About a year ago, Chaz Calitri, the head of operations for sterile injectables at Pfizer, was at home in suburban Philadelphia, when he got a call from his bosses. The company was moving forward with an experimental COVID-19 vaccine. Calitri, a chemical engineer by training, was in charge of Pfizer’s manufacturing site in Kalamazoo, Michigan, where the constituent parts of the vaccine would eventually be assembled before being shipped across the country. “At first, I was really excited,” he told me. “And then, after I sat down on the couch and started thinking about it, I was horrified, because I knew that it was going to take the full force of everything we could throw at it.”
Typically, vaccine manufacturing doesn’t begin until a candidate has proved to be both safe and effective in animal and human testing. In the past, that process could take ten years; Pfizer’s vaccine, which was developed in collaboration with BioNTech, a German biotechnology company, took a record ten months. The vaccine received emergency-use authorization from the F.D.A. on December 11th; two days later, the company began shipping tens of thousands of doses, all of which had been made while clinical trials were still under way.
In Kalamazoo, on a campus larger than Central Park, Calitri’s team worked around the clock. Pfizer hired roughly seven hundred workers, reassigned experienced engineers to the vaccine effort, and increased the number of vials that it could produce by installing additional “fill and finish” machines. Even so, by the end of 2020, the company had delivered only half of its initial production goal of a hundred million doses. A Pfizer spokesperson told me that, among other things, “securing enough raw materials took longer than we expected.”
Five days after his Inauguration, Joe Biden set a goal that a hundred and fifty million vaccinations would be administered in the first hundred days of his Presidency. At the time, about eight hundred thousand Americans were receiving a vaccine each day, most of them health-care and other front-line workers. The Trump Administration had left distribution planning up to the states; as vaccine appointments were made available to older Americans in many states, in mid-January, some vaccination sites were flooded with requests, but others sat relatively empty. “It was like running out on the field during the Super Bowl and telling the players to just do whatever they want,” Bruce Y. Lee, a professor of health policy and management at the City University of New York Graduate School of Public Health and Health Policy, told me. “So it’s actually not a big surprise, given the apparent lack of planning, that we’ve run into a lot of problems with the vaccine distribution.”
Even as more vaccines became available—by early February, roughly a million and a half shots were being administered a day—the quest for a vaccination appointment was being compared to winning the “Hunger Games” or beating the house at a Las Vegas casino. So far, just over ten per cent of the population has been fully vaccinated. The reason for the slow pace, as Representative Diana DeGette, of Colorado, said at a recent congressional hearing, is that “we still face a lack of vaccine supply to meet current demand.”
On February 11th, Biden announced that the government had signed a deal with Pfizer and Moderna—the latter’s vaccine had been authorized for emergency use by the F.D.A. on December 18th—for an additional hundred million doses each. Both vaccines are based on a messenger-RNA platform, or mRNA for short, that has previously never been produced commercially. At the end of February, Johnson & Johnson received an emergency-use authorization for its one-shot vaccine, which is based on a DNA platform similar to the company’s Ebola vaccine. A few days later, Biden announced that Merck, whose own COVID-vaccine efforts had failed, was going to help Johnson & Johnson to boost production.
The White House has sounded particularly optimistic this week. On Wednesday, the President announced a plan to secure an additional hundred million doses of the Johnson & Johnson vaccine. On Thursday, during his first prime-time Presidential address, he directed states to make all adults eligible for the vaccine by the beginning of May. The catch: none of the increased supply that has been established since Biden’s Inauguration will be available until late spring, at the earliest, and most of it will not arrive until the end of the year. All adults may be eligible to receive a vaccine in a couple of months. But whether doses actually will be available will depend on a lot of things going right.
When the President of the United States places an order for millions of doses of a COVID-19 vaccine, they do not simply appear, like Amazon packages, two days later. For much of the past year, it has taken Pfizer a hundred and ten days to produce each vial of vaccine. The time line starts at the company’s plant in Chesterfield, Missouri, outside St. Louis, which houses a cell bank of frozen E. coli bacteria. Scientists extract DNA from the E. coli cells to grow the template, called a plasmid, on which the vaccine’s mRNA will be built. Once the plasmid is made, purified, and tested, the double-helix structure of the DNA has to be linearized—literally, made linear. The process takes about ten days, after which it goes through additional testing. “We’re going twenty-four hours a day with three manufacturing shifts,” Christine Smith, the Chesterfield-site leader, told me. “And then there’s another shift making all the buffers and media to grow the cells in and getting ready for the next day. It’s a very regimented process. It’s not like we can just open up a door to the room next door and start making it.”
From Missouri, the plasmid is flown to Pfizer’s campus in Andover, Massachusetts, where it is incubated in a bath of enzymes and nucleotides—the building blocks of RNA—for several hours. The process, called in-vitro transcription, synthesizes the genetic material, the RNA, which carries the instructions to make a modified form of the spike protein that causes COVID-19. (These reëngineered spikes are what trick the immune system into creating antibodies to defend against the coronavirus.) A few days later, the RNA is placed in specially designed bags, frozen, and flown overnight to Kalamazoo, where Calitri’s team puts the final drug product into vials, and inspects and labels them before freezing them at ultra-low temperatures. When it’s time to ship them out, the vials are packed with dry ice—Pfizer has its own dry-ice-manufacturing facility on site—in thermal containers created specifically for this vaccine, each with its own G.P.S. unit and temperature alarm. (When the initial vaccine drive is over, one of Pfizer’s shipping containers will be sent to the Smithsonian, Tanya Alcorn, the company’s head of biopharma supply chain, told me.)
Both the Pfizer-BioNTech and Moderna vaccine candidates require rare ingredients that are in short supply, such as the lipids used to encase the mRNA and the enzymes used to transcribe it. Calitri, meanwhile, had been grappling with a series of engineering puzzles. “There’s a step in which the mRNA is coated with these lipids, and it’s done in a specialized mixer,” he told me. “The mixers we were using to develop the process are very small”—about the size of a silver dollar. His team didn’t have time to design a larger mixer, so they tied together a hundred of the silver dollars. When the filters on some of the filling equipment needed to be replaced, switching to a different filter was not an option, because any adjustment to the process would have to be approved by the F.D.A. Instead, the team had to learn how to “regenerate” the ones they had. It took six months and numerous prototypes to figure out how to store and ship a frozen product that needed to be kept at subzero temperatures. There were some misses, too. They thought the vaccine would need to be frozen as soon as it came off the filling line, so they installed blast freezers; the data have since shown such precautions to be unnecessary. “We needed to have options,” Calitri explained.
Even before the clinical trials were completed, it was obvious that Pfizer’s domestic operation would not have enough capacity to meet the U.S. demand. In July, Pfizer ordered two prefabricated modular manufacturing suites, but they took eight months to build and finally arrived in Kalamazoo in mid-February. “This is not like a production line for making cars or trucks,” Tim Manning, the supply coördinator for the Biden Administration’s COVID-19 response team, told me. “This is extraordinarily complex biochemistry. And it happens at the molecular level. . . . It’s really complicated, and made on extremely rare and difficult-to-make machinery.”
Most years, the health-care supply chain is fairly stable. Hospitals anticipate how many N95 masks, nitrile gloves, and various medications and vaccines they will need based on what they’ve needed in the past. The pandemic year exposed the fragility of that system. In early 2020, with a COVID-19 vaccine still on the distant horizon, Rick Bright, the director of the Biomedical Advanced Research and Development Authority, warned the Trump Administration that, once there was a vaccine, there would likely be a shortage of syringes, needles, and glass vials. At the time, manufacturers were producing around fifteen to twenty billion glass vials, for all of the world’s medications, in a typical year. Bright was fired in April; in a searing whistle-blower complaint filed in May, he predicted that it could take up to two years to produce enough vials just for the U.S. vaccination effort. “They can’t just crank out more vials,” Kelvin Lee, the director of the National Institute for Innovation in Manufacturing Biopharmaceuticals (NIIMBL), told me. “That glass gets manufactured through very specialized processes. And, ultimately, glass comes from sand. Their raw-material suppliers have to think about where they are going to get the right kind of sand to make sure the vial is of appropriate quality.”
The government has been able to use the Defense Protection Act to secure a sufficient number of vials so far. Some recent advancements in glass technology will likely help, too. The same month as Bright’s whistle-blower complaint was filed, Pfizer signed a multiyear contract with Corning, which is based in New York and manufactures a super-strong pharmaceutical-grade glass called Valor. In June of last year, the Trump Administration awarded Corning more than two hundred million dollars to scale up production. But that deal will address only a fraction of the need. Chandra Brown, who was the Deputy Assistant Secretary of Commerce for Manufacturing in the Obama Administration, recently wrote in an online editorial, “By this time next year, I predict Americans will covet borosilicate glass”—the material used in traditional vials—“the same way they do N95 masks.”
The vials also need rubber stoppers. Last fall, tropical storms in rubber-producing regions of Thailand, Vietnam, and India led to shortages that could have jeopardized the vaccination effort. The government used the Defense Protection Act to round up sufficient supplies, but it was clear that a tremendous strain had been placed on the world’s rubber supply. “The D.P.A. is allowing the U.S. to hoard some of these materials for production of U.S. vaccines, but is causing other shortages globally,” Robert Handfield, the executive director of the Supply Chain Resource Cooperative, and a professor of supply-chain management at North Carolina State University, told me. He also said that there is “very little visibility into the manufacturing bottlenecks that are occurring.” On March 5th, the Times reported that officials in the United States and Europe say that they may not have enough syringes to administer the vaccine.
The supply of lipids used in both the Pfizer and Moderna vaccines continues to be precarious, too. Vox recently reported that, even among the few companies whose facilities can be repurposed to make lipids, “not nearly enough of them are ready to make the kind of lipid nanoparticles we’d need to distribute billions of mRNA vaccine doses quickly.” As Stéphane Bancel, Moderna’s C.E.O., told investors in January, if “there’s one raw material missing, we cannot start making products, and that capacity will be lost forever because we cannot make it up.”
Recently, the Biden Administration has used the Defense Protection Act to acquire enough low-dead-space syringes to be sent out with every Pfizer-vaccine shipment. (Such syringes enable a sixth dose to be extracted from Pfizer’s vials, automatically increasing the company’s vaccine doses by twenty per cent.) With government support, a company called ApiJect is building a “Gigafactory” in North Carolina to manufacture single-dose injectables to reduce waste and simplify the distribution of vaccines. (It is expected to come online in 2022.) The White House is also investing in the construction of factories that would be able to make more than a billion surgical gloves a month. The goal is to move enough production Stateside, so that the domestic health-care supply chain is not dependent on other countries, which, in a crisis, will likely choose to prioritize their own citizens.
Perhaps most crucially, the government brokered a deal between Johnson & Johnson and Merck, paying Merck up to $268.8 million to upgrade two of its manufacturing facilities. But it will take months for Merck to retrofit its facilities; the vaccines it will be making for Johnson & Johnson are not expected to be ready until the second half of the year. Meanwhile, the company that Johnson & Johnson currently contracts with to produce its vaccines domestically has yet to receive F.D.A. approval. (The four million or so Johnson & Johnson vaccines that are now being distributed were made abroad.)
The most hopeful news is that Pfizer has cut the time it takes to make a batch of its vaccine to sixty days. As of mid-March, the company expects to deliver more than thirteen million doses a week, up from around five million last month. At a congressional hearing in February, John Young, Pfizer’s chief business officer, explained that the company has begun making its own lipids, and has increased capacity at its facilities in Kansas and Wisconsin (in addition to the new production suites in Michigan). It has also doubled batch sizes, increased yields per batch, and developed faster laboratory tests.
“We’re getting better at it,” Calitri said, of the manufacturing process. “I think people don’t know how challenging it is to make billions of doses of a product that you did not have a process for nine months ago. And then to scale that up even further. There’s so much involved from an engineering perspective, from a quality perspective, from a compliance perspective, and from a safety perspective. We’re not making widgets. We’re making a product that people inject into their bodies—into healthy humans—and it has to be perfect. We need to make sure of that for every single dose. That takes engineering, it takes science, it takes time.”
More on the Coronavirus
- There are three moments in the yearlong catastrophe of the pandemic when events might have turned out differently.
- Citizens around the world, from Brazil to Rwanda, share their experiences of the pandemic.
- In countries where the rate of infection threatens to outstrip the capacity of the health system, doctors are confronting ethical quandaries.
- Surviving a severe coronavirus infection is hard. So is recovering.
- Can the COVID-19 vaccine beat the proliferation of new virus mutations?
- The pandemic has presented companies with an unprecedented opportunity to rethink the fundamentals of the physical workplace.
