Where Medicine Now Stands In the Fight Against Covid

Doctors on the front line of fighting the disease are skeptical that a vaccine will be able to stop the pandemic. As medical knowledge about how to help patients fight the virus has improved, techniques not visible six or seven months ago to those in China, Italy and New York, are now widely shared and applied.

But, as in the fight against tuberculosis and HIV, incremental treatments with anti-virals and antibodies - perhaps in combination with a vaccine or possibly on their own - could eventually stop the spread and the severity of the disease. JL

Dhruv Khullar reports in The New Yorker:

We need to be prepared for the possibility that it will be the tuberculosis model—incremental progress on multiple fronts—that gets us through the coronavirus pandemic. Vaccine research programs aren’t the only ones progressing at unprecedented speed. Three kinds of therapies—antiviral drugs, antibodies, and immunomodulators—may be ready soon. In combination with a vaccine, they could help us turn the tide.giving antivirals to those who never require hospitalization could change the course of the pandemic.

The first outbreak of polio in the United States struck Rutland County, Vermont, in the summer of 1894. The disease began with fever, sore throat, and fatigue; it sometimes went on to damage the brain and spinal cord, paralyzing or even killing its hosts. Charles Caverly, a local physician, chronicled the devastation using detailed maps and notes. “Boy, 10 years; died within twenty-four hours with convulsions,” he wrote. “Boy, 10 months; died on sixth day, all extremities paralyzed. . . . Girl, 11 years; died on third day, no paralysis noted. . . . Male, 22 years; died on third day, both legs paralyzed.” Within weeks, a hundred and thirty-two people, mostly children, had been paralyzed, and eighteen had died.

In the coming decades, polio became a familiar menace. Summer, when the virus exacted its heaviest toll, was dubbed “polio season.” The virus crippled children and adults, often paralyzing their respiratory muscles and confining thousands to iron lungs. In 1916, New York City recorded nine thousand cases of polio and six thousand deaths. In August of 1921, Franklin Roosevelt, then a thirty-nine-year-old lawyer, fell off a sailboat and into the icy waters of the Bay of Fundy; the next day, he noticed lower-back pain, and within a week he could no longer stand. The pace and size of outbreaks accelerated. Even though the polio death rate declined in the decades that followed, owing to advances in medical care, the virus still disabled more than thirty-five thousand people a year during the nineteen-forties. In 1952—the year the virus peaked in America—nearly sixty thousand people were infected, and more than three thousand died. Parents refused to let their kids play outside. Cities introduced social-distancing measures. Summer camps were cancelled; schools were shut down; bars and theatres closed.

As bad as polio was, it wasn’t the only infectious disease stalking mid-century America. In the early nineteen-hundreds, tuberculosis began to compete with influenza for the title of the world’s deadliest disease. TB, which is caused by a bacterium that burrows into the lungs and other organs, spreads through coughing, sneezing, and talking; the Industrial Revolution, which brought huge crowds together in urban neighborhoods and workplaces, made it more likely to spread. In some people, TB lies dormant, sometimes for decades. In others, it becomes active, causing a quick and violent death: bedsheets drenched in sweat, sputum mixed with blood, a wasting away so severe that the disease was known as consumption. TB’s toll was greatest among vulnerable populations: immigrants, prisoners, the poor. The death rate for minorities was three times higher than for whites. In 1953, the United States saw more than eighty-four thousand new infections, and nearly twenty thousand deaths. In 1962, forty-one years after her husband was diagnosed with polio, Eleanor Roosevelt died, at age seventy-eight, from complications of tuberculosis.

The total elimination of polio in the United States and the near-complete eradication of tuberculosis are two of this country’s greatest public-health success stories. And yet they are strikingly different in their details. Polio was defeated by a silver bullet: Jonas Salk introduced a polio vaccine, administered by injection, in 1955; a few years later, Albert Sabin developed an oral version. As a result, in the nineteen-sixties, the number of polio cases plummeted to fewer than a hundred, and in the seventies it fell to fewer than ten; no one has contracted the virus in the United States since 1979. By contrast, there is no comparable vaccine for tuberculosis. The B.C.G. vaccine, introduced in 1921, is effective against TB only about half the time, and is not given to the vast majority of Americans.

Instead, TB has been beaten back incrementally, using a host of imperfect medical and public-health advances. In the nineteen-forties, Selman Waksman, a microbiologist at Rutgers University, isolated streptomycin, the first antibiotic effective against tuberculosis. Streptomycin had serious side effects, including nerve damage, and quickly led to bacterial resistance; as a result, several other drugs were developed. Used alone, none works particularly well, but a cocktail given over a period of months can cure the infection. Meanwhile, testing for TB became readily available across the United States. (For some people, such as hospital workers, it became required.) Public-health protocols were developed: new cases of TB were reported immediately to departments of health, which moved to isolate patients and trace their contacts. (Those systems remain in place today: as a physician, the first time I wore an N95 mask was while treating a patient with TB.) Along with ongoing improvements in American housing and sanitation, these developments made the disease both less infectious and less deadly. Today, around the world, ten million people contract TB each year, and a million and a half die of it; multidrug-resistant tuberculosis, which is harder to treat, is spreading in some developing countries. But the United States is recording fewer tuberculosis infections now than at any point in the nation’s history: about nine thousand cases each year, and around five hundred deaths.

In fighting the new coronavirus, we all want a silver-bullet cure: the polio model. In April, the federal government launched Operation Warp Speed, a ten-billion-dollar effort to expedite, through government-industry partnerships, the development, manufacturing, and distribution of a coronavirus vaccine. The goal is ambitious: three hundred million doses by January, 2021. Anthony Fauci, the director of the National Institute of Allergy and Infectious Diseases, believes it is plausible. “My hope, my expectation, is that we’ll have not one but multiple vaccines in 2021,” he told me.

In the popular imagination, a coronavirus vaccine will bring the pandemic to a decisive end. And yet not all vaccines are as powerful as the one Salk developed. Many vaccines are only partly effective, or work better for some age groups than others; the immunity a vaccine confers can wane with time, and a shot that’s hard to manufacture or distribute could remain unavailable to many of us. Last week, the Centers for Disease Control sent a letter to state governments telling them to prepare for the possible distribution of a coronavirus vaccine this fall; it described the progress of “Vaccine A” and “Vaccine B”—almost certainly the vaccines being developed by Pfizer and Moderna, respectively. Though these vaccines are promising, there is no guarantee that they will be cure-alls. “Unless you have a perfect vaccine, which very few are, you’ll always have people who end up getting sick,” Fauci said. “With or without a vaccine, we’re going to need other treatments.”

We could get lucky. But we need to be prepared for the possibility that, in the absence of a single-shot cure, it will be the tuberculosis model—incremental, simultaneous progress on multiple fronts—that gets us through the coronavirus pandemic. It’s a good thing, then, that vaccine research programs aren’t the only ones progressing at unprecedented speed. Three kinds of therapies currently in development—antiviral drugs, antibodies, and immunomodulators—may be ready soon. Alone or in combination with a vaccine, they could help us turn the tide.

There are lots of ways to fight back against sars-CoV-2 and covid-19, the disease it causes. We can limit the virus’s spread in the population at large; we can also build barriers against infection for at-risk people, such as caregivers or essential workers, in particular. We can devise therapies that prevent the newly infected from getting worse, and we can create interventions that target the sickest and give them a fighting chance. By surrounding the virus in this way, we can make it less contagious and lethal, changing the character of the pandemic.

Vaccines take the first approach: they fight the virus by hampering its spread in the general population. According to the World Health Organization, there are at least a hundred and seventy vaccine trials taking place around the world. The two vaccines that the C.D.C. is watching for this fall could be ready so soon because they are being produced using a new method, developed within the past decade, that promises to be vastly faster than anything that’s come before. In the past, a vaccine scientist might have isolated a pathogen, grown it in eggs or cells, and injected a weakened, dead, or sliced-up version into a patient, hoping to provoke a lasting immune response. Each virus required its own approach, the process took years, and the results were variable. Today, she can use new advances in vaccine “platforms,” or delivery vehicles, to move faster. A platform is like a car that can transport different passengers. Before, building a vaccine required building the car. Now, the genetic material of the sars-CoV-2 virus can be loaded onto a harmless, pre-designed virus—or, in the case of Moderna’s vaccine, onto messenger RNA—and delivered into human cells with relative ease.

Using its platform, Moderna developed its vaccine, mRNA-1273, within forty-two days of seeing the virus’s genetic sequence; the company took just six months to complete Phase I trials, a process that historically has taken three to nine years. “If you have a good platform, you can basically plug in any antigen,” Boris Juelg, an assistant professor at Harvard who is working on a vaccine with Johnson & Johnson, told me. “From a safety perspective, it’s nice to know a vector has already been given to thousands of people. You feel really good about that.” Still, no matter how a vaccine is developed, it takes time to prove it safe and effective. Even those leading the government’s vaccine efforts have signalled that the timeline laid out by the Administration is either extraordinarily optimistic or unrealistic. Larry Corey, who leads a network of vaccine-trial sites supported by the National Institutes of Health, has said that it will likely take seven months, from the start of the first clinical trials in July, to learn if a vaccine works: that “simple math,” he says, suggests that meaningful results won’t arrive until February, at the earliest. According to Moncef Slaoui, the head of Operation Warp Speed’s vaccine-development efforts, there is a “very, very low chance” of a vaccine being approved by Election Day. (Slaoui has said that he will resign if he feels political pressure to green-light an unsafe or ineffective vaccine.)

Validating a vaccine isn’t the end of the process. “Getting it rolled out from there will be massively challenging,” Juelg said. “Even if a vaccine is safe and effective, that doesn’t mean people can just walk into their primary-care doctor’s office the next day and get it.” In its letter, the C.D.C. indicated that, even if trial results come quickly and are favorable, only a few million doses of the new vaccines would be available at first—nowhere near enough to create herd immunity or fundamentally alter the economic and social disruptions of the pandemic. Both of the vaccines it described, moreover, must be stored at subzero temperatures—Moderna’s at four degrees below zero, Pfizer’s at ninety-four below—and require two doses, spaced weeks apart, to work. Distributing and administering them will not be easy.

Finally, of course, Americans must be convinced to take the vaccine in numbers high enough to conquer the coronavirus. Fauci told me that he worries about how vaccine science is being communicated to the public. “I never liked the ‘warp speed’ terminology,” he said. “It suggests, incorrectly, that you’re rushing things. Whenever people hear things are being rushed, they worry about safety. They think you’re prematurely putting something out there that isn’t entirely safe.” The speed in Operation Warp Speed, he said, really has to do with the reduction of financial risk: the government has committed to buying millions of doses of vaccines before they are proven. “If a vaccine doesn’t work, you’ve lost a few hundred million dollars,” Fauci said. “If it does work, if it’s safe and effective, you’ve saved four, five, six months of waiting to get people the vaccine. That’s huge.”

Even before the pandemic, American politicians and celebrities fomented vaccine skepticism with debunked safety concerns. In recent years, vaccine hesitancy has grown, and many Americans are now dubious about vaccines with proven track records. (In 2019, because of vaccine hesitancy, the country experienced its largest measles outbreak in decades.) A new covid-19 vaccine won’t have years of safety data behind it. “For a vaccine to be trusted, you need time and you need evidence that people who get it don’t end up having major issues later on,” Peter Bach, a drug-policy expert at Memorial Sloan-Kettering, told me. “Even if we get a vaccine, it wouldn’t surprise me to see many parents making the very rational decision not to vaccinate their children without long-term safety data.” Recent polling suggests that between a third and half of Americans would choose not to get vaccinated against the coronavirus, even if a vaccine were cheap and easily available. The case will be further complicated if a vaccine is only partly effective, or requires multiple doses, or if it’s unclear how long immunity lasts. If only part of the population gets vaccinated, any notion of immediate herd immunity is a pipe dream.

All this suggests that a silver-bullet vaccine may not be in the cards, at least not right away. “Hoping that a vaccine will quickly change the course of this pandemic is, I think, hoping for the least likely outcome,” Bach said. “I wouldn’t bet the future of our society on a covid vaccine.” And yet even a partially effective, partially adopted vaccine, combined with contact tracing and social distancing, could still make an incremental difference. Each person vaccinated would be a person who is less likely to get sick, and so less likely to pass the virus on to others. Depending on the vaccine’s effectiveness, those who’d received it might be able to rest easier. The virus’s momentum would slow, at least a little.

Antiviral drugs fight the virus in a different way from vaccines. A vaccine gets your immune system ready before an infection; it’s like an intelligence report telling your body to get ready for an incoming surprise attack. Antiviral drugs are useful after the invading force has arrived; they fight back by sabotaging it. As Matthew Hutson explained earlier this year, in his survey of antiviral-drug development, an antiviral might work by inserting faulty material into viral genomes, by restraining the enzymes that a virus uses to weave itself into our DNA, or by blocking the proteins that a virus needs to mature. Doctors now see covid-19 as a disease with two phases: in early and mild cases, replication of the virus seems to drive symptoms, but in later and more severe illness it’s our own overexuberant immune response that damages the body. Antivirals generally work best in the first phase, before critical illness has set in. Remdesivir, the only antiviral shown to work against the coronavirus so far, is largely ineffective for patients who are intubated or need high levels of oxygen—they are in the second, immune-focussed stage of the disease—but it can hasten covid-19 recovery by about four days in people with less severe illness, in whom viral replication is still driving symptoms. Researchers are working on other, more powerful antivirals, which could replace remdesivir or work in concert with it.

To understand the difference antiviral therapies could make in the fight against covid-19, you should first understand that, right now, doctors have very few therapies at their disposal. When I first began caring for covid-19 patients, in March, just as the pandemic exploded across New York City, we made great strides in understanding the course of the disease and reorganizing our hospital to accommodate the surge. But the few proposed remedies we tried, such as the drugs hydroxychloroquine and azithromycin, failed. We quickly discovered that the best we could do was support our patients while they battled the virus themselves. We laid people on their bellies to allow their lungs to expand; we tried multiple oxygen-delivery devices; we connected the most desperate to ventilators, hoping that their immune systems could mount the fight we couldn’t. Antiviral drugs would give us a weapon against the virus.

We’ve long used antivirals to treat viruses such as herpes and influenza. But they’re perhaps best known as a treatment for H.I.V. The first H.I.V. antiviral, AZT, was introduced in 1987, six years after aids was identified. In 1995, effective “combination antivirals” arrived; they used separate mechanisms to attack the virus simultaneously, “cornering” it so that it couldn’t mutate its way around a single drug, in a phenomenon known as viral escape. We’re likely to move faster with the coronavirus, both because there are so many more infections, giving researchers more opportunities for study, and because of how rapidly the disease unfolds. It takes nine years, on average, for an H.I.V. infection to progress to aids. The coronavirus leads quickly to acute illness, and so a researcher knows within weeks if a treatment worked.

So far, coronavirus research has focussed mostly on hospitalized patients. But it’s giving antivirals to those who never require hospitalization that could change the course of the pandemic. Treated people will walk around shedding less virus, reducing the chance that they’ll pass it to others. Treatment, if it’s given early enough, becomes a form of prevention.

The concept of treatment as prevention for viral diseases gained momentum in 2011, after a landmark study found that H.I.V. patients who started antiviral medications early were much less likely to pass the virus on to their partners. (The reduction in H.I.V. transmission was so large that the study’s results were released nearly four years ahead of schedule: it became unethical to continue withholding treatment from the control group.) As antiviral drugs became safer, with fewer side effects, the strategy was extended to people who didn’t yet have H.I.V. but were at risk of contracting it. So-called pre-exposure prophylaxis, or PrEP, in which high-risk, H.I.V.-negative individuals take antiviral drugs, is now a major tactic for preventing viral spread. When used correctly, and in concert with condoms, PrEP almost completely eliminates the risk of acquiring H.I.V. Doctors now use the same approach for influenza, prescribing the antiviral drug Tamiflu to nursing-home residents, postpartum women, and people with chronic conditions who are at risk for serious complications from the flu.

Could antivirals be used prophylactically for vulnerable people who haven’t yet been infected with the coronavirus? Remdesivir can be given only intravenously, which limits its applications. But an oral antiviral drug, known as MK-4482—previously known as EIDD-2801—is being developed by Merck and Ridgeback Biotherapeutics. In June, the drug entered Phase II trials; its creators hope that, in infected patients, it will be able to reduce the virus to undetectable levels within days. It’s easy to imagine how such a drug could be given to people who test positive for the coronavirus but who don’t require hospitalization—and even to those with whom they’ve had close contact. There are some preliminary signs that MK-4482, which was originally developed as an influenza treatment, may also work against other coronaviruses, including the ones that cause mers and sars. “At least in vitro, it seems to have pan-coronavirus activity,” Nicholas Kartsonis, who’s leading the medication’s development at Merck, told me. Ideally, it could be used against not just this coronavirus but others. “If the virus mutates again, and there’s a pandemic in the future, we might not have to ride this roller coaster all over again,” Kartsonis said.

In early April, I admitted John, a middle-aged man, to the hospital. In the days before I met him, John had visited two other emergency departments. He’d been feeling unwell—tired, feverish, coughing—but the oxygen in his blood had held steady, and he’d been told to isolate at home and return if the symptoms worsened. Eventually, his breathing grew labored; three days after his arrival, he was on a ventilator in the I.C.U. For weeks before John took a turn for the worse, both he and his doctors knew he had the coronavirus—but, during all that time, they had no medications to give him or the people he might have infected. I can’t say for sure that antivirals would have helped John, who remained debilitated and delirious weeks after coming off the ventilator. But it’s easy to see how antivirals prescribed after his first emergency-department visit—as azithromycin Z-Paks are prescribed to patients with pneumonia—could have halted his morbid progression and made a difference for others.

Like antivirals, antibodies interfere with an invader’s spread inside the body. Unlike antivirals, they are made by our own cells—and so the trick is to transfer them from one body to another. Therapies based on the transfer of antibodies have a long and storied history. In the early eighteen-nineties, two physician-scientists working in Berlin, Emil Adolf von Behring and Kitasato Shibasaburō, took blood from animals that had recovered from diphtheria and found that, transfused, it could prevent or cure the infection in other animals. They worked with guinea pigs, goats, and horses. Soon, though, the therapy was being used to treat people around the world. In 1896, it was hailed by The Lancet as “the most important advance of the Century in the medical treatment of acute infective disease.”

In 1901, von Behring won the first Nobel Prize in Medicine. In the following decades, blood from recovered patients—called “convalescent plasma” or “convalescent serum,” depending on which blood products were isolated—was used to treat pneumonia, meningitis, anthrax, measles, chicken pox, and other diseases. It was also deployed during the 1918 flu pandemic, when researchers found that it led to reductions in mortality. Studies from a century ago have limitations; still, most infectious-disease specialists agree that the treatment probably helped.

The use of serum therapies declined precipitously in the nineteen-forties, as antibiotics became widely available. That was partly because animal serum created substantial complications when it was used to treat people. It often caused fevers, chills, and other allergic reactions; its efficacy might vary from batch to batch; its antibodies had to be matched to specific strains of human pathogens. (There might be several varieties of a given virus or bacterium in circulation at any one time.) Antibiotics were equal-opportunity destroyers—safe, reliable, and easy to produce. In the nineteen-nineties, interest in convalescent plasma rebounded, because many previously tamed pathogens had developed resistance to antibiotics; by then, human blood was also being used, and the process was much safer. In patients sickened by sars, H1N1 influenza, and Ebola, convalescent plasma showed potential for lowering virus levels and, in some cases, death rates. But evidence from gold-standard randomized trials is still lacking. By the time well-controlled studies of the therapy had got under way, the viruses had moved on to other locations, which meant that new patients had to be enrolled.

In August, the F.D.A. granted an emergency-use authorization for convalescent plasma as a therapy for covid-19. Its decision was based on an unpublished study by researchers at the Mayo Clinic; the study lacked a control group and has not been peer-reviewed, but it found that patients who received convalescent plasma within three days of being hospitalized were somewhat less likely to die than those who got the treatment later in their stay. In a press conference, Stephen Hahn, the commissioner of the F.D.A., exaggerated the proven benefits of the treatment based on those preliminary findings, drawing widespread criticism from scientists and policymakers. (He later apologized.) Meanwhile, the panel of experts assembled by the N.I.H. to develop covid-19 treatment guidelines issued a statement arguing that there is currently “insufficient data to recommend either for or against the use of convalescent plasma for the treatment of COVID-19”; the panel concluded that “well-controlled, adequately powered randomized trials” are needed.

More than seventy thousand Americans have already received convalescent-plasma therapy for covid-19. Bach, the drug-policy expert, is appalled that the United States has not completed a single randomized controlled trial to judge its efficacy. “It’s an abomination,” he said. “I can’t believe none of these world-class institutions have run a proper RCT to get definitive evidence of effectiveness. Tens of thousands of Americans have received it, and we still don’t really know if it works.” Liise-anne Pirofski, the chief of the division of infectious diseases at Albert Einstein College of Medicine and Montefiore Medical Center, in the Bronx, is trying to fix that problem. Pirofski has studied antibody-based immunity for nearly three decades, and is now leading a large, multi-hospital, randomized trial studying the use of convalescent plasma. Over the years, she has often found herself advocating a snubbed therapy. Now the lack of effective treatment for covid-19 has created an opportunity. “Many of us are very excited right now,” Pirofski told me. “This could be the moment people realize that antibody-based therapies have an important role in the treatment of infectious disease.”

And yet questions persist about how useful, reliable, and widely available a blood-based treatment can be. The antibodies produced by the immune system are of variable quality. The plasma of a covid-19 patient might contain thousands of different kinds of antibodies against sars-CoV-2, only a portion of which are effective. (Some, for instance, might successfully target the virus’s crownlike spike proteins, which allow entry into human cells; others could attach to less important viral structures.) Moreover, in an evolutionary game of cat and mouse, viruses have developed decoy antigens specially designed to lure host immune systems into developing ineffective antibodies.

Pharmaceutical companies, therefore, are now trying to isolate the most effective sars-CoV-2 antibodies, so they can develop powerful antibody drugs, known as “monoclonal antibodies,” that can be manufactured at scale. These antibodies are grown in vats of cells, then separated out so that they can be given to people. In July, Regeneron, a biotechnology company based in Tarrytown, New York, announced that it would move its antibody drug, REGN-COV2, into Phase III trials; it is hoping to enroll two thousand patients across a hundred sites. “We’re trying to isolate the best antibodies to fight covid-19 and produce them in large amounts,” Christos Kyratsous, who is leading the drug’s development, told me. REGN-COV2 contains a combination of two antibodies thought to be particularly effective for neutralizing the virus; a combination approach could also prevent viral escape. “Viruses can mutate very quickly, but, given what we know about sars-CoV-2, I think it’s unlikely it will evade two antibodies at the same time,” he explained. Regeneron is testing the drug’s efficacy in different contexts: before people are exposed, after they’ve been exposed, and after they’ve contracted the virus.

To date, the F.D.A. has approved just one virus-targeting antibody drug: a medication for the respiratory syncytial virus (R.S.V.), which causes the common cold in healthy children and adults but can lead to devastating illness in preterm infants. A second medication—an Ebola-antibody drug that Kyratsous helped to develop—is currently under review. (Experience with the Ebola drug, he said, helped Regeneron move fast with the coronavirus.) Regeneron recently received a four-hundred-and-fifty-million-dollar contract from the federal government to start manufacturing REGN-COV2 while clinical trials are ongoing. If the drug works, it will be available to Americans immediately. (Because antibodies can’t be packaged into a pill—they’re easily digested by stomach acid—they must be given as an I.V. infusion, or as an injection into the muscle or under the skin.)

At the height of the city’s pandemic, I cared for a woman who worked as an I.C.U. nurse across town. She had started coughing many days earlier, and, as her condition worsened, her husband cared for her. I phoned him each afternoon—our hospital allowed no visitors—until, one day, he landed in the bed next to hers. Their paths diverged: when she was ready to go home, he was on his way to the I.C.U. It’s easy to picture how, in their case, antibodies and antivirals might have been used together. As a high-risk essential worker, she might have been given prophylactic antivirals to reduce her chance of getting the coronavirus; after she’d been infected, the drugs could have helped her beat the virus and made her less likely to pass it on. If she’d still developed symptoms, her husband could have begun antivirals, and both spouses could have benefitted from antibody therapy, which might have prevented their illnesses from progressing. (In the end, he joined her back at home—but after days in the I.C.U. and weeks in a recovery unit.) At every step, the virus would have been surrounded by hostile counterattacks. “It’s very possible that this disease will require more than one type of intervention,” Pirofski said. “The combination of an antiviral and an antibody could be superior to either alone.”

For physicians, there are few things as distressing as watching a patient’s slow, mortal decline with few remedies to offer. This spring, I cared for a woman I’ll call Ruth. She was advanced in years but youthful in spirit. When I met her, I wondered whether she might be better off recovering at home: she had mild symptoms and few preëxisting medical conditions. When I asked her if she felt up to walking, she told me that she’d rather dance. By the next morning, though, her demeanor had changed. She spoke in short sentences, pausing to catch her breath. Each day, her breathing deteriorated, and the oxygen levels in her blood declined. As her condition worsened, the level of inflammation in her body surged. She became delirious, and her organs started to shut down. I called her son, who told me that they’d had the conversation long ago: she didn’t want machines keeping her alive. He said goodbye through tears, over FaceTime, as his mother, eyes closed, took shallow breaths from an oxygen mask.

Autopsies of patients who have died of covid-19 reveal that the sars-CoV-2 virus infiltrates many vital organs. But the primary battlefield is the lungs—more precisely, the intricate network of air sacs, or alveoli, that allows humans to exchange carbon dioxide and oxygen. As the disease progresses, the virus saturates the lungs, and the body, in a frantic effort to respond, drums up an immunologic storm, unleashing a flood of immune cells that fights the virus but also causes collateral damage. This is when the second phase of the disease begins. Alveoli are exceedingly delicate, and during the immune system’s onslaught their membranes are compromised; fluid leaks into the air spaces, dead cells and proteins create blockages, and maintenance molecules cease to function. Many alveoli collapse entirely, closing off the body’s microscopic conduits to the outside world. In this second, later phase, a drug that shuts down viral replication might no longer help. “Early on, you want the immune system to be working hard,” Fauci told me. “You want to help it fight the virus. But, in advanced disease, it’s less about what the virus is doing and more about the immune system spiralling out of control.”

Immunomodulating drugs don’t stop the virus from replicating; instead, they try to restrain this devastating hyperinflammatory cascade. There’s mounting evidence that these medications could work for patients in the second phase of covid-19. The most widely cited example is dexamethasone, an inexpensive, generic steroid that quickly and powerfully suppresses multiple arms of the immune system. Early in the pandemic, it wasn’t clear whether steroids were helpful or harmful, but, in June, a major study from the United Kingdom found that dexamethasone can have large beneficial effects for severely ill covid-19 patients: it reduced by one-third the chance that intubated patients would die, and, for people on oxygen but not mechanical ventilation, it lowered the risk of death by a fifth. In American hospitals, dexamethasone is now routinely given to patients with severe covid-19; the drug may be partly responsible for our declining death rates. The search is on for more, better immunomodulators.

Michael Mansour, an infectious-disease doctor and an assistant professor at Harvard, is one of many researchers working to unravel the complex relationship between the coronavirus and the immune system. “We’re trying to figure out when someone crosses over from the viral phase to the immune-dysregulation phase,” Mansour said. “We need to thread the needle, to find the right balance.” Patients with covid-19 have dozens of immune-signalling molecules, called cytokines, surging through their blood at various times during the illness; cytokines serve as alert signs, telling the body where to mount its immune response. Which cytokines are provoking destructive reactions? Can they be targeted? “Steroids are really the nuclear option,” Mansour said. “They shut down pretty much everything. We’re trying to see if we can be more selective in how we curb the immune system.”

Mansour’s team is exploring whether inhibiting one particularly rambunctious cytokine, interleukin-6, might prevent complications later in the course of covid-19. IL-6 is known to set off a robust inflammatory cascade, and is implicated in a number of autoimmune conditions, including vasculitis and rheumatoid arthritis. A small study of twenty-one patients in China found that those who received a drug called tocilizumab, which is an IL-6 inhibitor, experienced marked reductions in inflammation and the need for oxygen. Mansour’s team enrolled hundreds of patients across seven Boston-area hospitals in a study, randomly assigning participants to tocilizumab or a placebo. Mansour expects to have results later this year, but, even if tocilizumab doesn’t help—two recent studies of IL-6 inhibitors have returned negative results—he believes that tinkering with the immune system will ultimately be a key part of our covid-19 strategy.

There are plenty of ways to modulate the immune system: different drugs target different parts of its incredibly complex machinery. “With immunomodulators, we’re in a land of riches,” Mansour said. “There have been so many agents developed over the years.” Many are used to treat autoimmune diseases, such as psoriasis, multiple sclerosis, or Crohn’s disease, and carry names that end with “mab”—short for monoclonal antibody. (Unlike virus-neutralizing antibody drugs, such as the one being produced by Regeneron, these drugs target parts of the immune system—not the virus itself.) There’s infliximab and adalimumab; tocilizumab and trastuzumab; ustekinumab and canakinumab. “There’s tons of stuff on the shelf here,” Mansour said. “I don’t think we’ll have to develop new covid-19 immunomodulators. We already have drugs that are safe and targeted. It’s a matter of testing different classes to see what works best.”

Patients like Ruth seem to cross a border in their experience with covid-19: once they’re on the other side, their bodies begin to work against them. Too often, in this latter phase, when the illness is less about viral replication and more about human immune systems, we’re powerless. Immunomodulators may help us bring both the body and the disease to heel.

How will we know when these new interventions are beginning to make a decisive difference? Epidemiologists quantify a disease’s lethality by calculating its “case-fatality rate”—a number that expresses the percentage of people who, having got the disease, go on to die from it. The number is crucial for determining how concerned we should be about a disease’s spread, and how aggressively we should move to stop it. The case-fatality rate for the coronavirus has been hard to pin down, in part because it’s never quite clear how many people have been infected. (Recent data from the C.D.C. suggest that there may be many times more infections than have been confirmed, which would make the virus more contagious and less lethal than we think.) And yet, whatever the rate, the dozens of therapies in development will almost certainly decrease it. Lower it enough and the world-stopping phase of the pandemic could come to an end.

Not long ago, I found myself talking with Rajesh Gandhi, an infectious-disease doctor at Massachusetts General Hospital. Gandhi is the incoming chair of the H.I.V. Medicine Association, and has treated patients with H.I.V. for decades. (In the past, many of them also developed active tuberculosis, because of their weakened immune systems.) He now serves as an expert on the N.I.H. panel developing covid-19 treatment guidelines. “At the beginning, there was nothing we could do,” he recalled, of the aids crisis. “We watched as people slowly died.” aids is a disease that’s been largely tamed using the TB model. There is no silver-bullet vaccine for H.I.V.; the virus still circulates around the world. But, in a relatively short span of time, in the mid-nineteen-nineties, a number of different therapies came together to make a pivotal difference in its lethality. Combined with public-health measures, those therapies have changed the world’s relationship to that virus. It is still a threat—but it’s a manageable one.

During this pandemic, I have often felt at a loss. At patients’ bedsides, I’ve experienced uncertainty and helplessness; the fear of getting sick myself, or of infecting my family, has followed me home from the hospital each day. In the evenings, when I’ve had the time and energy, I’ve turned to books and essays written by doctors during the first years of the aids epidemic. The challenges that they describe have felt familiar, and their accounts have given me hope. One step at a time, medical and scientific advances eventually subdued that deadly virus. They could do the same for the coronavirus. “With H.I.V., it was never just one thing,” Gandhi told me. “It was a lot of little things that added up to a big thing.”