MAY 12, 2020

RA CAPITAL’S COVID-19 MAP:

Insights into a Coordinated Response Strategy
PETER KOLCHINSKY, PHD
Managing Partner, RA Capital Management

For more information about RA Capital’s COVID-19 coverage, please visit:

www.racap.com/covid-19

TABLE OF CONTENTS
Introduction.............................................................. 2

Vaccines................................................................... 2

Therapeutics............................................................. 5

Planning for Access................................................... 7

Navigating a TechAtlas Map..................................... 8

200 BERKELEY STREET • 18TH FLOOR • BOSTON, MA 02116 • WWW.RACAP.COM 1

RA CAPITAL’S COVID-19 MAP MAY 12, 2020

INTRODUCTION by remdesivir’s. Some of these are oral and may be

easier to manufacture. Remdesivir’s success thus far
The world is now embarking on a massive initiative to should spark a shift in resources to make the most of
fight COVID-19. Hundreds of companies, non-profits, those other programs.
academic institutions, and government agencies are
playing a part in an effort that some have compared Meanwhile, the map teaches us that we should be
to the Manhattan Project. The US government calls its thinking ahead further: are there enough syringes
own portion of this effort Operation Warp Speed. and needles to meet an undoubtedly unprecedented
demand for a COVID-19 vaccine? Who should receive
When executing on a large-scale, multi-pronged the limited doses of COVID-19 vaccine that may become
initiative, it is essential to have an overview of the available in late 2020 or early 2021? What drugs might
totality of the ongoing work. If each individual vaccine work best in combination, or in what sequence? Will
or drug or test is a chess piece, then one needs to see there be knock-on effects from COVID-19 that we need
the chessboard as a whole. Lists of all the programs to prepare for, such as a surge in demand for flu vaccine;
fall short in conveying the interdependencies and America normally only procures enough doses for half
opportunities for using different technologies together the population – should we put in a larger order now?
or in a particular sequence. How do we ensure that everyone can access COVID-19
treatments and vaccines? COVID-19 has accelerated
Our team has created a map that captures many of the
all facets of drug and vaccine R&D and regulation. Only
major ongoing technology development efforts in the
by looking at the big picture can we see the gaps in our
world to solve our COVID-19 crisis. The goals of the
vaccination and treatment strategies. Only by looking
map are to not only lay out the competitive landscape
at the big picture can we more quickly anticipate
of a given space, but to also reveal strategic insights
where to shift resources when a project fails (there will
that emerge when looking at the space in its totality.
be failures along the way). And only by looking at the
Historically, such technology landscape maps of
big picture can we judge how to make the most of any
cancer, auto-immune disorders, diabetes, and other
success.
diseases have guided our own investment of time and
money into what we believed to be the most important Below, you will find many of our key insights around
projects. We hope that this map will be of use to vaccine, therapeutics, and diagnostic development
BARDA, CEPI, and other grant-giving organizations as that we have gleaned from looking at the big picture.
they consider the best uses of their resources, and we
are making it available to the global community in the
hopes that it allows leaders to formulate strategies to
make the most of all the ongoing work.
VACCINES
As one example, today there is considerable funding
The vaccine programs that were first to reach clinical
going to vaccines that focus the immune response
testing are considered difficult to manufacture at scale.
on just the viral spike protein. While that is logical, it
Therefore, the vaccines that reach the market sooner
leaves us all undiversified in our vaccine development
than others might not be the ones that will necessarily
strategy. We would argue that at least some funding
be produced in sufficient amounts to meet the world’s
should go towards vaccines that rely on inactivated
needs. Some of the vaccines that may be among the
and attenuated virus.
first scaled to billions of doses may not be able to be
On the treatment side, while remdesivir is authorized for re-dosed every year if it turns out that we need to get
emergency use to treat patients with serious COVID-19, seasonal booster shots. Some vaccines might be more
it is in short supply, difficult to manufacture, and must readily combined with flu vaccines than others, which
be given intravenously (IV). But there are related drugs would be useful if we do someday need to give regular
in testing whose probabilities of success are informed booster shots and want to avoid having to give people

200 BERKELEY STREET • 18TH FLOOR • BOSTON, MA 02116 • WWW.RACAP.COM 2

RA CAPITAL’S COVID-19 MAP MAY 12, 2020

two injections. So, the various vaccine programs may is also a risk that a second dose of Oxford’s adenoviral
serve different purposes. The earliest ones to market vaccine does not work as well because our immune
might only benefit front-line workers; the first to scale systems generate antibodies against the viral vector
to high levels of production will unlock the global in response to the first dose (a problem particular to
economy; the ones that can be readily re-dosed and vectorized vaccines). In that case, we could possibly
combined with flu vaccine may take a bit longer to boost with a different vaccine.
develop but help keep SARS-CoV-2 permanently in-
check. Exploring Combinations
Combining an adenoviral vaccine for the prime dose
mRNA and other vaccines for the boost dose might even
Vaccines that deliver mRNA encoding SARS-CoV-2 maximize vaccine production capacity due a vaccine
Spike antigen could be among the first to be approved, phenomenon known as fractional dosing in which the
but most companies working with this technology can booster dose can sometimes be much lower than the
only manufacture them on the scale of a few kilograms. initial priming dose. So rather than using 1B adenoviral
The most important two questions for those programs doses to vaccinate 500M people twice and 500M
are how much mRNA will be required to generate an doses of a protein-based (or mRNA-based) vaccine
immune response and will the vaccines require more to vaccinate 250M people twice for a total vaccination
than one dose. If two doses are required (e.g. at least of 750M people, we might vaccinate 1B people with a
50ug each, which Moderna suggests is likely), then combination of one dose of the adenoviral vaccine and
those programs will initially serve to vaccinate tens a half-dose of the protein-based vaccine. The various
of millions of people per month globally. But if some large-scale vaccine groups should be collaborating
of these programs need only a single 5ug dose, then on joint animal studies with a goal of exploring these
mRNA could quickly vaccinate hundreds of millions “heterologous prime-boost” human studies. To our
of people per month by the end of 2020 or by early knowledge, such collaborations are not yet ongoing.
2021. We will know what doses are needed within a
few months: trials for vaccines by CureVac, BioNTech, Protein Antigens and Combinations with
Moderna, and others should read out by fall 2020. Flu Vaccines
We will eventually have options that are more
Adenoviral Vectors conventional than mRNA or viral-vector vaccines.
Johnson & Johnson (J&J) is scaling up production of its Novavax, for example, will start trials in May 2020 of
adenoviral-vector-based COVID-19 vaccine and plans two doses of its nanoparticle vaccine. With enough
on initially delivering hundreds of millions of doses, investment, they might be able to scale up production
with capacity to make 1 billion doses/year sometime considerably. Behind Novavax is Sanofi, which unlike
in 2021. However, J&J is not yet in the clinic. Their trials all the other companies mentioned so far, is one of the
will likely start this fall. But in late April, the University world’s few large vaccine manufacturers (Sanofi has
of Oxford ambitiously launched a controlled trial of a support from GlaxoSmithKline, another large vaccine
similar adenoviral vaccine in over 1000 people. This player, which has contributed a novel but proven
will tell us how good such a vaccine could be, point adjuvant to Sanofi’s effort). Sanofi sells conventional
to an effective dose, and indicate whether it needs to flu vaccines and is now working on a protein subunit
be boosted with a second dose. The Oxford vaccine version of the SARS-CoV-2 vaccine. That vaccine will
might scale up more slowly than J&J’s but it will give be slow to market but is potentially the easiest to
us a sense of how well and how efficiently J&J’s might scale using tried and true methods. Depending on
work. In the event that it only needs to be given once the adjuvant, it could also be combined with Sanofi’s
and at a low dose, then we might speculate that J&J’s existing flu vaccine, allowing for streamlined global
manufacturing capacity will stretch to vaccinate many distribution over the long run. It makes sense for
more people than originally planned. However, there governments and payers to think now about how and

200 BERKELEY STREET • 18TH FLOOR • BOSTON, MA 02116 • WWW.RACAP.COM 3

RA CAPITAL’S COVID-19 MAP MAY 12, 2020

how much they would want to pay for a combination Monitoring for Antigenic Drift (that would
flu/SARS-CoV-2 vaccine versus just the individual require updating vaccines)
parts. Making sure that people get the flu shot is going Based on sequencing of SARS-CoV-2 from patients
to be critical to minimizing false alarms from people all over the world, we are seeing the emergence of
getting the flu but scaring themselves and others into mutations in the Spike protein, including the critical
thinking it is COVID-19. Receptor Binding Domain (RBD) that we consider the
virus’ greatest point of vulnerability to antibodies. It
Vaccines and Antibody Tests is feared that if the RBD changes too much, then our
When we do have a vaccine, we need to anticipate vaccines will stop working against these new strains.
that antibody tests of the future need to be able to tell This phenomenon of antigen drift is one way the flu
the difference between antibodies from a vaccine and evolves slightly away from our vaccines each season,
antibodies from an infection. We have both of these requiring us to update flu vaccines to keep them
kinds of tests and need to make sure that the right ones effective. Therefore, we need vaccine manufacturers
are deployed and used at the right time in synchrony to regularly test whether the antibodies their vaccines
with vaccines to track the vaccines’ real-world generate can neutralize these new RBD-mutated
effectiveness. We will also need antibody tests that can strains. If they detect waning efficacy, we need them
accurately tell when a person has been reinfected; that to start working on updated versions of their vaccines.
is possible by looking for IgM antibodies, though those Reporting these findings to the public would go a long
suffer from poor specificity (high false-positive rate). way towards reassuring the public that news of new
Better IgM tests are required if we hope to be able to strains with feared mutations in the RBD region does
detect rates of reinfection without having to use PCR not mean that vaccines will not work.
to catch each reinfection (especially since reinfections
might be milder or even asymptomatic in patients with Just in case: Preparing for Potential
some immunity after a vaccination or first infection). Challenge Studies
There is a complex relationship between social
Passing the Safety Hurdle distancing and vaccine development. On the one hand,
Hanging over the whole vaccine field are several social distancing helps to flatten the curve until we get
questions about COVID-19 vaccine safety and efficacy. a vaccine. On the other hand, unless there is a certain
We all want to know if they will work in humans, and rate of new infections in the tested population (i.e. the
we know that several vaccines have been able to “attack rate”, such as 2% of people getting infected
generate antibodies in animals that protect those each month), it is impossible to demonstrate that a
animals from infection. We also know that these first vaccine actually works to prevent real-world infections.
few experiments have not observed a phenomenon Therefore, it will be important for vaccine developers to
called “enhanced disease” in which a vaccine helps the be able to run studies wherever in the world there are
virus infect certain immune cells and leads to worse high rates of infections.
outcomes. However, before we can be sure these
vaccines do not cause enhancement, it would be good Some have suggested testing vaccines on volunteers
to know that we can somehow induce enhancement in who are willing to then let themselves be exposed to
animals. Several laboratories are trying to do this (e.g. the virus to see if the vaccine protected them (so-called
use an intentionally poorly prepared vaccine to cause “challenge studies”). However, we currently neither
enhancement). As soon as one does, that will allow know the right dose of virus to simulate an infection
everyone to compare their vaccines in this model. in the real world nor do we have a supply of the virus
It will be essential that these preclinical studies are that would be considered suitable for such a trial. It
published as quickly as possible so that the field can probably makes sense to solve those two problems
learn from them. quickly in case it turns out that most countries in the
world have flattened their curves by this fall to such
an extent that vaccine trials are undoable. Regardless

200 BERKELEY STREET • 18TH FLOOR • BOSTON, MA 02116 • WWW.RACAP.COM 4

RA CAPITAL’S COVID-19 MAP MAY 12, 2020

of their feasibility, the ethics of doing such challenge Antivirals – Repurposed and Novel
studies are unclear and also require study. The first generation of drugs being tested for COVID-19
Making doses of remdesivir and any other proven are those that are already on the market or were
antiviral therapeutics available to such volunteers in being developed to treat other viral infections, such
challenge studies would provide an added level of as flu or hepatitis C. Gilead’s polymerase inhibitor
safety. After all, the point of a vaccine is primarily to remdesivir, originally developed to fight Ebola, has
prevent an infection. Volunteers could be vaccinated, already been proven to work to some extent and now
exposed to virus, and then tested every day or two has Emergency Use Authorization for the treatment of
to catch an infection early. Anyone infected can be hospitalized COVID-19 patients. We will likely know by
started right away on a therapeutic, lowering the risks June or July whether any other repurposed drugs are
to volunteers and making such a trial ethically more effective against COVID-19. Especially promising are
acceptable. oral polymerase inhibitors; they work in the same way
as remdesivir but are more convenient. Some of them
may even be easier to manufacture. The challenge with
these drugs is twofold: they may differ in their ability to
get to the lungs and the potency with which they work
on SARS-CoV-2.

THERAPEUTICS Behind these repurposed drugs are those specially

developed to target SARS-CoV-2, most prominently
Drugs can help infected people recover more quickly,
monoclonal antibodies. Antibodies are similar to
with fewer symptoms and reduced chance of death.
antibiotics and antivirals in that resistance is possible.
Safe and effective therapies to treat COVID-19 and its
If the virus’ Spike protein mutates in the region that an
symptoms can help bridge the way to widely available
antibody binds to, then the antibody may lose its ability
vaccines (likely in late 2021). Understanding what
to bind to the virus. Regeneron is developing a cocktail
kinds of treatments might be on the horizon can help
of antibodies, much like we use a cocktail of antivirals to
prioritize resources and anticipate future challenges,
manage HIV and cure hepatitis C. In case a virus has one
directing efforts to help plug gaps in our collective
mutation that prevents one antibody from binding, then
COVID-19 response. Once we know what types of
the other antibodies in the cocktail will still neutralize
drugs are showing signs of working, we can redouble
the virus. But there are a number of companies that
our efforts to find the best among them and the best
are, at least initially, only testing one antibody in the
ways to use them. Such incremental improvements
clinic. They should consider collaborating to create
have turned good older types of drugs into much
multi-antibody cocktails, similarly to how HIV pioneers
better newer ones, as we’ve seen with blood pressure
created combination pills to treat HIV.
medications and insulins. This same kind of iterative,
incremental innovation can be accelerated to address One challenge with all novel drugs is that making them
COVID-19 if we look a few moves ahead. at scale might turn out to be harder than proving that
they work. As we have seen with remdesivir, initial supply
Our COVID-19 arsenal can be split into two general
must be rationed. It is important that the discovery
buckets: drugs that attack SARS-CoV-2 itself, either by
or approval of an effective drug does not prompt an
preventing it from getting inside our cells or disrupting
uncontrolled easing of social distancing. Premature
replication once it is in there; or drugs that prevent or
celebrations or overconfidence will lead to more
limit the damage the virus can inflict, by protecting vital
infections than our drug supply can accommodate.
organs or properly calibrating our immune responses
– prodding an underactive immune response or Further complicating the question of supply is whether
dampening an overactive one that is doing more harm drugs that each are partially effective would be best
than good. used together or individually. If we have 250,000
courses each of remdesivir and a similarly effective

200 BERKELEY STREET • 18TH FLOOR • BOSTON, MA 02116 • WWW.RACAP.COM 5

RA CAPITAL’S COVID-19 MAP MAY 12, 2020

antibody, should we treat 500,000 people with one are taking a “throw everything at the wall” approach to
of the drugs or should we treat 250,000 people with treating this cytokine storm. However, especially as we
both? Remdesivir is only about 30% effective in terms get the rate of infection under some control, we need
of shortening the time to discharge from the hospital to think carefully about how we test these drugs in
and it appears to reduce the chance of dying by about the context of fewer severely ill patients. Repeatedly
the same. An antibody might offer similar efficacy alone, searching for the particular circumstances where any
but together maybe they would 95% effective. We need one drug might work, as we have seen the world do with
to answer these questions with clinical trials to make hydroxychloroquine, a generic drug that initially seemed
sure we get the most benefit from our early batches promising before failing in multiple studies, is inefficient.
of proven drugs. For example, that means making sure
It is critical to learn from early trials and accept when
that remdesivir is available at hospitals that are studying
a drug does not seem to work. We should carefully
therapeutic antibodies and that antibodies are tested at
allocate the finite supply of patients (at facilities that can
hospitals that have access to remdesivir.
run high-quality trials) to other drugs that might have a
Some groups are working on generating polyclonal better chance of working. This is less of a problem with
antibodies in animals. But they need a SARS-CoV-2 antiviral drugs since those can always be tested in milder
vaccine with which to vaccinate the animals so that they patients. But drugs meant to treat severe disease will
develop antibodies. Yet vaccine companies are focused face enrollment constraints. Like a hundred cars trying
on making vaccines, not helping other groups generate to get through an intersection without a system, no one
polyclonal antibodies for therapy. Still, it would be ideal makes progress. But if we are systematic, we can avoid
if some of the companies that have had success making a traffic jam and test the most promising candidates as
batches of their vaccines could collaborate with groups quickly as possible.
developing polyclonal antibodies in cows and horses to
How we prioritize which drugs get tested when patients
see if those might be a source of therapeutic antibodies
are in short supply is a key question. If one IL-6 inhibitor
before monoclonal antibodies come to market.
fails, we might argue that it makes sense to run one
Realistically, those would be only a few months apart at
more IL-6 inhibitor trial because of how plausible it is
this point, so if this work does not start right away, then
to believe that this mechanism should work to rein in
we may as well wait for monoclonal antibodies, which
cytokine storm. But if that second one fails, then all
will be in clinical testing in 3Q20 and could plausibly
IL-6 inhibitors should be deprioritized in favor of other
come to market in 4Q20, about the same time as the first
approaches. We can work our way through the various
low-scale vaccines and about four to six months before
categories laid out on the map systematically, instead
we estimate some of the large-scale vaccines become
of judging each drug by its own trial.
available.

Taming inflammation and Efficient

Development Strategies
There are many drugs that could potentially mitigate the
severe symptoms of COVID-19 by preventing a patient’s
immune system from overreacting and causing more
harm than good (these drugs are typically used to treat
rheumatoid arthritis and other autoimmune disorders
and therefore are all being repurposed). The trouble with
the immune system is that we really do not understand
it well enough to calculate the odds of any one of
these drugs working. It therefore makes sense that we

200 BERKELEY STREET • 18TH FLOOR • BOSTON, MA 02116 • WWW.RACAP.COM 6

RA CAPITAL’S COVID-19 MAP MAY 12, 2020

PLANNING FOR ACCESS

We hope you agree that it is inspiring to see how humanity has mustered its technological know-how in response
to this crisis. We have long seen this level of ingenuity and determination applied to hundreds of other diseases that
represent urgent unmet needs – our team has made over 100 maps spanning hundreds of disease areas covering
thousands of companies and technologies. We are certainly heartened to see what science can do to ease human
suffering, if we continue to invest in biomedical progress.

Today, COVID-19 is an urgent unmet need for all of us that has revealed the major gaps in America’s healthcare
system, particularly early in the crisis when insurance plans started to deny coverage of testing and care that the
CDC was recommending. In response to outcries and recognizing that patients disregarding guidelines due to out
of pocket costs was a threat to public health, insurance plans and the government have patched up some of those
holes. Tests are now generally well covered and the federal government has vowed to cover the costs of care of
millions of uninsured Americans. We now need to likewise secure affordable access future vaccines, drugs, and
tests for the long run, beyond the current crisis.

We should also remember that cancer, diabetes, and countless other diseases are a personal COVID-like crisis for
all those who suffer from them and their families. The reforms we consider for COVID-19 are the reforms we should
consider for every disease and for all of healthcare. With proper insurance, biomedical innovation can offer America
great value and remain affordable to each of us when we become patients in need of care.

200 BERKELEY STREET • 18TH FLOOR • BOSTON, MA 02116 • WWW.RACAP.COM 7

 RA CAPITAL’S COVID-19 MAP MAY 12, 2020

NAVIGATING A TECHATLAS MAP

MOST TECHATLAS MAPS ARE STRUCTURED IN A SIMILAR WAY:
Start at the center to learn some of the basics about the disease and how the map is structured.
In this case, we teach that the major goals are prevention and treatment. Then follow the branches outwards, a bit
like a “choose your own adventure” story, to learn about different approaches to achieving each goal. We teach in
the nodes what might make one approach better than another to guide your choice.

Similar technologies are grouped together. That allows us to spot when data from one program changes
the probabilities of success for related programs.

At the end of each branch, you get to the specific programs.

IN THESE ENDNODES, YOU’LL FIND:

SPONSOR: The name STAGE: Stage of development TRIAL DESIGN:

of the company or (in case of drugs, we use a If a clinical study has started, we indicate
companies (sometimes mouse icon to indicate that it’s with “RCT” if the trial is randomized and
universities) working still preclinical and a person controlled, which is important because
on the technology (i.e. icon to indicate that clinical uncontrolled trials are hard to interpret and
drug/vaccine/test). studies have started). learn from, so we’re primarily focused on
RCT data to know whether a drug really
works.

TECHNOLOGY NAME:
ROUTE OF ADMINISTRATION: Some programs have
Whether the drug is given orally, special designations that
IV, or via some other way. their sponsors refer to. DATA DATE:
When we expect to see
data from a clinical study

ORAL Fujifilm Holdings (Avigan/favipiravir) RCT P DATA:3Q20

FOR DRUGS:

P
REPURPOSED/NOVEL: THERAPEUTIC vs PROPHYLAXIS: IN HOSPITAL OR POTENTIALLY
Some drugs were already available or We illuminate the corresponding icon OUT-PATIENT:
in development to treat something else to indicate whether a drug is being We illuminate an icon to teach whether a
when COVID-19 struck and could be tested to treat infected patients or drug is likely to be limited to the hospital
repurposed into testing for COVID-19 to prevent exposed people from setting for various reasons or might be
quickly. If the Repurposed icon is not becoming ill. available at a pharmacy.
illuminated, that means the drug is Novel,
which means it likely will take longer to
start clinical studies.
200 BERKELEY STREET • 18TH FLOOR • BOSTON, MA 02116 • WWW.RACAP.COM 8
RA CAPITAL’S COVID-19 MAP MAY 12, 2020

FOR VACCINES:

1D Altimmune (AdCOVID)UAB INTRANASAL

# OF DOSES: Most vaccines will need to be given as a two-dose course (2D)

but there is a chance that some might work with only one dose (1D).

GEOGRAPHY: Because some companies working on vaccines may prioritize their home
regions (possibly because they are receiving grants from local agencies), we include a flag
to indicate what region a company is based in. Where there are collaborations, we indicate
the flags for each of the companies.

UPDATES:
With each update, we will highlight any changes since the previous update. This will include new programs being
added, new Map structures being added, and updated Milestones.

1D Altimmune (AdCOVID)UAB INTRANASAL

THE MAP ALSO HAS MANY OTHER INFOGRAPHICS DESIGNED TO TEACH:

• how long it will take to develop vaccines and drugs,

• how quickly courses of various vaccines will become available,
• how COVID-19 progresses from the point of infection and when certain antibodies start showing up that
diagnostic tests can detect,
• basic protective measures and how they work to flatten the curve,
• how the virus replicates and what that means for vaccine development.

200 BERKELEY STREET • 18TH FLOOR • BOSTON, MA 02116 • WWW.RACAP.COM 9