Breathless: The Scientific Race to Defeat a Deadly Virus

by David Quammen

presumably someone read these books to Trump or, more likely, summarized them for him.

But community spread would come, Messonnier admitted. It was not a question of if but of when, she said, “and how many people in this country will have severe illness.”

(Nonpharmaceutical interventions, or NPIs, is a fancy term for behavioral modifications to slow the spread of a disease, such as school closures, stay-at-home orders, social distancing generally, and wearing masks.)

thanks to draconian versions of nonpharmaceutical intervention, China would flatten its curve like a mesa. America’s curve over the coming months, into the following year, would resemble the Grand Tetons.

A national ethos of “rugged individualism,” by which I mean not real individualism but programmatic self-concern, to the detriment of community weal, was probably a large part of the reason.

China is the world’s leading consumer of pork, also the world’s largest source, producing about 50 percent of the global supply.

In 2018, the average consumption in China was seventy-five pounds per person. But in late summer that year an outbreak of African swine fever swept into the country, eventually affecting more than 150 million pigs. The disease is caused by African swine fever virus (ASFV), a DNA virus endemic to sub-Saharan Africa, where its reservoir hosts are bushpigs and warthogs, and ticks vector the virus from one animal to another. With the arrival of European colonizers, bringing domestic swine to Africa, ASFV infected those pigs too. In the twentieth century it got to Europe, then was eradicated, then returned in this century, possibly by way of wild boar imported to southern Belgium from Eastern Europe for the amusement of hunters. Among domestic pigs, the virus is highly virulent, and the strain that reached China, by August 2018, was almost 100 percent lethal. Within eight months, because China normally produces such a large share of the world’s pigs, commodities analysts in the West were cheerily telling their industry clients that ASFV’s impact there would “lift all protein boats.”

For the first nine months of the pandemic, SARS-CoV-2 seemed to be mutating slowly and evolving little or not at all. Some scientists noted “remarkably low” genetic diversity among the many samples that had been sequenced. One group of researchers, based at the Walter Reed Army Institute of Research, in Silver Spring, Maryland, looked at 27,977 genome sequences from infected people in eighty-four countries and found “little evidence” of natural selection for anything new. The virus, they wrote, “is being transmitted more rapidly than it evolves”—meaning that the discernible mutation rate was less than one nucleotide change per human case. It was passing from person to person, in many instances, without a single copying mistake in its thirty-thousand-letter genome. That’s notable constancy, especially for an RNA virus. It didn’t mean that the virus might not have mutated at all, within a given person, but that, if there were mutations, they proved unsuccessful in competition with the unmutated strain at getting themselves replicated and transmitted. Evolutionary biologists call that purifying selection. Purify what you’ve got. If it ain’t broke, don’t fix it. Maybe this virus didn’t need to evolve because it was already succeeding so well among humans.

Coronaviruses lie at one end of the RNA virus spectrum. Their genomes are exceptionally long and their mutation rates are exceptionally low, less than a tenth the rate in other RNA viruses. Those two atypical characteristics are mediated by a nifty mechanism, a special protein called nsp14 (which could stand for “nifty special protein 14,” but doesn’t). This protein performs a proofreading function, tracking along the genome while it replicates itself, letter by letter, and correcting most mistakes before they can get into new virions. With genomes so long, if they didn’t have such a mechanism, coronaviruses would accumulate so many errors that they would rattle to pieces, like a Model A Ford in which someone had neglected to tighten the bolts. That outcome is called “error catastrophe.” Evolution supplied nsp14, allowing coronaviruses to avoid error catastrophe—by letting them tighten their bolts. A long genome was an advantage, in ways still unmeasured, and the proofreading function of nsp14 made it possible.

Everybody who has ever thought about evolution knows the story about the moths, right? The change in the moths in the trees,” Korber said. A famous story, yes: she meant the replacement of light-winged forms of the peppered moth (Biston betularia) by dark-winged forms, as the tree trunks of industrialized England became darkened with soot. A mutation for dark wings allowed some moths to escape the notice of predatory birds, and those dark-winged mutants left the most offspring.

People grow cynical: What’s the latest scariant?

by May 10 the state of New York had more than 330,000 cases, most of those in New York City, accounting for 8 percent of the world total.

His young colleagues assigned nicknames, within their informal lab culture, to the mutations. D614G was known as “Doug.” If the genome lacked that mutation, retaining the original form, it was “Douglas,” as in Doug-less. N501Y became “Nelly.”

The virus seems to have first reached Brazil from Italy, by way of four travelers arriving in São Paulo during February 2020, just at the time when Italy’s own initial outbreak began in Lombardy. From there it moved north quickly, meeting a disorganized public health response exacerbated by political turmoil in some states (such as Rio de Janeiro, where the governor faced impeachment), acute shortage of resources in others (Amazonas, a huge state with few ICU beds and all of those concentrated in Manaus), prolonged circulation of the virus before the first case was reported (Ceará state, on the north coast), socioeconomic inequities, and a poorly informed president (Jair Bolsonaro) with autocratic tendencies who rejected nonpharmaceutical interventions (masks, social distancing), promoted hydroxychloroquine as a treatment, fired his minister of health in April, lost another minister of health in May, and then appointed to the job an army general with no medical qualifications. On April 28, a day when Brazil’s total cases exceeded 73,000 and the death toll passed five thousand, Bolsonaro faced a group of reporters at a barricade, and someone mentioned the numbers. “So what?” he said with a shrug. “I’m sorry. What do you want me to do?”

Gamma

George Macdonald studied malaria, a very complicated disease caused by a protozoan that passes from host to host in the bites of mosquitoes, and the R0 for that microbe depends on several factors, such as the density of mosquitoes, their longevity, how many times each mosquito may bite, and how long an infected human remains infectious to other mosquitoes. The reproduction number could be wildly various, Macdonald noted, ranging from 1.0 upward to who knows what. If there were long-lived and voracious mosquitoes biting people with long-enduring active infections, it might be 735. No wonder Macdonald and his public health colleagues couldn’t eradicate malaria, and seven decades later the disease still kills almost a half million children each year.

Here’s why I’m torturing you with these math details, which seem dry but are actually very wet.

Ha

The number R0 plays a big role in determining the supposed threshold at which herd immunity takes effect. That threshold is a percentage level—this much of the population has passed from susceptible to infected. And what does the threshold mark? It marks the point at which R has fallen to 1.0. Transmission has slowed because susceptibles have become fewer, and therefore harder for the virus to locate. Chains of infection come to dead ends. R keeps falling. From then on, so long as the reproduction number remains below 1.0, the outbreak subsides. All things being equal, it will decline to a tolerably low level of endemic disease or, best-case scenario, the virus will disappear entirely from that population because it can’t find anyone new to infect. That’s how smallpox eradication was achieved. The last active smallpox strains in the last few human cases were not killed by antiviral drugs. They died of isolation and loneliness, without leaving offspring.

Smallpox was special. It produced visible symptoms in each infected person, so cases could be identified and isolated while health workers vaccinated everyone around them, stopping viral spread by containment; and it dwelt within no other reservoir than humans, so it couldn’t reemerge from a nonhuman animal host. Polio eradication has been harder. That virus too inhabits no other host, but many polio infections are asymptomatic, making them harder to identify and isolate. Eradication ...

The equation used to calculate herd immunity is very simple, so simple that even I can understand it and you can too: threshold = 1 - 1/R0. He prints equation. Eyes roll back in heads. But no, wait, look how easy this is. If the reproduction number is three, meaning three secondary cases infected by each primary case, you need nothing more than grade-school fractions to work it out: 1 minus 1/3 equals what? Two thirds, right? So, in that case, the threshold for herd immunity is two thirds: 67 percent of the population.

Measles virus was zoonotic in origin, having diverged from the cattle virus that caused a disease called rinderpest. Here’s our punishment (among others) for domesticating bovids: measles got into humans from our cattle and has been with us for roughly two thousand years. Rinderpest has now been eradicated but the measles virus in humans, not. The reason for that, I think, is that it has been easier for us to contain and control bovine behavior than our own.

Hydroxychloroquine has been around for a long time, but it entered the headlines in late March 2020, after Fox News host Laura Ingraham, with help from Larry Ellison, the cofounder of Oracle, and possibly also Elon Musk, the entrepreneur and spaceman, planted that word like an earwig in Donald Trump’s brain. Trump, as you may have heard, is not a scientific sophisticate. He would become slightly confused some months later when discussing what he called “herd mentality,” by which he seems to have meant herd immunity (although no one can know for sure).

And this well-organized trial ended early, before reaching the hoped-for number of enrolled patients, for a reason that may seem, to those of us in the rest of the world, enviable and weird: Wuhan ran out of Covid patients. The strict lockdown of that city, the testing and tracing and isolating, all the nonpharmaceutical interventions, had made pharmaceutical interventions unnecessary. On the day that the Wuhan remdesivir study was published, April 29, 2020, China at large had a total of four new cases and no new deaths.

Then again, the claw hammer is an important and trusted tool among carpenters, but it’s not recommended for use in dentistry.

Re Ivermectin

RSV had a low profile in public awareness, but it was the leading cause of hospitalization for children under five years old, and it killed upward of sixty thousand in a typical year, possibly as many as 200,000, with 99 percent of those in developing countries. A brutal threat to poor kids in unprivileged circumstances. But this virus presented some special challenges for creating a vaccine: it attacked victims at such a young age; it possessed some tricks for immune evasion; it could reinfect a child, or an adult, after recovery from a first infection; and there was an unfortunate history, dating back to the 1960s, of RSV vaccine efforts gone amiss during clinical trials, in which a trial vaccine made matters worse, provoking a crisis of inflammation when the virus hit. The polite name for this dark sort of event was “vaccine-enhanced disease.”

The ingenious concept of using mRNA was that the antigen would be produced—in mass quantities, from a set of genetic instructions—inside the patient’s own body. That idea went back decades, to several sources, among the earliest of whom was Jon A. Wolff, at the University of Wisconsin School of Medicine and Public Health, who published on it in 1990.

The name itself, Moderna, was a portmanteau of “modified” and “RNA,” with the added appeal that it contained “modern.”

On March 16, 2020, the day Moderna began its Phase 1 trial, the United States recorded “only” twenty-two COVID-19 fatalities, but within a month the daily death count rose a hundredfold. On December 18, the day Moderna’s vaccine received emergency use authorization from the U.S. Food and Drug Administration, 3,171 Americans died of Covid, as well as ten thousand other people around the world.

A year after the second field trip, this team published a paper containing some of these details (including the forest cobras) and one headline result: they had isolated live virus from five of the bats. The reservoir host—or anyway, a reservoir host—of Marburg virus had been found. It was safe to say, thanks to Towner and Amman and Swanepoel and their colleagues, that the outbreaks at Kitaka and Durba, probably also those cases linked to the cave on Mount Elgon, and possibly too the virus in the African monkeys shipped to Europe in 1967, originated from Egyptian fruit bats. That discovery had taken only forty-one years.

During that brief period when the preprint was up, while numerous scientists tore into it on Twitter and elsewhere, Tony Fauci expressed some reactions privately, later revealed among his published emails. To the director of the NIH, Francis Collins, Fauci forwarded an op-ed on origin stories, noting, “The Indian paper is really outlandish.” To colleagues within his office, on an email thread requesting his guidance for a possible response, he wrote, “Geeeez.”

Deer mice are the most abundant (nonhuman) mammals in North America.

In the lingo of disease ecologists: a sylvatic cycle. That term comes from the Latin word sylva, meaning forest. A virus with a sylvatic cycle is two-faced, like a traveling salesman with another wife and more kids in another town. Yellow fever virus, for example: transmitted by mosquitoes, it infects humans in cities (the urban cycle) when the right mosquitoes are present, but it’s broadly enough adapted to infect monkeys also, and it does that in some tropical forests (the sylvatic cycle), circulating in monkey populations. Yellow fever can be eliminated in cities by vaccination and mosquito control, but whenever an unvaccinated person goes into a forest where the virus circulates, that person can become infected, return to the city, and trigger another urban cycle, if some mosquitoes are still there to help. Yellow fever virus has never been eradicated, and travelers to many tropical countries are still obliged to be vaccinated, because the sylvatic cycle will persist, and threaten another urban cycle, until you kill every mosquito or vaccinate every monkey.

if the world’s forests or other natural ecosystems contain populations of wild animals in which that virus circulates, either because they are the original reservoir hosts (horseshoe bats in southern China?) or because they have become infected by contact with humans (mink in Utah? deer mice in Westchester County?), then there is no end to COVID-19. (There is probably no end to it regardless, but that’s another matter, to which I’ll return.) There is no herd immunity where there is a sylvatic cycle. An unvaccinated person has contact with an infected wild animal (a mink, a cougar, a monkey, a deer mouse) during some activity (hunting, cutting timber, picking fruit, sweeping up urine-laced dust in a cabin) and becomes infected with the virus, potentially triggering a new outbreak among people. You could vaccinate every person on Earth (that’s not gonna happen) and the virus would still be present around us, circulating, replicating, mutating, evolving, generating new variants, ready for its next opportunity.

Oof

The two gorillas recovered within weeks, although not before one animal, a forty-eight-year-old silverback with heart disease named Winston, had been treated with monoclonal antibodies.

His heart disease was named Winston 🤣

Being 96.2 percent similar at the level of nucleotides, as it is, RaTG13 differs by 3.8 percent. Given the pace at which coronaviruses generally mutate and evolve, that reflects about fifty years of evolutionary divergence. RaTG13 differs from the baseline virus first detected in Wuhan (known as Wuhan-Hu-1, the sequence released by Zhang and Holmes) at about 1,150 nucleotide positions, and those positions are scattered throughout the genomes. Some of the world’s best evolutionary virologists (professionals in that field, not amateurs visiting it) and coronavirus experts, including Susan Weiss, Stanley Perlman, David Robertson, Robert Garry, and Kristian Andersen, assure us that RaTG13, with or without lab manipulation, is not the answer to the origin question for SARS-CoV-2.

gain-of-function (GOF) work is any sort of laboratory experiment that increases some biological capacity of an organism. More specifically, “gain-of-function research of concern” (GOFROC) is work on a pathogen (viral or whatever) with pandemic-causing potential that might make it more able to infect humans, to transmit among them, or to cause them greater harm.

The core issue over gain-of-function research is whether it creates what policy people call potential pandemic pathogens (PPP, another baleful string of letters for your mental toolbox). A potential pandemic pathogen is a bug that’s highly transmissible, capable of uncontrollable spread among humans, and could cause widespread illness and death. By that definition, the first milestone of PPP work in the twenty-first century came in 2005, when a team of researchers from the U.S. CDC and elsewhere reconstructed the 1918 influenza virus. They did that by assembling its genome from old autopsy specimens and some frozen lung tissue of a victim buried in the Alaska permafrost, then activating that into a live virus by reverse genetics, expressing the viral genome in cultured cells. This was controversial. One scientist called it “a recipe for disaster.” The researchers defended their work on grounds that reviving the virus and studying it in a secure lab not only illuminated why that virus had been so deadly but also led toward important insights about influenza viruses generally, with relevance to vaccine development, antiviral treatments, and prediction of virulence in other viruses.

the seventh-largest city in the world. Mumbai contains twelve million people in the city proper, with a population density of roughly 73,000 people per square mile, among the highest anywhere on Earth.

Scientists have cloned cats, dogs, deer, and other animals. Using microsurgical equipment, you extract the nucleus from one cell of a leopard in captivity at SGNP. This is the leopard you seek to clone—to replicate as closely as possible in a new animal, its genetic twin. It might be a male or a female. Call it Donor 1. You also remove the nucleus from an egg cell, an oocyte, taken from another leopard, necessarily a female (because only females produce eggs). Call her Donor 2. You insert your chosen nucleus into that egg cell, then you activate the thing to begin dividing. When it has developed to the point of comprising a couple hundred cells, those cells will be contained inside a spheroid mass, with a protective layer, called a blastocyst. This is your preimplantation embryo. The nuclear genome in each of those cells is identical to the genome of your clone target, Donor 1. And the cytoplasm of each cell, the gelatinous fluid surrounding the nucleus, contains mitochondrial DNA, a subsidiary genome, inherited only from that female who gave you her egg cell, Donor 2. You surgically implant the embryo in the uterus of a foster mother. For convenience, you might choose a smaller and less fearsome cat for this motherhood role—for instance, a Bornean clouded leopard, a gracefully spotted feline about the size of a border collie. If you are lucky and adept, your Bornean clouded leopard carries the embryo to term and gives birth to a true leopard, the genetic twin—in its nucle...

The mayor of Wuhan ignored the warnings of a highly contagious virus spreading through the city, and on January 19, 2020, the potluck went ahead as planned. Four days later, the city government suspended all public transportation, the provincial government cut the highway links, and Wuhan was locked in quarantine. But that was too late.

Again, they described the elements of their model: the circulation phase, when the virus is spilling from one kind of animal host to another, remaining broadly adapted, capable of infecting them all; the first accident, when a genetic change happens to give the virus added advantage in humans; and the second accident, a societal circumstance that allows the virus to amplify and spread, approaching a critical point. (This is like what Malcolm Gladwell described in his book The Tipping Point.

Most of that noise was made by amateurs and pundits who had swotted up a bit of virology for the occasion. (I’m an amateur in this field too, but early in 2021 I decided to keep quiet for a while and listen.)

Suspicion that the virus might have a lab origin, they wrote, “stems from the coincidence that it was first detected in a city that houses a major virological laboratory that studies coronaviruses.” But the link to Wuhan, they added, more likely reflects the facts that it’s the largest city in central China, a hub for travel and commerce, densely populated by eleven million people, with multiple animal markets in its midst. Yes, Holmes and his coauthors agreed, the possibility of a lab accident can’t be entirely dismissed. Furthermore, that hypothesis may be nearly impossible to disprove. But it’s “highly unlikely,” they judged, “relative to the numerous and repeated human-animal contacts that occur routinely in the wildlife trade.” Failure to investigate that zoonotic dimension, with collaborative studies, crossing borders between countries and boundaries between species, would leave this pandemic festering and the world still very vulnerable to the next one.

ascertainment bias, a scientific pitfall. In plain words: if you think that a certain factor is what you’re looking for, in a scientific study, you will be more likely to find it, because you will look in places where that factor lives.

Have you ever tried to write a report by committee? This was worse.

Among the Chinese points of counterevidence, Daszak recalled, was a law constraining the sale of live, wild mammals. If it was illegal, it must not have happened.

Critics in the West suggest that China has been covering up a laboratory accident. Maybe that’s not the point. Maybe that’s not the explanation for why Chinese officials have resisted a phase 2 study, and why they favor the cold chain hypothesis, which is plausible but for which there seems to be no evidence. Maybe the point is that China has been embarrassed by, and is covering up, not a lab leak but an animal leak.

The wild animals of China are considered state property; penalties for capturing and trading animals of a protected species (such as raccoon dog or bamboo rat or civet) include a three-year prison term and fines. The takeaway from Xiao’s study is that, as of November 2019, none of this seemed to be enforced.

Gronvall noted such negligence as a possible reason why the Huanan market was quickly closed, emptied, and sterilized on January 1, 2020, immediately after city officials learned of its connection to the outbreak of pneumonias. The wildlife traders disappeared. “The swift clear-out of the market may have been intended to protect them as well as the law-enforcement officers and local politicians who had looked the other way,” Gronvall wrote. “Once it looked as if a disease that came from the market was spreading, sweeping illegal activity under the rug would be a priority to avoid blame and hold onto profits.” For the pandemic to have originated in China was “embarrassing” enough, she wrote. Worse still if it reflected negligence in law enforcement and possible corruption. “Although the wildlife trade was a known disease risk, and there were laws restricting it on the books, it wasn’t stopped.”

Yes, the Intelligence Community (IC) of the U.S. is a thing, established in 1981 as a formal assemblage of agencies that now number seventeen, including of course the CIA and the Office of Naval Intelligence (ONI), the Defense Intelligence Agency (DIA), the National Security Agency (NSA), and the Intelligence Branch (IB) of the FBI, but also Marine Corps Intelligence (MCI), Space Delta 7 (within the United States Space Force, whatever that is), as well as branches of the Coast Guard, the Department of the Treasury, the State Department, the Department of Homeland Security, and the Drug Enforcement Administration, among others. A bodacious aggregation of intelligences.

argumentum ab auctoritate, argument from authority,