On a Wednesday morning in late June, Junior Mhlongo arrived at the Chris Hani Baragwanath hospital in Soweto, the predominantly Black township outside Johannesburg, South Africa. Sitting inside an examination room, he unzipped the purple jacket he wore to guard against winter in the southern hemisphere. But he kept on the red cloth mask he wore to guard against the spread of COVID-19.

Mhlongo slipped his left arm from his coat sleeve and hiked up the cuff of his shirt. He winced as a nurse stuck a needle into his arm and injected him with a syringe of clear liquid. “I feel a little bit scared, but I want to know what is going on with this vaccine, so that I can tell my friends and others,” he told the reporters who had come to watch.

It’s not just Mhlongo: The whole world is awaiting an answer. The young volunteer is among the first of some 2,000 patients in South Africa—and thousands more in the U.K., the U.S., and Brazil—participating in clinical trials for a vaccine for COVID-19, pioneered by scientists at the University of Oxford in England.

A vaccine is the key to ending the long, global nightmare of the coronavirus pandemic. It can unlock economies and save countless lives. Without it, millions more could die, and business and government will remain hostage to the virus.

As of late July, there were at least 168 different vaccine candidates in some stage of development, according to the World Health Organization. But the Oxford vaccine is arguably the furthest along in human testing. Preliminary human trials have been promising. In anticipation of success, the U.S. government has preordered 300 million doses at a cost of $1 billion. The U.K. has ordered 100 million; the European Union, 400 million. Deals have been struck to provide the developing world with more than 1.3 billion doses.

Still, vaccine development is always a gamble—a 2018 MIT study found that 66% of vaccine candidates for infectious diseases fail. And no one has staked more on the Oxford vaccine than Pascal Soriot, chief executive officer of AstraZeneca, the British pharma giant. Soriot scored a coup in late April when AstraZeneca swept in to become Oxford’s commercial partner on its COVID-19 vaccine, displacing rivals with better track records in vaccine development to seize the lead in this global race.

The Oxford partnership has drawn unprecedented attention to AstraZeneca. But it’s only the latest in a series of bold collaborations, strategic shifts, and calculated risks at the company over the past eight years—moves that have elevated AstraZeneca from one of Big Pharma’s least successful drug developers to its most dependable. Soriot is the engineer of that metamorphosis. Born in France, the 61-year-old pharma lifer has spent most of his career abroad but still speaks in French-inflected English. He has jet-black hair that contrasts with the crisp white dress shirts he favors, and he speaks with a directness rare in executives of any nationality. Soriot tells Fortune he believes in “casual intensity.” “We have to take what we do seriously, but we should not take ourselves too seriously,” he says.

Of course, few pharma CEOs have faced a challenge as existentially serious as COVID-19. The Oxford partnership is risky, requiring AstraZeneca to commit hundreds of staffers to expanded clinical trials and manufacturing—long before the vaccine is even proven. Whether the vaccine is truly effective won’t be known until sometime this fall. But if clinical trials show a positive result, U.S. and U.K. regulators are expected to approve the vaccine on an emergency basis, and Soriot has promised to have doses ready as soon as late September so mass vaccination programs can begin.

On average, developing a new vaccine takes more than a decade. In this case, it is being done in about six months. Even if it works, AstraZeneca, which brought in $24.4 billion in revenue last year, will make no money at first. It has agreed to provide most of those initial 2 billion doses, and perhaps billions more, at cost. The vaccine will boost AstraZeneca’s bottom line only if COVID-19 proves to be an endemic, seasonal menace, like the flu, for which people require regular vaccinations. Still, AstraZeneca (or AZ, as it is informally known) has seen a stock market bump. On the day the Oxford deal was announced, its shares soared to an all-time peak. They have since traded as high as 97 pounds ($124), up 38% from their pre-pandemic price, and AZ now commands the highest market capitalization on London’s FTSE 100 index.

How AstraZeneca got the inside track on the Oxford vaccine is a story, like all business stories, about relationships. But it’s also about COVID-19’s fraught geopolitics. And it’s a tale of a remarkable corporate turnaround—one that holds lessons for any business trying to profit at science’s cutting edge, and for any executive trying to lead through a radical cultural transformation.

****

When news of a mysterious respiratory illness in Wuhan, China, surfaced in December, AstraZeneca was quick to take notice. More than any other Western pharmaceutical firm, AZ has bet big on China. Almost a quarter of its 61,500 people are based there, and last year its Chinese revenues were some $4.8 billion. Beginning in late December, Soriot held almost daily video calls with AstraZeneca’s Chinese executives, tracking the outbreak as it inched toward becoming a pandemic.

“Then we quickly thought, ‘What can we do to help?’ ” Soriot recalls. AstraZeneca bought 9 million FFP3 respirators, which can filter out the virus, and donated them to health care workers worldwide. It began investigating whether any of its existing drugs—including a leukemia treatment that fights inflammation and a diabetes drug that can protect the heart and kidneys—could be repurposed to battle COVID-19. It is collaborating with Vanderbilt University on synthetic antibodies that might treat the disease. And after it became apparent that the U.K. lacked adequate COVID-19 testing capacity, AstraZeneca teamed up with its British rival, GlaxoSmithKline (GSK), and the University of Cambridge to set up a new test-processing lab.

One thing was conspicuously absent from AstraZeneca’s response: a vaccine. AZ is not a significant player in that field. It currently sells just one vaccine, a nasal spray that protects against seasonal influenza. But events taking place not far from its Cambridge, England, headquarters, in a rival college town, were about to change that.

Amid a cluster of research labs in Oxford, the Jenner Institute is housed in a low-slung modern office building with smoky glass windows and green cladding. Named for Edward Jenner, the 18th-century physician who first inoculated people against smallpox, it has emerged in the past two decades as one of the world’s foremost centers for vaccine development.

Oxford was able to leap out in front in the sprint for a coronavirus vaccine because of similarities between SARS-CoV-2, the coronavirus that causes COVID-19, and those that cause severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS). Sarah Gilbert, a Jenner researcher, had been working on a MERS vaccine, and she realized almost immediately it might be possible to use the same method for SARS-CoV-2.

The majority of vaccines are made using a weakened version of the virus they target. Gilbert’s vaccine is different: It takes a harmless chimpanzee virus and genetically modifies it to produce the surface spike protein found on SARS-CoV-2, allowing the body to develop a potentially effective immune response. Vaccines produced this way have at least one major advantage: They don’t have to be kept at subfreezing temperatures, which is especially important in developing countries that lack reliable cold transport and storage.

The Oxford scientists knew from previous work that the modified chimpanzee virus was safe. And they knew it could be manufactured in large quantities. But John Bell, the professor who helps oversee the Oxford medical-science division’s work with external partners, says the university also knew it would need help from Big Pharma. “Making billions of doses of vaccine for the whole world, let’s be clear, that’s not what universities could or should be doing,” says Bell, an athletic 68-year-old Canadian who habitually keeps his eyeglasses propped atop his silver-gray hair.

Oxford set out to find a partner with proven vaccine expertise—but politics made negotiations difficult. Oxford balked at some U.S. companies’ insistence on exclusive worldwide manufacturing rights. The Trump administration was prepared to offer large sums to guarantee the American public was first in line for a vaccine. Various governments, including those of the U.S., France, and Germany, vowed to block exports of vaccines made in their countries. Ultimately, Bell says, the British government, which steers millions of pounds in funding to Oxford each year, grew anxious that it could wind up without adequate vaccine access, and in April it told Oxford to find a British partner.

There were only two U.K. companies with sufficient manufacturing capability: GSK and AstraZeneca. GSK might seem like the obvious choice. It is the world’s top vaccine producer, with a portfolio of more than 30, including inoculations for measles, meningitis, and pneumonia. But GSK was already working on its own COVID-19 vaccine efforts, including one with French drug giant Sanofi and another with Chinese biotech firm Clover Biopharmaceuticals. (The project with Clover is now in early-stage human testing.)

AstraZeneca had virtually no vaccine track record. But it had experience with a manufacturing process similar to the one Gilbert uses for her modified chimp virus. (AZ uses it to make synthetic antibodies.) What’s more, Bell had a personal connection to Soriot: Until last year, Bell served on the board of Swiss drugmaker Roche, where Soriot had been a fast-rising star, serving as chief operating officer for two years before taking the top job at AZ. “I knew Pascal was somebody who was absolutely committed to the highest-quality science, but also prepared to take risks,” Bell says. “And, to be crystal clear, this vaccine is a big risk.”

Oxford also had another important factor in mind: AstraZeneca’s remarkable R&D rebound under Soriot’s leadership. “It’s one of the great turnaround stories in pharmaceutical companies over the past 20 years,” Bell says. Michael Leuchten, an equity analyst for Swiss bank UBS, says of AstraZeneca’s recovery: “It’s remarkable. I’ve never seen anything like that before, and I don’t think we will again anytime soon.”

****

When Soriot took the helm of AstraZeneca in 2012, the company faced a bleak future. Its bestselling medicines—Crestor for high cholesterol, Nexium for acid reflux, and antipsychotic drug Seroquel IR—were about to plunge off the industry’s dreaded “patent cliff.” As soon as they lost intellectual-property protection, generic drug manufacturers would be able to sell cheap copies. Half of AZ’s annual sales—$17 billion in total—were likely to disappear in the next five years.

Worse, the company, having underinvested in R&D, had no pipeline to replace its blockbusters. “At the time, the company was run like a spreadsheet,” Soriot says. “It was all about cost savings, increase the profit and use that for a buyback, and mechanically increase the share price. But the company was on the road to nowhere.” Many investors agreed. “He has to rethink the strategy,” Stephanie Maher, a fund manager at asset management firm AGF, said when Soriot was appointed.

The minute Soriot took over, he stopped the buybacks and plowed money back into science. AZ’s R&D spending as a percentage of sales was 16% in 2011, the year before Soriot was appointed; in 2019, it was 26%, the highest percentage of any of its peers. And reviving the drug pipeline wasn’t just a case of throwing money at a problem: It involved refocusing the entire company.

A decade ago, AZ’s R&D operation was extremely inefficient: Just 4% of its drug candidates made it to market. “We were spending $5 billion a year on R&D and not delivering any medicine,” says Mene Pangalos, who heads AZ’s biopharmaceuticals R&D. A damning 2012 analysis from the InnoThink Center for Research in Biomedical Innovation, a consultancy, found that AZ was spending, on average, $11.8 billion in R&D for every drug that received U.S. Food and Drug Administration approval, the worst record in the industry.

Soriot’s predecessor, David Brennan, hired Pangalos to revive the research division. After Soriot took over, he accelerated the effort to restock the medicine cabinet. In one early leadership meeting, a sales executive said a researcher’s presentation “had been a bit technical and over my head.” Soriot publicly upbraided him. “If you want to be a senior leader in our organization, you need to be interested and engaged in the science that we do and the patients that we are treating,” Pangalos recalls Soriot saying. “There was stunned silence and quite a few gulps. But if you were an R&D person in the room, it was the best we’d ever felt.”

The fledgling overhaul was almost smothered in its nest. In the spring of 2014, the larger U.S. drugmaker Pfizer launched a hostile takeover attempt. Investment bankers initially told Soriot the company had a less than 10% chance of remaining independent, the CEO recalls. Soriot repelled the effort with hefty promises to investors—pledging to sustain dividend growth and boost revenue to a lofty $40 billion, a 75% increase, by 2023. In retrospect, Soriot says, the takeover attempt galvanized his strategy, making it crystal clear to the entire company that its survival depended on more effective, and thus more profitable, R&D.

AstraZeneca had been focused on primary-care drugs, medications for chronic conditions such as allergies and high blood pressure. Soriot shifted its emphasis to specialty care—drugs that command high reimbursements from insurers and better profit margins, even though their market size is smaller.

Nowhere has that change in priorities been more dramatic than in oncology. Under Soriot’s predecessor, AstraZeneca had been divesting from cancer treatments. But early on, leading oncologists outside the company told the CEO that AZ was about to write off a potential blockbuster in Lynparza, a drug for late-stage ovarian cancer that had shown promise in clinical trials. Soriot reversed course and revived the drug.

It was a momentous decision. Lynparza emerged as a pioneer in a class of drugs called PARP inhibitors. These drugs target enzymes critical to cell repair; by blocking them, they keep cancer cells from rejuvenating. It soon emerged that Lynparza could treat some pancreatic, breast, and prostate cancers, too. It’s now the bestselling PARP inhibitor: In 2019, it generated $1.2 billion in revenue, with sales growing 85% year over year. Its versatility also makes it attractive to corporate partners. In 2017, Merck agreed to pay AZ up to $8.5 billion to share in Lynparza’s development.

Lynparza is the flagship example of AstraZeneca’s oncology strategy, which has yielded highly specialized medications that merit premium prices. Tagrisso, which targets a specific, highly dangerous mutation of lung cancers, has become AZ’s brightest star: It brought in $3.2 billion last year. Imfinzi, an immunotherapy treatment for lung, urinary, and bladder cancers, made $1.5 billion in 2019. (AZ has also struck partnerships for oncology drugs, including two multibillion-dollar deals with Japan’s Daiichi Sankyo in as many years. In July, it agreed to pay as much as $6 billion for the right to codevelop a Daiichi drug that could limit chemotherapy’s side effects.) Oncology now accounts for 37% of AZ’s annual product sales, more than any other area, and up from just 12% five years ago.

In addition to oncology, Soriot reorganized AstraZeneca around drugs for cardiovascular, renal, and metabolic disease, respiratory illness, and immunology. Across all those categories, Soriot and Pangalos instituted a rigorous process for assessing drug candidates. Known as “the five R framework,” it mandates that before the company commits to developing a drug it must be satisfied it has identified five “rights”: the right target, the right patient group, the right body tissue, the right safety regime, and, crucially, the right commercial opportunity. The checklist sounds like Pharma 101, but it’s the kind of discipline AstraZeneca had once lacked.

Disciplined deployment of the framework has drastically reduced the number of compounds AstraZeneca had under development. But it has lifted its success rate in moving promising molecules from preclinical investigation through to completion of late-stage clinical trials. Once a paltry 4%, that rate is now 20%, three times the industry average.

AZ has also used new digital technologies to reinvent its approach to clinical trials. Software called Merlin enables AZ to select trial sites 70% faster than before, according to Cristina Duran, chief digital health officer for the R&D division. Merlin produces trial cost estimates in minutes (it used to take days). In a research ecosystem where trial volunteers often skew disproportionately male and white, Merlin helps select trial populations that better reflect real-world demographics. Another software system, called Control Tower, allows managers to get a visual snapshot of all AZ’s trials on a single dashboard, helping them predict problems in patient recruitment. “We’ve taken systems that the company was using for 20 years, and in the past two years we changed them—which is either insane or brave, but it has been successful,” Duran says.

To cement its science-led ethos, in 2016 Soriot moved both AZ’s corporate headquarters, which had been in London, and its research headquarters, which had been in Macclesfield, to Cambridge. Here Soriot is overseeing construction of a signature, $1.2 billion strategic R&D center, a round glass doughnut to showcase the work of AZ’s scientists. The building is steps from the government molecular biology lab known as “the Nobel factory”—its scientists have won 12 of them—as well as top-flight cancer and stem cell research labs. Since moving, AZ has increased its collaborations with Cambridge University to more than 130, up from fewer than 10 five years ago.

AstraZeneca is a smaller company today than it was before its reinvention, thanks largely to the loss of sales from those older blockbusters that fell off the patent cliff. AZ’s revenues last year were 27% lower than in 2011. But investors seem to think the company has made up in drug quality for what it lacks in sales quantity. AZ’s stock price has risen 137% in dollar terms since Soriot took over, trouncing the 61% increase in the S&P Pharmaceuticals Select Industry Index over the same period.

****

AZ’s deal to produce Oxford’s vaccine was brokered in a flurry of Zoom calls during the last week of April. Oxford had two main conditions: that AstraZeneca forgo any profit on the vaccine until the pandemic is over, and that it commit to making the vaccine as widely available as possible. AZ agreed to both.

For AstraZeneca, the vaccine is a showcase project where it can demonstrate its ability to be a partner for academic groups and biotech companies at the forefront of science. That’s worth more than bragging rights: Those relationships are critical to feeding AZ’s pipeline of winning medicines. The speed of the vaccine project, meanwhile, is a key test of the systems AstraZeneca has built to run clinical trials more efficiently.

To condense testing that would normally take years into months, Oxford took unprecedented steps: Rather than recruiting just a few dozen patients for Phase I trials, which are designed to show a drug is safe by testing it on a small cohort, Oxford recruited 1,100 and accelerated the pace of inoculations. Phase II trials, which test for safety and efficacy on a much larger number of volunteers, and Phase III trials, which aim to prove that a drug is effective, would normally take place sequentially. Instead, they are happening simultaneously. Oxford is recruiting 10,000 people in the U.K. for this testing, plus 5,000 in Brazil and 2,000 in South Africa, while AZ is enrolling 30,000 volunteers in the U.S.

Results are expected anytime between September and November, Soriot says. But the pandemic’s urgency means large-scale vaccine production must start even before testing concludes. With funding from governments and international agencies, AZ has licensed manufacturing—again, at no profit—to firms across the globe, including a landmark deal with the Serum Institute of India, a private biotechnology company, to provide 1 billion doses for low- and middle-income countries.

****

In early June, AstraZeneca’s soaring stock was momentarily rattled. Bloomberg News, citing anonymous sources, reported that AZ had approached the U.S. pharmaceutical company Gilead about a supernova-scale merger—one that would be the largest in the sector’s history. The pandemic has made Gilead a household name, thanks to its production of remdesivir, one of the few drugs that has been shown to reduce hospitalization time for severely ill COVID-19 patients.

Gilead declined to comment on the report, as did Soriot. But, speaking in general terms to Fortune, he implied that initiating merger talks would make little sense in the current climate. “I would just invite you and everybody else to look at the facts,” he says. Figuring out how to execute clinical trials under pandemic conditions is an all-consuming undertaking, he says. So too is working on the vaccine. “It is energizing everybody, but it is a big job for a number of us.” Plus, given travel restrictions and quarantine measures, negotiating teams would be unable to meet face-to-face. “Imagine the biggest merger in the history of the industry, done by Zoom? I mean, come on,” he says.

Industry and stock market analysts, on the whole, discounted the possibility of a tie-up. Except, some noted, there was a possible scenario where the merger made sense—and it was a scenario that turned investors’ blood cold: What if AstraZeneca had to do a deal because it was running out of cash?

Gilead is sitting on a cash hoard of $24 billion. AZ, for all its recent success and highly valued stock, is cash poor. Last year the company didn’t generate enough net cash from operations to cover its $3.5 billion in dividend payouts—the ones Soriot promised to keep growing during his fight with Pfizer. In fact, the company had to raise $3.5 billion through an additional share offering, its first in 20 years, in order to meet that and other obligations.

AZ’s cash issues stem in part from its reinvention push. The company has taken restructuring charges to discontinue legacy projects, made lump-sum upfront payments to new drug-development partners, and taken on debt to fund R&D—all strategies that hurt cash flow today. Naresh Chouhan, an analyst at London’s Intron Health, an equity research boutique focused on health care, says the company has used “a lot of creative accounting”—legal but aggressive tactics, such as amortizing those lump-sum payments over the life of the partnership—to keep these transactions from depressing the non-GAAP “core earnings per share” figure AstraZeneca tells investors to pay attention to.

That high-wire act has left some investors wary. UBS’s Leuchten admires AZ’s turnaround, but he is also one of three analysts (out of 24 that cover AZ) to have a sell rating on the shares. Leuchten calls the company “a concept stock.” He says that investors have convinced themselves that AZ has a “platform” for churning out a succession of blockbusters, but that so far, its performance doesn’t justify its outsize valuation, currently 106 times its trailing earnings per share.

Soriot accuses critics like Leuchten of “looking at their shoes instead of looking at the horizon.” He says, “The question people have to decide about is, ‘Do I believe in the future of this company?’ ” He defends AZ’s dividends, saying they reward shareholders for their patience. Cash flow will eventually improve, he says. “And soon enough the people who say, ‘You should cut the dividend’ will move to another criticism: They will say, ‘Why won’t you increase the dividend?’ ”

****

For now, any debate about AstraZeneca’s long-term financial health is superseded by a single question: Will the Oxford vaccine work?

On July 20, the first glimmer of an answer emerged when medical journal The Lancet published the results of Oxford’s Phase I trials. The headlines were positive. (“First human trials of Oxford coronavirus vaccine show promise,” Reuters declared.) Yet AstraZeneca’s stock price fell 5.9% on the news, its largest one-day drop since March, and subsequently drifted lower still.

Phase I trials are designed to prove a vaccine doesn’t threaten users’ health, but they also offer clues to its effectiveness. The AZ-Oxford vaccine appeared to succeed on both counts. The drug showed no serious side effects. Of the 25 volunteers vaccinated with a single dose whose blood was analyzed in detail, 91% produced antibodies capable of neutralizing the virus. In 10 volunteers who also received a second booster dose, 100% produced antibodies—and in those subjects, antibody levels equaled those found in patients who recover from COVID-19.

So what unnerved the markets? In part, it was that booster-dose group. The disparity between their results and the first group’s may indicate that Oxford’s vaccine will require two doses. As Pangalos pointed out at the time, most competing vaccine efforts are also looking at two-dose protocols. Still, the need for multiple shots would make any global vaccination program more fiendishly complicated and expensive. If another vaccine could confer immunity with a single dose, it could become the preferred option, negating AZ’s first-mover advantage.

The Oxford candidate could also lose out to a rival two-dose vaccine. Analysts noted after the Lancet report that initial trials of some competing vaccines—notably, one that Pfizer is developing with biotechnology firm BioNTech (see box)—indicated higher levels of antibody production than Oxford’s. Another question hovering over the effort: Whether it will confer “sterilizing immunity,” which would keep inoculated people from spreading the virus to the unvaccinated.

For those reasons and more, Oxford’s Bell has put his vaccine’s odds of success at no better than 50/50. “This is not a slam dunk, at all,” he says.

Still, the Phase I study provided reason to believe AZ and Oxford could score. The antibody counts were encouragingly high. And COVID-19 antibodies may not be the only key to any eventual vaccine’s effectiveness (indeed, an increasing number of studies indicate that those antibodies may fade rapidly). AstraZeneca and Oxford were keen to note that their vaccine also prompted a strong response from T cells, which seek out and destroy pathogens. Thanks in part to these cellular warriors, Pangalos told reporters, he is “increasingly confident” Oxford’s vaccine will provide immunity for at least a year.

As for its strength relative to other vaccines, Oxford and AstraZeneca researchers note that, so far, no two of the trials have used an identical blood sample assay, making apples-to-apples comparisons impossible. What’s more, Soriot observes, there’s room for more than one winner: Safely inoculating the entire world will certainly require multiple vaccines.

And as it turns out, Oxford is not AZ’s only ticket in the COVID-19 vaccine sweepstakes. Shortly after Soriot became CEO, AstraZeneca invested $240 million in a biotechnology startup in Cambridge, Mass., called Moderna; AZ later invested a further $140 million, upping its stake to 9%. Moderna is developing a vaccine based on messenger RNA (mRNA), which carry instructions from a cell’s nucleus to parts of the cell that manufacture proteins. Modify mRNA and you can get the cell to produce whatever proteins you want—including ones that (hopefully) elicit an immune response. Moderna’s vaccine is now entering large-scale human testing, and its stock has soared on the news—inflating the value of AstraZeneca’s stake to more than $3 billion.

With a bit of luck, both vaccines might succeed. Now all eyes are on those later-stage trials. The testing needs to happen in places where coronavirus infections are rampant, so that scientists can make a comparison between inoculated individuals and a control group. That’s why Oxford researchers are racing to hotspots where cases are still surging—to Junior Mhlongo’s hospital in South Africa, as well as to Brazil and India and, yes, to the still-stricken U.S. Oxford is also contemplating a controversial “challenge trial”—where healthy young volunteers will be inoculated and then purposefully infected with the coronavirus to see if the vaccine works.

If both vaccines fail, AstraZeneca may not be materially worse off. But the world will be that much further away from obtaining the get-out-of-lockdown card it so desperately needs.

****

Leaders of the vaccine pack

At least 168 COVID-19 vaccine candidates are now in development—a fact that hopefully increases the likelihood that one or more will succeed. Here are six that look particularly promising. All are in Phase II or Phase III trials, which means they did well in early studies and are now being tested in larger groups for safety and efficacy.

Oxford/AstraZeneca

This vaccine uses a modified chimpanzee virus to produce a protein found on the surface of SARS-CoV-2, the virus that causes COVID-19. That antigen prompts the body to develop an immune response. In early trials, it increased production of antibodies and T cells; Phase II and III trials are underway.

Sinovac

This Chinese biopharma company bases its vaccine on an inactivated version of SARS-CoV-2 (the approach used in the majority of vaccines). The drug has appeared safe and produced antibodies in two rounds of trials. Sinovac recently launched Phase III studies in Brazil, where the virus has spread rapidly.

Pfizer/BioNTech

U. S. pharma giant Pfizer and Germany’s BioNTech are developing a vaccine that relies on messenger RNA (mRNA), bits of genetic code that instruct cells to make proteins. The goal: Modify mRNA so it prompts cells to make proteins that look like COVID-19, eliciting an immune response. Early trial results led the U.S. government to put up $1.95 billion to secure doses if the vaccine works.

Moderna

The Cambridge, Mass., startup also uses a vaccine based on mRNA. (AstraZeneca owns a 9% stake in Moderna.) The project has been backed by $950 million in U.S. government funding. The vaccine provoked an immune response in all 45 patients tested in a Phase I trial; a study involving 30,000 volunteers launched in late July.

Cansino

A Chinese company that specializes in vaccines and is partly backed by U.S. drugmaker Eli Lilly, Cansino is testing a vaccine based on a genetic mutation of an adenovirus (the kind that causes colds). The Chinese military has approved the vaccine for use in its ranks; wider safety trials are in progress.

Murdoch Children's Research Institute

This Australian institution is experimenting with the bacillus Calmette-Guérin (BCG) vaccine—a drug introduced in the 1920s to fight tuberculosis. Studies have suggested that BCG protects against other respiratory infections; Phase III trials are underway to see whether that immune response extends to COVID-19.

A version of this article appears in the August/September 2020 issue of Fortune with the headline “A reborn pharma giant takes the lead against COVID-19.”