. Michael Mina made a deal with a cold storage company. With many of its restaurant clients closed down, the firm had freezers to spare. And Dr. Mina, an epidemiologist at the Harvard T.H. Chan School of Public Health, had a half-million vials of plasma from human blood coming to his lab from across the country, samples dating back to the carefree days of January 2020.

The vials, now in three hulking freezers outside Dr. Mina’s lab, are at the center of a pilot project for what he and his collaborators call the Global Immunological Observatory. They envision an immense surveillance system that can check blood from all over the world for the presence of antibodies to hundreds of viruses at once. That way, when the next pandemic washes over us, scientists will have detailed, real-time information on how many people have been infected by the virus and how their bodies responded.

It might even offer some early notice, like a tornado warning. Although this monitoring system will not be able to detect new viruses or variants directly, it could show when large numbers of people start acquiring immunity to a particular kind of virus.

The human immune system keeps a record of pathogens it has met before, in the form of antibodies that fight against them and then stick around for life. By testing for these antibodies, scientists can get a snapshot of which flu viruses you have had, what that rhinovirus was that breezed through you last fall, even whether you had a respiratory syncytial virus as a child. Even if an infection never made you sick, it would still be picked up by this diagnostic method, called serological testing.

“We’re all like little recorders,” keeping track of viruses without realizing it, Dr. Mina said.

This type of readout from the immune system is different from a test that looks for an active viral infection. The immune system starts to produce antibodies one to two weeks after an infection begins, so serology is retrospective, looking back at what you have caught. Also, closely related viruses may produce similar responses, provoking antibodies that bind to the same kinds of viral proteins. That means carefully designed assays are needed to distinguish between different coronaviruses, for example.

But serology uncovers things that virus testing does not, said Derek Cummings, an epidemiologist at the University of Florida. With a large database of samples and clinical details, scientists can begin to see patterns emerge in how the immune system responds in someone with no symptoms compared to someone struggling to clear the virus. Serology can also reveal before an outbreak starts whether a population has robust immunity to a given virus, or if it is dangerously low.

“You want to understand what has happened in a population, and how prepared that population is for future attacks of a particular pathogen,” Dr. Cummings said.

The approach could also detect events in the viral ecosystem that otherwise go unnoticed, Dr. Cummings said. For example, the 2015 Zika outbreak was detected by doctors in Brazil who noticed a cluster of babies with abnormally small heads, born seven to nine months after their mothers were infected. “A serological observatory could conceivably have picked this up before then,” he said.

Serological surveys are often small and difficult to set up, since they require drawing blood from volunteers. But for several years Dr. Mina and his colleagues have been discussing the idea of a large and automated surveillance system using leftover samples from routine lab tests.

“Had we had it set up in 2019, then when this virus hit the U.S., we would have had ready access to data that would have allowed us to see it circulating in New York City, for example, without doing anything different,” Dr. Mina said.The Coronavirus Outbreak ›

Although the observatory would not have been able to identify the new coronavirus, it would have revealed an unusually high number of infections from the coronavirus family, which includes those that cause common colds. It might also have shown that the new coronavirus was interacting with patients’ immune systems in unexpected ways, resulting in telltale markers in the blood. That would have been a signal to start genetic sequencing of patient samples, to identify the culprit, and might have provided grounds to shut down the city earlier, Dr. Mina said. (Similarly, serology would not be able to spot the emergency of a new virus variant, like the contagious coronavirus variants that were discovered in South Africa and England before spreading elsewhere. For that, researchers must rely on standard genomic sequencing of virus test samples.)

The observatory would require agreements with hospitals, blood banks and other sources of blood, as well as a system for acquiring consent from patients and donors. It also faces the problem of financing, noted Alex Greninger, a virologist at the University of Washington. Health insurance companies would be unlikely to foot the bill, since serology tests are usually not used by doctors to treat people.The Coronavirus Outbreak ›

Dr. Mina estimated that the observatory would cost about $100 million to get off the ground. He pointed out that, according to his calculations, the federal government has allocated more than twice that much to diagnostics company Ellume to produce enough rapid Covid tests to cover the American demand for only a handful of days. A pathogen observatory, he said, is like a weather forecasting system that draws on vast numbers of buoys and sensors around the globe, passively reporting on events where and when they arise. These systems have been funded by government grants and are widely valued.

The predictive power of serology is worth the investment, said Jessica Metcalf, an epidemiologist at Princeton and one of the observatory team members. A few years ago, she and her collaborators found in a smaller survey that immunity to measles was ominously low in Madagascar. Indeed, in 2018 an outbreak took hold, killing more than 10,000 children.

Now, the half-million plasma samples in Dr. Mina’s freezers, collected by the plasma donation company Octopharma from sites across the country last year, are starting to undergo serological tests focused on the new coronavirus, funded by a $2 million grant from Open Philanthropy. Testing had to wait for the researchers to set up a new robotic testing facility and process the samples, but now they are working through their first batches.

The team hopes to use this data to show how the virus flowed into the United States, week by week, and how immunity to Covid has grown and changed. They also hope it will spark interest in using serology to illuminate the movement of many more viruses.

“The big idea is to show the world that you don’t have to spend huge dollars to do this kind of work,” Dr. Mina said. “We should have this happening all the time.”

Tiny Blobs of Brain Cells Could Reveal How Your Mind Differs from a Neanderthal’s

In recent years, scientists have figured out how to grow blobs of hundreds of thousands of live human neurons that look — and act — something like a brain.

These so-called brain organoids have been used to study how brains develop into layers, how they begin to spontaneously make electrical waves and even how that development might change in zero gravity. Now researchers are using these pea-size clusters to explore our evolutionary past.

In a study published on Thursday, a team of scientists describe how a gene likely carried by Neanderthals and our other ancient cousins triggered striking changes in the anatomy and function of brain organoids.

As dramatic as the changes are, the scientists say it’s too soon to know what these changes mean for the evolution of the modern human brain. “It’s more of a proof of concept,” said Katerina Semendeferi, a co-author of the new study and an evolutionary anthropologist at the University of California San Diego.

To build on the findings, she and her co-author, Alysson Muotri, have established the UC San Diego Archealization Center, a group of researchers focused on studying organoids and making new ones with other ancient genes. “Now we have a beginning, and we can start exploring,” Dr. Semendeferi said.

Dr. Muotri began working with brain organoids more than a decade ago. To understand how Zika produces birth defects, for example, he and his colleagues infected brain organoids with the virus, which prevented the organoids from developing their cortex-like layers.

In other studies, the researchers studied how genetic mutations help give rise to disorders like autism. They transformed skin samples from volunteers with developmental disorders and transformed the tissue into stem cells. They then grew those stem cells into brain organoids. Organoids from people with Rett Syndrome, a genetic disorder that results in intellectual disability and repetitive hand movements, grew few connections between neurons.

Dr. Semendeferi has been using organoids to better understand the evolution of human brains. In previous work, she and her colleagues have found that in apes, neurons developing in the cerebral cortex stay close to each other, whereas in humans, cells can crawl away across long distances. “It’s a completely different organization,” she said.

But these comparisons stretch across a vast gulf in evolutionary time. Our ancestors split off from chimpanzees roughly seven million years ago. For millions of years after that, our ancestors were bipedal apes, gradually attaining larger heights and brains, and evolving into Neanderthals, Denisovans and other hominins.

It’s been difficult to track the evolutionary changes of the brain along the way. Our own lineage split from that of Neanderthals and Denisovans about 600,000 years ago. After that split, fossils show, our brains eventually grew more rounded. But what that means for the 80 billion neurons inside has been hard to know.

Dr. Muotri and Dr. Semendeferi teamed up with evolutionary biologists who study fossilized DNA. Those researchers have been able to reconstruct the entire genome of Neanderthals by piecing together genetic fragments from their bones. Other fossils have yielded genomes of the Denisovans, who split off from Neanderthals 400,000 years ago and lived for thousands of generations in Asia.

The evolutionary biologists identified 61 genes that may have played a crucial role in the evolution of modern humans. Each of those genes has a mutation that’s unique to our species, arising some time in the last 600,000 years, and likely had a major impact on the proteins encoded by these genes.

Dr. Muotri and his colleagues wondered what would happen to a brain organoid if they took out one of those mutations, changing a gene back to the way it was in our distant ancestors’ genomes. The difference between an ancestral organoid and an ordinary one might offer clues to how the mutation influenced our evolution.

It took years for the scientists to get the experiment off the ground, however. They struggled to find a way to precisely alter genes in stem cells before coaxing them to turn into organoids.

Once they had figured out a successful method, they had to choose a gene. The scientists worried that they might pick a gene for their first experiment that would do nothing to the organoid. They mulled how to increase their odds of success.

“Our analysis made us say, ‘Let’s get a gene that changes a lot of other genes,’” said Dr. Muotri.

One gene on the list looked particularly promising in that regard: NOVA1, which makes a protein that then guides the production of proteins from a number of other genes. The fact that it is mainly active only in the developing brain made it more attractive. And humans have a mutation in NOVA1 not found in other vertebrates, living or extinct.

Dr. Muotri’s colleague, Cleber Trujillo, grew a batch of organoids carrying the ancestral version of the NOVA1 gene. After placing one under a microscope next to an ordinary brain organoid, he invited Dr. Muotri take a look.

The ancestral NOVA1 organoid had a noticeably different appearance, with a bumpy popcorn texture instead of a smooth spherical surface. “At that point, things started,” Dr. Muotri recalled. “I said, ‘OK, it’s doing something.’”

The proportion of different types of brain cells was also different in the ancestral organoids. And the neurons in the ancestral organoids began firing spikes of electrical activity a few weeks earlier in their development than modern human ones did. But it also took longer for the electrical spikes to get organized into waves.

Other experts were surprised that a single genetic mutation could have such obvious effects on the organoids. They had expected subtle shifts that might be difficult to observe.

“It looks like the authors found a needle in a haystack based on an extremely elegant study design,” said Philipp Gunz, a paleoanthropologist at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, who was not involved in the research.

Simon Fisher, the director of the Max Planck Institute for Psycholinguistics in the Netherlands, said the results must have come from a mix of hard work and some good luck. “There must have been some degree of serendipity,” he said.

Although the researchers don’t know what the changes in the organoids mean for our evolutionary history, Dr. Muotri suspects that there may be connections to the kind of thinking made possible by different kinds of brains. “The true answer is, I don’t know,” he said. “But everything that we see at very early stages in neurodevelopment might have an implication later on in life.”

At the new research center, Dr. Semendeferi plans to carry out careful anatomical studies on brain organoids and compare them to human fetal brains. That comparison will help make sense of the changes seen in the ancestral NOVA1 organoid.

And Dr. Muotri’s team is working through the list of 60 other genes, to create more organoids for Dr. Semendeferi to examine. It’s possible that the researchers may not be so lucky as they were on their first try and won’t see much difference with some genes.

“But others might be similar to NOVA1 and point to something new — some new biology that allows us to reconstruct an evolutionary path that helped us to become who we are,” Dr. Muotri said.

Drugmakers Look for New Ways to Test Covid-19 Vaccines


As more Covid-19 vaccines become available in the U.S., it is getting tougher to run large clinical trials to test a new vaccine’s ability to prevent disease because people are less willing to take a placebo—forcing drugmakers and researchers to look for workarounds as they vet the next generation of shots and test new uses for authorized ones.

One potential workaround would be to determine what level of immune response a vaccine has to trigger to protect people from the coronavirus, as measured in blood samples, and to use that information to create smaller, faster and less-expensive clinical trials.

Instead of requiring tens of thousands of volunteers and costing several hundred-million dollars, such trials could involve only hundreds of people at a fraction of the cost. They could be used to speed the availability of new vaccines targeting emerging variants.

Moderna Inc., MRNA +1.71% Pfizer PFE -0.98% and its partner BioNTech SEBNTX -1.21% and a federally funded network of researchers are conducting analyses to learn what immune response is necessary for protection with current vaccines, known as an immune correlate of protection. They say it could come in handy for new studies of already-authorized vaccines—such as testing the shots in children or whether reduced doses are effective—as well as for trials of the next generation of shots, including those targeting new coronavirus strains. Over time, such knowledge could also help determine how long protection from the vaccines lasts.

Another workaround is to run future large efficacy trials outside the U.S., in places where viral transmission is high and vaccine availability is more limited. Arcturus Therapeutics Holdings Inc., whose Covid-19 vaccine is in mid-stage testing, may run a large Phase 3 trial of its experimental shot outside the U.S. because of the diminishing feasibility of running it in the U.S., Chief Executive Joseph Payne said in an interview. The company hasn’t disclosed which country or countries.

Large studies involving tens of thousands of people have been launched in the U.S. for five Covid-19 vaccines, including the two authorized for use from Moderna and Pfizer and a vaccine from Johnson & JohnsonJNJ -0.41% which plans to seek U.S. authorization soon. In these studies, researchers randomly assign the volunteers to receive either the vaccine or a placebo and then compare how many get sick with Covid-19 in each group.

Vaccine vs. Placebo

But it is becoming more difficult to run these placebo-controlled efficacy trials because prospective recruits increasingly want one of the highly effective authorized shots rather than an experimental shot or a placebo, researchers say. The challenge is heightened among groups that now have access to the vaccines, like health-care workers and the elderly.

In a large study of Novavax Inc.’s NVAX +1.15% vaccine, about 1.5% of the volunteers who were assigned to receive a placebo subsequently decided to get one of the authorized vaccines, Gregory Glenn, the company’s head of research, said in an online scientific forum this week. More than half of those making that choice were over the age of 65.

“People in the U.S. don’t want a placebo anymore if they’re in a group that can get the authorized vaccines,” said Dr. Kathleen Neuzil, a vaccine researcher at the University of Maryland who helps lead the federally funded Covid-19 Prevention Network, made up of research sites running large clinical trials of Covid-19 shots.

Covid-19 vaccines are designed to work by inducing a person’s immune system to produce antibody proteins that can neutralize the coronavirus. The immune correlate of protection is the concentration of those antibodies at a level that prevents Covid-19 disease; antibodies below that level aren’t protective, while at or above that level are protective.

The immune correlate of protection wouldn’t be definitive proof that a vaccine is effective at protecting people from disease, but it could be sufficient to guide regulatory authorization of new vaccines or new uses for existing vaccines, said Peter Gilbert, a biostatistician at the Fred Hutchinson Cancer Research Center in Seattle and part of the Covid-19 Prevention Network. Regulators would still require reports of any side effects to determine safety and might require further studies to confirm efficacy.

A Predictive Blueprint

Such correlates of protection have been used for past vaccine development. The Food and Drug Administration has approved certain meningitis vaccines based on their ability to induce an immune response that correlates with protection, rather than requiring large placebo-controlled efficacy trials.

“It saves time, it saves money and it may be the only thing that’s logistically feasible going forward,” Dr. Neuzil said.

Covid-19 vaccines researchers expect to determine the immune correlate of protection by comparing antibody levels in blood samples taken from vaccinated people who stayed healthy with antibody levels in the relatively small number of vaccinated people in the studies who still got sick from Covid-19.

Researchers from the Covid-19 Prevention Network are running analyses to try to determine the immune correlate of protection for Moderna’s vaccine within the next couple of months. They are examining some of the blood samples taken from all subjects about one month after the second dose in the large clinical study of the Moderna vaccine.

They plan to conduct similar analyses for other Covid-19 vaccines from J&J, AstraZeneca PLC and Novavax, which are being tested in trials run by the researchers’ network.

Pfizer and BioNTech are conducting their own analysis to determine the correlate of protection for their Covid-19 vaccine. A Pfizer spokeswoman said the company would explore the use of immune responses in additional studies of its vaccine, such as in pregnant women, children and people with compromised immune systems.

Leaders of the Covid-19 Prevention Network expect that future vaccines could be approved based on trials of only several hundred people, if results show that they had an immune response believed to be protective.

Moderna is exploring the use of an immune correlate of protection to test whether a half-dose of its vaccine could offer sufficient protection against Covid-19 disease, Chief Medical Officer Tal Zaks said at a recent investor conference.

An FDA spokeswoman said that when immune correlates of protection are established for these vaccines, they will be useful for a variety of studies including the evaluation of vaccines in children and assessing the response to new variants.

There are challenges for determining protective immune responses. There were relatively few cases of symptomatic Covid-19 in people who received the Moderna and Pfizer vaccines in the large studies at the time they were authorized, making statistically significant comparisons difficult. But researchers continue to follow study subjects and expect to see higher numbers.

The fast-spreading coronavirus variant is turning up in US sewers

A hyper-transmissible form of the coronavirus that causes covid-19 has been found in US sewer systems in California and Florida, confirming its widening presence in the US.

Buckets of dirty water drawn from sewer pipes near Los Angeles and outside Orlando starting in late January are among those in which genetic mutations shared by a so-called UK variant have been detected.

The UK strain B.1.1.7, a mutated form of the coronavirus  first discovered in southeast England in December, doesn’t seem to resist vaccines, but it does appear to spread more easily and has already taken over in countries including Israel, where it’s now responsible for 80% of cases. Some researchers have warned that if the variant takes hold in the US, it could become the dominant form by March.

The new sewage data is consistent with other estimates that the variant is increasing its reach. On January 7, the company Helix and researchers in California used patient test results to estimate that the variant is now responsible for 1% to 2% of cases in the US as a whole and 4% in Florida, about four times the proportion found in early January.

In the sewers of Altamonte Springs, near Orlando, tests on waste water suggest that 4% of those infected have the new variant. “I was thinking it would be in Miami or larger areas. But that was wishful thinking,” says Ed Torres, director of public works and utilities in Altamonte Springs, who oversees the sewage testing program.

A strain of covid-19 that appears to spread faster is colliding with the campaign to vaccinate Americans.

Local spread

Sewer tests are now providing a direct glimpse of just how many people are infected with the variant in some cities. Torres says a model he works with indicates that more than 200 people are infected with the variant just in his wastewater collection region, an exurb of 77,500 people.

Health officials in central Florida initially blamed their B.1.1.7 cases on visitors who tested positive, but the sewer tests in Altamonte indicate that the variant is spreading locally, too. Florida remains largely open for business, including theme parks, which are operating with the use of masks and physical distancing.

According to the US Centers for Disease Control and Prevention, only 611 cases of B.1.1.7 have been directly confirmed nationally via genetic sequencing of the viral samples collected in patient nose swabs, with many of the positives coming from California and Florida. Because only a small percentage of hospital swabs are ever analyzed for what form of the virus is present, the true number of B.1.1.7 cases is certainly much larger.

The US has less ability than some European countries to track variants of the virus because test swabs are not subjected to complete genomic sequencing as often, a situation some experts have likened to “flying blind” in the face of a changing pandemic.

According to the New York Times, as of January about 1.4 million people were testing positive for the coronavirus every seven days, but fewer than 3,000 of those clinical samples were being sequenced letter for letter, a step that’s usually necessary to see what mutations the virus has acquired.

Sewage surveillance

Wastewater offers a chance to monitor the variant more widely, and at lower expense. A single liter of dirty water carries the remains of viruses shed into toilets by everyone who shares a sewer system, offering a readout on the health of thousands, even millions, of people.

Since last spring, some cities have used molecular tests on sewage as an early warning system, since the amount of coronavirus in sewage can predict how many people will turn up in hospitals a week to 10 days later. The reason sewage results go up or down before official case numbers do is that people seem to start shedding the virus into toilets a day or two before they feel ill, and it often takes even more time to receive a test result.

“You can see the post-Thanksgiving spike, and the after-Christmas boom,” says Raul Gonzalez, who carries out tests on sewer water for a utility in Virginia Beach.

The dashboard COVIDPoops19, maintained by the University of California, Merced, tracks readings from over 1,000 sites in 47 countries. Testing experts say the virus or virus fragments in sewage aren’t alive or dangerous.

Initial results

Sewage tests use a sensitive version of PCR, the testing technology employed in hospital tests. Called digital PCR, it is also employed in so-called liquid biopsies to spot signs of cancer in a blood draw.

When the first reports of the B.1.1.7 variant emerged in December, GT Molecular, a company in Fort Collins, Colorado, was the first to reformat its sewage test to search for the variant. That involves checking sewage for two mutations characteristic of the B.1.1.7 strain.

The CDC ordered software that was meant to manage the vaccine rollout. Instead, it has been plagued by problems and abandoned by most states.

“We measure the amount of the virus that has the parental sequence and the amount of virus with the mutant sequences,” says Rose Nash, director of R&D at the company. “Most samples are not coming back positive, but what we are seeing across samples that do come back positive is about 5% levels of the variant. We are expecting that to increase.”

Because the sewage test has only been in use for a few weeks, it’s too soon to say on the basis of that evidence alone if variant levels are rising or falling. “There is no distinct trend yet in the variant,” says Mike Shaffer, who manages data for Oxnard, California, a beach city northwest of LA, which found low levels of the variant in January.

Shaffer says he also wants to confirm that the test is really finding the UK variant, not some look-alike strain, and says the Oxnard sewer samples were sent to Stanford University for sequencing. Christopher McKee, CEO of GT Molecular, agrees that the science behind the variant test “is still pretty nascent.”

According to GT Molecular, several other districts in California and Florida have positive results but haven’t announced them. “To me it’s worthless if you don’t put the information out there,” says Torres. “We need to communicate it to people who can do something about it.”

Soo far, however, news of the variant in sewage or its spread has not led to any major change in public policy. As yet, there is no national plan to deal with the threat that the variant will spread quickly and cause case numbers to rise.

Robert Levin, the public health officer for Ventura County, where Oxnard is located, presented the sewage findings to a meeting of a supervisory board last week. “They found a tiny amount of the variants that are considered hyper-transmissible,” Levin said. “The impact that this will have on our county cannot be predicted. This is uncharted territory.”

How vaccines are made, and why it is hard

Nine vaccines against covid-19 have already been approved in one jurisdiction or another, with many more in various stages of preparation. That this has happened within a year of the illness coming to the world’s attention is remarkable. But it is one thing to design and test vaccines. It is another to make them at sufficient scale to generate the billions of doses needed to vaccinate the world’s population, and to do so at such speed that the rate of inoculation can outpace the spread and possible mutation of the virus.Listen to this story

Broadly, there are two ways of making antiviral vaccines. One, tried and trusted, involves growing, in tanks called bioreactors, cell cultures that act as hosts for viruses which are then used in one way or another to make the vaccine in question. Cells grown this way can be of many types—insect, human kidney, monkey kidney, hamster ovary—as can the resulting vaccines. These may be weakened or killed versions of the virus to be protected against, or live viruses of a different and less-dangerous sort that carry a gene or two abstracted from the target virus, or even just isolated target-viral proteins. The point is that the vaccine should introduce into the body, or induce that body to make, something which the immune system can learn to recognise and attack if the real target virus should ever turn up.

In with the new

The alternative method, developed recently and employed to make the mrna vaccines, such as those of Moderna and Pfizer, that the pandemic has stimulated the invention of, requires culturing cells only at the beginning of the process. mrna is the substance that carries instructions about how to make a protein from a cell’s dna to the molecular factories, known as ribosomes, which do the actual manufacturing. In the case of covid-19, the instructions in question generate spike, a protein found on the surfaces of particles of sars-cov-2, the virus that causes this illness. Suitably packaged and delivered, such mrna can induce some of the body cells of the inoculee to turn out spike, which the immune system then learns to recognise. To make this type of vaccine you therefore have to generate lots of the relevant mrna.

That process does indeed start with cells, though they are bacterial cells, rather than those of animals. But it does not end with them. The bacteria used, normally a well-understood species called E coli, have spliced into them a dna version of the part of the sars-cov-2 genome which describes spike. (Confusingly, as is true of many viruses, sars-cov-2’s actual genes are made of rna.) The bacteria are then allowed to multiply for a few days before being broken open, their dna filtered out, and the dna versions of the spike gene extracted as what is known as a dna template.

Once purified, this template is mixed with a soup of pertinent enzymes and fed molecules called nucleotides, the chemical “letters” of which rna is composed. Thus supplied, the enzymes use the templates to run off appropriate mrnas by the zillion. These are extracted and packaged into tiny, fatty bubbles to form the vaccine.

Both the cell-culture and the mrna approaches have benefits and drawbacks. The former has the advantage of being well established. Versions of it go back to vaccine-making’s origins. But keeping cultured animal cells alive and healthy is a tricky business. A whole subfield of bioengineering is dedicated to this task. Vaccine-makers who rely on live cultures constantly struggle with yields. Using this method to make a lot of vaccine, fast, is hard.

It was difficulties of this sort that Pascal Soriot, boss of AstraZeneca, cited on January 26th in defence of his firm’s failure to provide vaccine supplies which the European Union claimed it had been promised. AstraZeneca is an Anglo-Swedish company that, in collaboration with Oxford University, created one of the first vaccines to be approved. As Mr Soriot told La Repubblica, an Italian newspaper, “You have glitches, you have scale-up problems. The best site we have produces three times more vaccine out of a batch than the lowest-producing site.”

De-necking the bottles

Maximising a bioreactor’s yield is as much an art as a science. The underlying health of the cells involved matters. So do environmental conditions at the manufacturing site. That AstraZeneca has not been able to meet its own production targets shows how hard it is to predict when the right balance of biology will be found. The company says it can take six to nine months to start a production site up from scratch, and that even this timetable is possible only by working with experienced partners and at an accelerated pace. At the moment, AstraZeneca is working with 25 manufacturing organisations in 15 countries to make its vaccine.

Producing mrna vaccines at scale has problems, too. The biggest is how to protect the mrna molecules both from the environment they must travel through in order to reach the arm of their recipient, and from the recipient’s own body, which will attack them as they journey to the ribosomes which will transcribe them.

Protection from the environment is mainly a matter of having a strategically located set of refrigerators, known as a cold chain. Protection from the body, though, is where the fatty bubbles come in.

Production of these bubbles was a cottage industry before the pandemic. A small Austrian firm, Polymun Scientific, is one of just a handful that can make them. Their main previous use was in niche cancer treatments. Scaling up their production, which is happening right now, has never been done before and adds uncertainty to the continued supply of mrna vaccine.

There are other bottlenecks, too. In particular, the factories in which vaccines are made must be built to a high standard, known as gmp, for “Good Manufacturing Practice”. There is currently a shortage of gmp facilities. Andrey Zarur, boss of GreenLight Biosciences, a firm in Boston that is developing an mrna vaccine, says his company has employees whose entire job, at present, is to work the phones trying to find gmp facilities in which to make their vaccine. There is, though, nothing available. He is therefore looking to buy firms whose vaccine candidates have turned out not to work, simply in order to acquire the facilities in question.

Supplies of raw materials such as nucleotides are also tight. According to Dr Zarur, Thermo Fisher, an American chemical-supplies company, has spent $200m on a new facility in Lithuania to make these molecules, though the firm itself would not confirm this.

On top of all this, the transport and distribution of vaccines once they have been made presents yet further challenges, and concomitant potential for hold ups. Vaccines must be stored in special non-reactive glass vials. Some, such as the current version of Pfizer’s mrna vaccine, must also be kept at extremely low temperatures, though that problem may go away soon. Drew Weissman, one of the inventors of mrna-vaccine technology, says producers are now testing shots which are stable for three months when kept at 4°C.

Once supply chains for both cell-culture and mrna vaccines have been scaled up, and bottlenecks unblocked, the manufacturing processes may face a different test—how quickly they can produce new vaccines to deal with new viral variants as these emerge. The continued efficacy of approved vaccines against such variants is not guaranteed, and it may be necessary to make others (see article).

Here, the mrna approach may have an advantage. Its production systems will require a simple tweak—the dropping in at the start of a dna template describing the new variant’s spike protein. Cell-culture systems, by contrast, will have to be rebuilt to some degree for every new variant they aim to vaccinate against.

Scale models

Producers, such as those in China, who use older-fashioned cell-culture techniques, will have to recalibrate their entire operations. Newer systems, like AstraZeneca’s, which use cells specially designed so as not to be influenced by the new version of the spike gene in the viruses they are carrying, should be able to get on track in the time it takes to start a culture from scratch—about a month. For mrna systems, Drs Weissman and Zarur say it would take a couple of months to go from new variant to large-scale vaccine production. If variants resistant to the current crop of vaccines do evolve, then that speed and certainty in making new vaccines to combat them will be essential.