Six Months To A Vaccine: A New Manhattan Project for COVID-19

by Nikolai Eroshenko and Hannu Rajaniemi (vaccinemanhattanproject@helixnano.com)

Executive Summary (TL;DR)

At the time of writing, COVID-19 has forced one third of the global population into lockdown, killed nearly two hundred thousand people, and destroyed the livelihoods of millions.

It is widely reported that we can expect a vaccine in 18 months. On this trajectory, we will see millions of deaths. Extreme social distancing measures, continued indefinitely, may slow this down but will not prevent trillions of dollars in economic losses.

Even more sobering, an 18 month timeline to a vaccine is considered optimistic and even unrealistic by most experts. It is becoming clear that classic vaccine development is not fit for purpose in a global pandemic. And while many COVID-19 vaccine programs are underway worldwide, we are still thinking too small.

There Is An Alternative

While it may seem that we are already doing all we can, we are in fact at a crossroads. We can either continue on our current path or come together with a coordinated effort — a modern-day Manhattan Project — to develop a vaccine in six months.

The original Manhattan Project met an existential military challenge by pursuing multiple solutions to seemingly intractable technical problems in parallel, and by rapidly modernizing the necessary infrastructure.

Now, facing an existential challenge to public health and the economy, we have to do the same.

There have been other recent proposals to promote more parallelized vaccine development and manufacturing buildup, but trying to squeeze more speed out of existing technologies and processes will not get us under the 12–18 month mark.

We need a 6-month Vaccine Manhattan Project: a large concerted effort to: 1) upgrade our vaccine technology stack and 2) upgrade the methods we use to assess vaccine safety.

We can change our trajectory with:

• $10 billion
• Project leadership with appropriate experience (DARPA or similar).
• A simple collaboration framework with the right incentives that enables any company or academic group to contribute.
• An FDA/EMA task force to evaluate and develop novel clinical trial designs.

Given the enormous benefits of even a partial success, the cost of doing this is trivial. A $10Bn vaccine program amounts to the economic cost of eight hours of US-wide partial quarantine.

By approaching this crisis with a Vaccine Manhattan Project, we can:

1. Aggressively develop several parallel solutions to every technical problem (design, delivery, manufacturing, distribution logistics, etc). This can take place at an unprecedented speed (see Figure below) and removes the bottlenecks faced by vertically integrated companies.
2. Systematically address the safety of the vaccine (the single biggest development bottleneck) with the tools of modern biotechnology. Allergen tests and proteomics may be able to identify potential safety issues before a large group of Phase III clinical trial subjects are asked to subject themselves to the risks of new vaccines

A Manhattan Project vs linear vaccine development

Beyond the COVID-19 crisis, a Vaccine Manhattan Project will also: create the infrastructure to counteract future pandemics and other diseases, deter bioterror, create new jobs and vastly improve access to new medicines in the developing world.

For all these reasons, world leaders and philanthropists should marshal the resources and will to execute a Vaccine Manhattan Project to develop a COVID-19 vaccine as soon as possible.

History teaches us that combining money with determined leadership can have an extraordinary effect on scientific and human progress. We have done this before for tools of war. Now let’s do it for tools of health.

A more detailed rationale and some thoughts on what a Vaccine Manhattan Project could look like can be found below.

Why a Vaccine in Six Months?

The default answer to the question “When will we have a COVID-19 vaccine?” is currently 12–18 months. This is far too slow given the colossal human and economic impact of the current crisis. Purely from a dispassionate financial perspective, the cost of a partial shutdown in the US is ~5% of real GDP growth per month. This means that getting a vaccine six months earlier would be worth ~$6.4 trillion.

Dollars aside, a staggering number of lives are at stake. Mid-range estimates predict ~96 million Americans will get infected with SARS-CoV-2, with ~4.8 million hospitalizations, ~2 million requiring the ICU and ~480,000 deaths, unless extreme social distancing measures are kept in place. We can also expect many deaths from stroke, cancer, and other diseases, due to COVID-19 soaking up medical resources.

The alternatives to a fast vaccine are simply not good enough:

• New or repurposed therapeutics will eventually save many lives, but may still require hospitalizations — and therefore continued social distancing to manage ICU capacity.
• Using other human coronaviruses as a benchmark, a pure social distancing strategy may require societal shutdowns for more than two years.
• Widespread testing and targeted quarantines may be more effective — given frequent re-testing, highly accurate and comprehensive contact tracing, and travel restrictions. But given that as many as 60% of SARS-CoV-2 infected people are asymptomatic, doing this in a large and sparsely populated geography like the US may prove challenging — not to mention risk infringing on what many would consider basic civil liberties.

Finally, if we do not act quickly, the pandemic may become like the CoVs that cause common colds, i.e. highly prevalent and seasonal. Right now is a unique moment in time: we might still be able to keep the SARS-CoV-2 lineage from becoming a common human pathogen for all of future history.

For all these reasons, we need a vaccine fast — even faster than is traditionally considered feasible.

How does a Vaccine Manhattan Project make a 6-month vaccine possible?

Many vaccine experts consider even the 12–18-month timeline for the 80+ vaccine programs currently underway to be extremely accelerated (as in this recent proposal) and unrealistic. FiercePharma quotes SVB Leerink analyst Geoffrey Porges:

We view the current expectations for a vaccine in this timeframe as the equivalent of standing 24 feet (the usual distance is 8 feet) from a dartboard, with one dart in hand, and counting on a bullseye from one throw. It is theoretically possible, but highly unlikely, that such expectations are correct.

We agree — given traditional vaccine technologies and regulatory frameworks! But there are no laws of physics preventing us from going faster. We simply have to break the experts’ assumptions on technology and clinical trial designs. This requires a large concerted effort to upgrade both our vaccine technology stack and the methods we use to assess vaccine safety — in other words, a Vaccine Manhattan Project.

Such a program can outperform the prevalent model of vertically integrated companies pursuing independent programs in two important ways:

1. Better and faster safety and efficacy trials. The Manhattan Project can work with the FDA to develop a unified regulatory strategy that leverages modern proteomics to detect rare adverse events early without multi-year Phase III follow-ups, and utilize biomarkers and population-level statistics to assess efficacy in a matter of weeks.
2. Superior speed and risk profile through massive parallelism. A Manhattan-style project achieves speed by 1) identifying all points of technical risk, and 2) attempting to solve them in as many ways as possible in parallel.

Speeding up clinical trials while improving safety

Vaccines must pass a higher safety bar than other types of drugs because they will be given to an extremely large number of healthy people. As a consequence, large Phase III vaccine trials that track patients for serious adverse events for two years or more are traditionally used. This represents by far the biggest bottleneck in the development timelines. Here, we outline representative strategies to illustrate how to overcome these obstacles — a fully realized Vaccine Manhattan Project would naturally develop many more.

We believe that fully embracing the tools of modern biotechnology can enable faster Phase IIIs, while increasing patient safety by identifying potential safety issues before a large cohort of Phase III subjects are asked to subject themselves to the risks.

The safety risks for SARS-CoV-2 vaccines fall under three broad categories: 1) Acute anaphylaxis in response to the antigen or the adjuvant, 2) vaccine-induced allergies or autoimmune diseases, and 3) antibody-dependent enhancement (ADE) effects that can make future SARS-CoV-2 infections worse.

Acute issues typically occur within minutes of vaccination, and therefore can be rapidly flagged in Phase I and Phase II trials. Allergies and autoimmune disorders are the main reason that trials take a long time. However, it should be possible to detect warning signs of adverse immune response without actually causing them in people. For instance, allergen tests can be carried out before and after vaccination to identify unexpected shifts in allergen profiles. Proteome protein arrays and mass spectrometry methods can be used to detect the development of autoantibodies, allowing potential autoimmune issues to be identified in early vaccine subjects who do not go on to develop any diseases. These data can then be used to build custom assays for early diagnosis of potential side-effects in Phase III trials, eliminating the arbitrarily long follow-up period.

These clinical safety assays could also be used after approval for continued monitoring of possible safety issues — a significant improvement over the current approach of relying on patients and doctors to report serious adverse events.

ADE presents a special danger, as it might present as simply severe cases of COVID-19, making it non-trivial to detect under standard clinical trial designs. Basic virology studies carried out on patient sera could be used to determine whether ADE is observed in immune responses to SARS-CoV-2. If it is, serological assays should be developed to detect vaccine-induced ADE. Not only would this approach lead to faster clinical safety data, but it would allow subjects in danger of ADE of COVID-19 to be identified early. Such data would be actionable, as the patients can self-isolate more aggressively and their doctors would have more information when diagnosis and treat them if they do get infected.

For the efficacy aspects Phase II and Phase III trials, there are several options to speed things up. For example, it should be possible to characterize the immune response of patients that have recovered from COVID-19 and use these profiles as a “gold standard” efficacy biomarkers in Phase II clinical trials, dramatically reducing trial duration.

The prevalence of SARS-CoV-2 could also enable a new Phase III strategy for using population-level statistics, such as changes in new cases identified with widespread testing. Given the high contagiousness of CoV-2, and how prevalent it already is in many American communities, an effective vaccine should shift the dynamics of local spread. This may occur within a few days of the vaccinated subjects developing a strong IgG response, enabling efficacy measurements as early as five weeks into a clinical trial. In contrast, traditional clinical trials rely on measuring the protection conferred to each clinical trial subject, which may require >1 year of observation to obtain sufficient statistical power.

In summary, compressing SARS-CoV-2 vaccine clinical trials to a handful of months is not only possible using modern technology, they can actually be made significantly safer than traditional vaccine trials. Implementing novel clinical trial concepts such as the ones outlined above would permanently transform the vaccine development landscape and potentially unleash a wave of new vaccines in infectious diseases, cancer, and beyond.

Speed and lower risks through massive project parallelism

The Manhattan Project initially worked on three different methods for uranium enrichment and one for plutonium enrichment. The vaccine Manhattan Project can follow the same approach for antigen design, adjuvants, manufacturing, vaccination logistics, clinical trial design etc, using technologies sourced from several companies and academic groups. This carries significantly less risk than relying on vertically integrated companies, for two reasons:

1. Companies typically have to hit certain milestones to raise money. This forces them into a linear path of preclinical → scale-up → safety → efficacy. As a consequence, they are vulnerable to small “risk bottlenecks” in their research, development, and clinical processes. In a Manhattan Project, problems in each of those stages could be tackled in parallel from the start. See Figure 1 below.
2. A successful vaccine requires successful execution in multiple domains including antigen design, delivery platform, manufacturing and regulatory strategy. Companies, especially startups, are usually stronger in specific areas. For example, companies with sophisticated manufacturing can be weaker in antigen design, and vice versa. A Manhattan Project can simply take the best components from each company and rapidly build safe and inexpensive vaccines on unprecedented timescales.

Figure 1. A Manhattan Project vs linear vaccine development

Below, we present an example program skeleton with components that can be addressed in parallel.

Basic SARS-CoV-2 clinical virology

• Study patient sera for evidence of antibody-dependent enhancement (ADE) to guide the design of maximally safe vaccine candidates.
• Track the dynamics of immunological memory of recovered patients.
• Correlate patient genetics (for ex. ACE2 alleles) to disease severity.

Antigen design

• Use high-throughput biochemical methods to map antigens that are unlikely to give rise to ADE antibodies to maximize safety.
• Use computational structure prediction to design antigen fragments.

Adjuvant

• While the FDA has traditionally had concerns with “novel adjuvants”, next-generation protein-based adjuvants may reduce the effective dose of the vaccine by a factor of 10–100, and vastly accelerate scale-up . Additionally, long-term protection may require a more powerful immune response than can be achieved using traditional vectors. We believe advanced adjuvants should be considered: deep immunoprofiling in Phase I and II as discussed above should flag up potential issues.
• We propose screening 100s of adjuvants + antigen-adjuvant combinations in mice for potent humoral responses.

Manufacturing

• Scale-up mRNA/replicon, DNA, protein, and delivery lipid manufacturing.
• Carry out fundamental technology development aimed at radically reducing manufacturing costs across all four platforms.
• Develop a low-cost, painless electroporation device.
• Vaccine distribution logistics nationally and globally.

Clinical:

• Provide a unified point of coordination with the FDA.
• Characterize the immune response of patients that have recovered from COVID-19 and use these profiles as a “gold standard” to be 1) guide for developing meaningful readouts in pre-clinical models; 2) used as efficacy biomarkers in Phase II clinical trials.
• Pre-enroll patients into clinical trials, including using epidemiological data to inform study design to decrease the necessary sample sizes required for Phase III, while increasing the direct public health benefits of the trial.
• Develop allergen and autoantibody panels as discussed above.

Program Management / Business Development

• Reduce overhead and friction for coordinating between the different entities.
• Identify potential opportunities for even closer integration between companies or academic labs, and act as an honest broker/matchmaker for structuring deals.

A Pan-CoV vaccine

In parallel to developing a lead vaccine aimed at SARS-CoV-2, resources should be marshaled to lay the foundation of a broad, multi-epitope vaccine that can guard against viral evolution and provide protection from future coronavirus outbreaks:-

• Use basic virology and epidemiology studies to predict epitopes that are expected to be highly constrained by evolution.
• Apply approaches from cancer vaccine technologies to making a multi-epitope T cell vaccine that broadly targets the viral proteome.
• Coordinate access to BSL3 resources for carrying out viral challenge animal tests involving live SARS-CoV-2.

How to Run a Manhattan Project?

We already know how to do this. A lot of the institutional memory from the original Manhattan Project was captured in the DARPA model, which has shown how to structure project teams and recruit the right leadership to solve massive technical problems

The participating organizations can also be provided with strong incentives to collaborate. For example, an obligatory licensing option to prevent individual companies from blocking the final lead candidate integration can be used as a stick. And there are many carrots to go around:

• Smaller companies get access to manufacturing/clinical/etc resources at an early stage of their program.
• Larger companies get access to bleeding edge technologies from smaller companies.
• Everybody benefits from funding for relevant projects.
• Every participant gets equity in the final product.

We believe that with these playbooks, a vaccine Manhattan Project can be executed by the federal government, a strong alliance of super-philanthropists, the international community, or a coalition of all three.

We will outline ways for rapid Manhattan Project initiation and leadership requirements in more detail in future posts, but suggest that technical program management skills be emphasized over traditional biopharma experience when recruiting project leadership.

Vaccine Manhattan Project Benefits Beyond a SARS-CoV-2 Vaccine

Leveraging the full power of the US innovation ecosystem to successfully deliver a SARS-CoV-2 vaccine has several benefits beyond the vaccine itself:

• Astonishingly rapid response will send a powerful message about the US biodefense capacity and discourage bioterror attacks.
• The network of collaborators will create an ecosystem that will lead to greater company formation — a “biotech Silicon Valley” effect. This will help the US to cement its position as a global leader in biotechnology and demonstrate the power of the open innovation model.
• Building out manufacturing across a range of new modalities will create the foundation of a manufacturing base. This makes sense to set up in non-coastal states, which could create new hubs of biotech manufacturing — “biotech Detroits”.
• Multi-antigen T cell vaccines may provide a path towards pan-influenza vaccines, which could save 250,000–650,000 lives each year globally.
• This effort will speed up vaccine technology development by many years, not only catalyzing the infrastructure to prevent the next pandemic, but potentially creating the foundation for advanced vaccine applications such as prevention of cancer and even neurodegenerative diseases.
• Biomanufacturing innovations developed in the project will enable manufacturing of high-purity, safe nucleic acids, and proteins at an unprecedented scale. This will radically reduce the cost of nucleic acid and protein vaccines. Furthermore, adjuvant technologies should lower cost per dose by 1–2 orders of magnitude. Taken together, these improvements will allow the global poor to access gene therapies, cancer vaccines, and other medical technologies that have to date only been available to patients in the developed world.

For all the reasons outlined above, we urge world leaders and philanthropists to marshall the resources and leadership to execute a Manhattan Project scale effort to develop a SARS-CoV-2 vaccine as soon as possible. To mirror Oppenheimer’s famous quote from Bhagavad-Gita, let us become Life, the Preserver of Worlds.

How You Can Help

This is a living document and we welcome feedback (vaccinemanhattanproject@helixnano.com) from anyone who has read this far.

If you agree that we need a Manhattan Project to develop a SARS-CoV-2 vaccine, please share this post on social media — or, better yet, send it to your congressperson, senator, or anyone else who can help make it happen.

Acknowledgments

We thank Seth Bannon, George M. Church, Tom Kalil, Zuzana Krejciova-Rajaniemi, Ela Madej, Michael Nielsen, Lenny Raymond, Cooper Rinzler, Victoria Song and the HelixNano research team for useful comments; and Carina Namih for very helpful edits.