By Nicholson Price, Rachel Sachs, Jacob S. Sherkow, and Lisa Larrimore Ouellette
Many worries dog the race for a COVID-19 vaccine. Is the FDA going to bow to political pressure and possibly approve a vaccine prematurely? How much will a vaccine be rewarded by governments and how will it be allocated to patients? How will distribution work, when the world faces potential supply-chain shortages ranging from glass vials to low-temperature freezers to manage the challenging cold chain? Will enough people take the vaccine to reach herd immunity? Can the government overcome the understandable skepticism of many Black Americans toward a vaccine produced by a medical system with a long history of systemic racism?
Alongside these concerns is another substantial challenge: how are firms going to make hundreds of millions—eventually billions—of doses quickly and effectively? As one of us (NP) recently discussed in work with Professors Arti Rai and Timo Minssen, vaccine manufacturing is hard, and the world needs it to happen fast. How should innovation policy scholars and policymakers think about this issue?
What are the challenges with large-scale manufacturing of vaccines?
Antiviral vaccines are famously finicky to manufacture. The oldest, most reliable manufacturing method uses—believe it or not—chicken eggs. Eggs are carefully injected with a “live” inoculant virus; the virus then replicates inside the eggs; the eggs are then carefully “decapped” to remove their tops; the viral “aspirant” is then sucked out; and the virus is then purified, chemically separated, and heat inactivated or “killed.” These killed bits of virus are then, through additional steps, used as the basis for making a pharmaceutically acceptable vaccine.
The process—like egg farming itself—is labor intensive. In addition, it requires a host of specialized facilities and specialized equipment not found on typical poultry farms. It also requires lots and lots of eggs. The U.S. government has its own specialized egg producing facilities that can manufacture emergency stockpiles of vaccines; by some accounts these facilities require upwards of 900,000 eggs at a time. This is a major, if not the major, difficulty in scaling egg-based manufacturing. According to Dr. Martin Friede, coordinator of the Initiative for Vaccine Research at the World Health Organization (WHO) in Geneva, “Egg production is a huge bottleneck . . . . You can’t just call your local egg farm and say tomorrow I need 10 million more eggs.” In addition, even while the vaccine egg market is relatively small to the commercial egg market, they do compete—making sudden shifts in demand on either side potentially shortage-causing.
There are newer approaches to be sure—vaccine manufacturing doesn’t rely just on eggs. Some vaccines are made by culturing the virus in mammalian cells, rather than chicken eggs. The famous polio vaccines developed by Jonas Salk and Albert Sabin were both based on monkey kidney cells, for example. Others are manufactured by using bacteria to produce recombinant versions of the viral proteins thought to induce protective immune responses. These methods are complicated for a variety of reasons and require an intense focus on preventing contamination and ensuring that the product’s protective effect is against the targeted virus. As a consequence, such vaccines are not readily scalable (at least relative to the incredible, edible egg).
Two leading COVID-19 vaccine candidates, Moderna’s and Pfizer/BioNTech’s, stray even further from traditional techniques and don't rely on viral proteins to produce a protective immune response. Instead, the active portion of these vaccine candidates consists only of mRNA—DNA’s molecular cousin. In theory, a key advantage of mRNA vaccines is the speed with which they can be synthesized and manufactured, although the technology has never previously been scaled up for commercial production. The most critical aspect, of course, is ensuring the mRNA sequence produces the “correct” immune response with minimal side effects. This method of vaccine production—while promising—is entirely novel; no other mRNA vaccine has been approved anywhere, for any other condition.
At a broader level, all of these vaccine manufacturing processes require enormous amounts of infrastructure—whether government-owned egg farms or massive RNA synthesizing facilities. And because each vaccine product is slightly different, different plants cannot readily be switched back and forth from different types of vaccines. Moreover, they typically take around five years to build. Finding appropriate incentives to invest in establishing and operating such facilities is problematic, especially when the results are uncertain. This is one of many reasons why encouraging excludability for vaccine manufacturing is both insufficient and unnecessary. In addition, “the basic vaccine strategy has not changed [for over 60 years], primarily because the low profit margins provided by the influenza vaccine product did not support investment in new production technologies.” This dynamic produces a disconnect; firms have little incentive to innovate in manufacturing or invest in at-risk capacity, but producing vaccines at-risk for a problem like COVID-19 produces huge social gain relative to its cost.
What current incentives and programs are addressing this challenge?
The concern that social gains from building manufacturing capacity at-risk exceed private benefits does not mean that no investments will be made in vaccine manufacturing without intervention; Pfizer, for one, apparently began to shift manufacturing capacity very early in its development process. Rather, it means that the level of manufacturing investment will be less than society should want. To address this problem, policymakers are deploying a range of programs both to encourage early manufacturing of COVID-19 vaccine candidates, even before one is authorized for marketing, and to promote innovation in vaccine manufacturing technology.
First, some governments and nonprofits are providing direct support to firms to help scale up production. In the United States, for example, efforts under Operation Warp Speed include funding provided in the spring for vaccine developers Moderna, AstraZeneca, and Johnson & Johnson (with the latter two using adenovirus vector approaches) to invest in manufacturing. This summer, HHS also entered agreements to advance domestic manufacturing capabilities, such as with Texas A&M’s Center for Innovation in Advanced Development and Manufacturing, which is at least partially reserved for the Novavax recombinant-protein vaccine candidate. These efforts are supporting vaccine candidates with at least three different mechanisms of action—and three different manufacturing strategies. Hopefully, then, even though retooling a facility for a totally different type of vaccine is hard to do quickly, there will be some extra capacity available for whatever types of vaccines are eventually demonstrated to be useful. The need for vaccine-specific manufacturing investments is why Bill Gates announced in April that the Gates Foundation would fund factories for seven promising vaccine candidates, although we have not seen subsequent news about these investments.
In addition, we have previously written about the use of advance purchase commitments to support the at-risk manufacturing of vaccines before they complete the relevant clinical trials. These commitments are negotiated between particular firms (such as Moderna and AstraZeneca) and governments (including the EU) or with the COVAX facility run by the WHO, Gavi, and the Coalition for Epidemic Preparedness Innovations, which is a framework for both pooling incentive efforts and attempting to equitably allocate resulting doses. In some cases, nonprofits such as the Gates Foundation or Mexico’s Carlos Slim Foundation have helped finance these agreements. Advance purchase agreements can also be made with manufacturers, such as the Gates Foundation’s payment for up to 100 million doses of vaccines manufactured by the Serum Institute of India, the world’s largest manufacturer of vaccines, which is working on the AstraZeneca and Novavax candidates. Advance purchases mitigate the financial risk companies would otherwise face in manufacturing vaccines even before gathering evidence about their efficacy, and some of the agreements include milestone payments to encourage developers to meet particular goals. Continuing these efforts can help get more vaccine doses over the finish line quickly, and we think there is more room for international coordination—such as a reconsideration of the U.S. decision to avoid COVAX.
What else should policymakers do to help large-scale manufacturing?
Investing in pre-approval scale-up of manufacturing is important for increasing the speed with which an effective vaccine can reach billions of people, but it is not the only step policymakers can take. Governments also need to encourage the transfer of knowledge about vaccine manufacturing, as emphasized in recent pieces by Professor Ken Shadlen and Professors Nicholson Price, Arti Rai, and Timo Minssen.
As we explained above, vaccine manufacturing is finicky. Vaccines are not easy to reverse engineer or copy—there are no generic vaccines (known as biosimilars) or even regulatory guidance on how to do it. The lack of biosimilar vaccines likely stems both from low profits in vaccine markets and protection of manufacturing details as trade secrets. Currently, to expand manufacturing, the developer’s involvement for transferring know-how about their vaccine candidates is essential, making patents largely redundant. The problem of technology transfer for vaccines is not new, but the COVID-19 pandemic has given it increased urgency.
In some cases, firms have been willing to share biopharmaceutical manufacturing details independently. For example, as Price et al. note, six firms recently agreed to share information about manufacturing monoclonal antibody treatments for COVID-19 (for which they received the Department of Justice’s blessing under antitrust laws). For an effective COVID-19 vaccine, initial demand will far exceed supply, so developers will have a strong profit incentive to license their technology and transfer related knowhow to multiple manufactures. Vaccine developers currently have little private incentive, however, to share their manufacturing insights with firms making competing products.
To encourage greater sharing of vaccine manufacturing knowledge, policymakers can use incentives or mandates. For example, Price et al. suggest that when governments or NGOs make the large financial commitments described above—whether through direct funding or advance purchase agreements—they should attach a requirement to transfer manufacturing know-how. Such incentives could help spur a longer-term shift in developing a broad knowledge about how manufacturing works, for vaccines and otherwise; incentives for drug manufacturing innovation are generally problematic. And investments in greater knowledge-sharing now could have payoffs not just for this pandemic, but also for future ones.
This post is part of a series on COVID-19 innovation law and policy. Author order is rotated with each post.
