There is an immense effort underway to develop vaccines that can offer partial protection against COVID-19. Pharmaceutical companies have raced through the early stages of development with unprecedented speed (and an unprecedented Niagara of federal funding, $11B and counting), and four of them have candidate vaccines in the US FDA’s Phase III trials.
This is a truly spectacular and unprecedented achievement. If all goes well, the companies will begin to see useful data from these Phase III trials as early as the end of October, and more definitively by the end of the year. Several more vaccines are close to entering Phase III, and over 150 others are in some stage of development. There has never been a vaccine effort like this before, in part because there hasn’t been a pandemic on this scale since vaccines were invented. If useful vaccines emerge, they will set a world record for speed by a considerable margin, and demonstrate what big pharma can do when generously motivated.
Despite this monumental effort, there’s a lot we don’t know yet about these potential vaccines, which means there’s a lot we don’t know about just how much benefit we’ll get from them. Our ignorance can be usefully organized into four categories: efficacy, durability, safety, and scalability. The rock star vaccines we all know about—polio, smallpox, measles, mumps, etc—all score highly in all four categories, but many other vaccines do not. The first generation of COVID-19 vaccines now racing through the trials will be far from perfect, even if they’re still very much worth having.
Here’s your handy guide for the perplexed:
Efficacy is a measure of what percentage of those who receive the vaccine will actually be protected from infection. Polio, naturally, is the star, at pretty much 100%, while influenza ranges from a barely noticeable 10% up to a reasonably valuable 60%. Anything over 50% for SARS CoV-2 would be likely to win approval. Handicappers right now are looking for 55-60%. One company believes its candidate has a chance to hit 70%, but no promises yet. Any or all of these candidates could crash and burn in Phase III—it’s not uncommon, or, at the other extreme, one might hit an unlikely home run with unexpectedly high efficacy (not likely).
At any efficacy between 50% and 70%, the vaccines won’t be able to create the vaunted condition known as “herd immunity.” For this virus, herd immunity might require 80% to be immune, and there may not be 80% of Americans even willing to take it, so herd immunity is a heavy lift reserved for a future year. For the near future, relative pandemic safety will still depend on help from non-vaccine medical treatments, and we’ll still need behavioral protocols—ubiquitous self-testing for example, continuing use of masks and distancing, limitations on indoor gatherings, and so forth. This is not to say that partially effective vaccines should not be embraced—they will save lives and probably reduce severity of infection—but they are not a pandemic panacea. It’s possible that the extreme rush to get these early vaccines out, called Operation Warp Speed, will result in less effective vaccines dominating in the first round, which will slow the arrival of better vaccines and therefore cost lives. We’ll see how it plays out.
Durability is a measure of how long protection lasts after vaccination. Some vaccines—polio is the poster child—give lifetime immunity. Others—influenza is the rotten child—last 90 days, more or less. We’ll learn about the vaccine’s durability as it goes into use. Durability of one year or more will be vastly preferable to anything less. It’s possible to imagine the whole world getting on an annual vaccination cycle, but hard to imagine anything tighter.
Safety is a measure of what proportion of the population can safely receive the vaccine, expressed as a percentage. The vaccine developers learned in small Phase II trials that their vaccines were not obviously dangerous, but vaccines against global pandemics need to be exceptionally safe, since they work by being administered to very large numbers of people. One risk, for example, is that a new vaccine might make the infection worse by confusing a patient’s immune system. This is the kind of issue that might not be discovered until Phase III or even later. This kind of risk has to be considered with these vaccines, which have been developed so quickly, some using very novel methods, against a new and unknown virus.
In addition to risks of adverse effects that only show up at large scale, some vaccines are not safe for specific demographic groups. The groups at greatest risk of being unable to be vaccinated are older people and very young ones. Since older people are especially vulnerable to this virus, having a vaccine that is safe and effective for older people will be critically important, but is not guaranteed. The reason for the risk to older people is that immune systems weaken with age, so vaccines must be stronger to stimulate them. Higher vaccine doses can be dangerous to older people, so there’s a crossover point where a vaccine dose strong enough to stimulate a useful immune response is dangerous in itself.
The other risk incurred by new vaccines to a new virus is that the vaccine will put selection pressure on the virus, which will adapt: that is, mutate to a new version that is not controlled by the vaccine. Until the new vaccines are in the field, we won’t know how this novel virus responds to selection pressure. A worst case scenario is that by putting a relatively weak first-generation vaccine into very broad circulation very quickly, we’ll maximize the odds that the virus will adapt—and the world will have to deal with a resistant virus which will have made our new vaccine instantly obsolete.
Scalability is a measure of how difficult it will be to manufacture, distribute and administer each vaccine. There are a welter of factors, including the scale of the scaling needed: it’s very rare to require a new vaccine to be administered to billions of people in a short period of time to respond to a global pandemic, but that’s what’s needed from this one. Here’s a sampling of other factors:
A typical vaccine contains over 1,000 ingredients in addition to the active ingredient—components to stabilize it, amplify its potency, and so forth. Some of these ingredients can be difficult to acquire in sufficient quantities at sufficient speed, and there are, in many cases, few sources.
More mundane things—the special glass and needles used for vaccines, for example—are also in limited supply. The vaccine supply chain is global, with many choke points, and the US has, during the Trump years, pulled back from many of its global relationships, which increases risk now.
The vaccines in development must be kept very cold from the moment they are manufactured until the moment they are delivered. This requires what’s called a “cold chain” of custody to ensure that the vaccine is kept constantly cold as it traverses the globe. One of the vaccine types in development must be kept at -80 degrees Celsius: no cold chain in the world today is capable of this temperature end to end. (The other type can travel at a balmy -20C.) Imagine the logistics of administering this vaccine to people in remote equatorial locations.
Compounding the challenge is that some of the vaccines require two doses several weeks apart, and there’s talk of a booster a few months later. Now imagine that the duration of the vaccine is short, so all of this must be repeated every six months or nine months. If you wanted to give 80% of the world three shots every nine months, that’s 36.5 billion doses/year. Nothing close to that has ever been tried.
Scalability is ultimately limited by trust: how many Americans are willing to be vaccinated once it is offered? Recent polls show that roughly 30% say they probably or certainly won’t get the vaccine, and another 20% will “wait and see.” The combined influence of the anti-vaccination movement and the “don’t trust anything Trump touches” movement is substantial. To achieve herd immunity against this tricky virus, roughly 80% of the population needs to be immune. If 30% of the population refuses vaccination, we’re out of luck before we start. If we assume 70% will be vaccinated and the vaccine itself is only 60% effective, the share of the population with actual immunity would be only 42%, only half-way to the promised land.
Our Unique Bet
Effectiveness, durability, safety, scalability: all of these uncertainties are going to be excruciatingly important to Americans over the coming year, because our country, uniquely, has bet everything on vaccines. Lacking a national plan and vigorous national leadership, we’ve stumbled through partial shutdowns and half-hearted distancing and masking. We’ve failed to set up meaningful testing, tracing, and quarantining. We’ve saved perhaps a million lives, but we’ve lost at least 150,000 unnecessarily, and now we are told it’s going to get worse: many epidemiologists believe that we will suffer a surge this fall and winter substantially worse than the surges last spring and summer.
While we’ve stumbled, the virus has been spreading. IHME, the highly-regarded modeler based at the University of Washington, currently predicts that the daily death rate will more than double between Sept 1 and December 1, and it predicts 410,000 total deaths by January 1, by which date the daily death rate is projected to be about 2,600, on par with the peak death rate last spring, and poised to rise further in January. The vaccines won’t arrive in time or at scale to make much difference until later in 2021.
One scientist sadly remarked, “I guess America doesn’t do NPI.” NPI means “Non-Pharmaceutical Interventions.” It’s jargon for all the masking and testing and other protocols that have been deployed by most other developed nations to keep their cases and deaths far below ours. It seems a fair conclusion at this point. Our polarized politics and commitment to freedom of individual action over collective action has combined with toxic lack of leadership to get us here. But here is where we are, so the best we can do is look forward to how we might realistically improve life in America after the next surge.
Looking Forward: Partial Vaccines, Improvement Cycles
With luck, the pharmaceutical companies will make incremental progress on vaccines after the first rush. Any new vaccine will need to be at least 10% better than those currently in the market to win approval and market share. A speculative road map for vaccine evolution might have these targets:
|1st Generation||2nd Generation||3rd Generation|
|Durability||12 months||2 years||4+ years|
|Safety||90% of population||98% of population||99.9999% of pop.|
|Scalability||700M doses/year||2B doses/year||Globally scalable|
|(Thanks to Fred Brown for estimates)|
If the 2nd generation vaccines meet these goals and earn high public trust, herd immunity might be within reach. It will be much more within reach by the 3rd generation, but that’s a decade away. If we get a decent 1st generation vaccine in the next six months, the coming period of partial vaccines will be gradually become better that what we’ve been living through in 2020. The winter surge will begin to fade and, even without herd immunity, the partial vaccines will lower the risks to health care workers and other frontline workers, and will make the lift easier for other pandemic control measures. Death rates will become lower, the economy will have more potential to operate effectively, and freedom of movement will be improved as the vaccines go global.
Looking Forward: The Cocktail Era
In some cases, we may find that “cocktails” of several vaccines are more effective than any single vaccine, so mixology will become a life science. There will also be cocktails of other treatments, much as there are today for HIV-AIDS. These other treatments may include drugs that disrupt the virus in various ways, improve the performance of the immune system, and suppress the worst symptoms. Fewer people will need to be hospitalized when sick, and fewer will die when hospitalized. Life won’t be perfect, but it will be less fearful. With luck, we’ll be back in the global pool next year, so mixing-and-matching from global resources, not just our own.
Looking Forward: Trace-Free Testing
As mentioned above, “America doesn’t do NPI.” Apart from the polarized mask wars, the biggest sticking point has been our inability to develop and execute an effective program of testing, tracing, and quarantine. Such programs have been the backbone of pandemic control success in places such as South Korea and New Zealand. Here, would-be tracers report little willingness to cooperate, and in any case our use of PCR testing is so slow that the usefulness of the results, sometimes a week delayed, has been termed “garbage” by no less an optimist than Bill Gates. A new approach to testing might save us from ourselves, spare us from the horrors of tracing, and let the economy function incrementally better. The new testing model is to make it ubiquitous, cheap, fast, and self-administered. That’s a tall order, but it’s being hotly pursued, despite uncertainties about accuracy. To its credit, the federal government recently promised to buy a lot these tests, which will serve to speed up development. With such testing available and robustly promoted and supported, businesses might have a way to assure relatively safe workplaces, some kinds of events could happen again, and face-to-face establishments such as retailers, restaurants, hotels, and airlines would have at least a partial path forward. 2021 could be a big year for this kind of testing and the consequent benefits.
Looking Forward: The Long Haulers
This is the darker side of the future. There have been numerous reports of people who have recovered from COVID-19 without recovering. People with a wide array of persistent, even disabling problems that have not gone away: chronic pain, shortness of breath, general weakness, heart inflammation, chronic fatigue, brain fog, organ dysfunction… it’s a long list, and the people suffering – now called “long haulers”—are not just those who were severely ill. Some had asymptomatic or mild cases, and yet ended up in this compromised state. Numbers are hard to come by, but the hospitals in Bergamo, Italy, which was the epicenter of Italy’s terrible spring outbreak, are now calling back every patient that survived for a follow-up exam. They report that 50% of these patients are still suffering health consequences that appear to have come from their COVID-19 illnesses 5-6 months ago. The US, with so many more cases per capita than other major developed nations, could have a significant shadow over its economy and health care system for some time to come.
Looking Forward: Mask On, Mask Off.
Masks are annoying. They muffle speech, fog glasses, hide smiles, make breathing less comfortable, and don’t accommodate eating and drinking very well at all. However, in places where mask-wearing has been sustained at high levels, two things have happened: rates of transmission have been reduced, and, it appears, average severity of cases has been reduced. Masks alone can’t control the virus, but masks everywhere can dampen the consequences significantly. Having partial vaccines, treatment cocktails and ubiquitous testing doesn’t do away with the need for masks, distancing, and hand-washing. Instead, the combination becomes more effective. In a world where the virus has been tempered by a partial vaccine, mask-wearing can do proportionally more good in reducing transmission. Some parts of the world now have such good control of the virus that mask-wearing has become something for special occasions—indoor groups, for example. The US is not one of those parts of the world. Here, masks will continue to make a big positive difference for the foreseeable future.
We are heading for quite a four-juggernaut convergence over the next few months. There is, of course, an election coming. The possible appointment to the Supreme Court of a replacement for the late Justice Ginsburg is roiling the country. There’s very likely to be a worst-yet wave of COVID-19 cases and deaths coming. And the pharmaceutical companies with vaccines in Phase III trials are eager to announce something positive about how the trials are going.
One company, Pfizer, has already said it “could know” by the end of October if its vaccine is working. If any candidate seems to have legs before Election Day, the President is widely expected to short-circuit the FDA approval process and sign an “Emergency Use Authorization” to permit the process of vaccine deployment to begin as soon as possible, as he has done with hydroxychloroquine and convalescent plasma.
It could be an election bombshell, as well as a logistical nightmare. The pharmaceutical companies are highly motivated to try to give the president what he wants, because having the president override the usual FDA process is an enticing carrot—it gets them to market faster, lets them gobble up scarce vaccination process resources first, gives them an opportunity to claim large market share (and share dominance has historically has proven to be long-lasting), and it gives them exceptional protection from liability in case the vaccine is found to be harmful when scaled up.
Their deal with the US government is for just two years, after which they are free to sell their vaccines on the open market to the highest bidder. That might not be the best thing for beating the pandemic, but it will certainly reward the shareholders of the companies with the strongest grip on the early brass ring. Between November 3 and January 21, we could see a terrible scramble for power as Trump tries to hang on with charges of election fraud and Biden fights to get mailed-in ballots properly counted, while at the same time COVID-19 deaths are climbing to new heights and the results of the Phase III trials are arriving. Only in America.
It’s also possible that none of the vaccines currently in Phase III will survive. In that case, the US will face a bleak future: we’ll have made our big Live Free or Die bet on the vaccine, the pandemic will be back worse than ever, and we’ll have no vaccine and no functioning alternative to fall back on. If the surge in cases and deaths materializes as predicted, and no vaccine is approved, or one is approved but isn’t very effective or durable, the US will face a worst-in-the-world virus challenge in 2021, in which deaths could be much greater than in 2020, and we may have no choice but to start over with the shutdown we never finished last spring, shut down much harder, and stay shut down until we can, at last, develop a ubiquitous testing program and other protocols to regain control of the virus, and get a good vaccine deployed at some uncertain future date. This is a terrible scenario, so let’s hope the vaccines now in trial are good enough. We’ve burned our bridges and burned our boats; let’s not set the water on fire too.
The Bottom Line
As I write this on September 19th, the United States is approaching 200,000 confirmed deaths, a number growing at a rate of nearly 1,000/day. If we had done nothing to prevent the virus from sweeping across the country, that number could be over a million now, so our sacrifice has not been for nothing. However, if we’d managed this as well as, say, Germany, we’d be mourning about 37,000 deaths now, growing at a rate of about 15 a day. If we’d done it as well as South Korea, which had its first cases the same day we did, we’d have had, by now, only 2,432 deaths, growing at a rate of about 23 a day. As we head into the fall and winter, the virus is out of control in the United States, and yet our president and his party are committed to pretending that the problem is mostly behind us, that they’ve handled it well, and that in any case a vaccine is about to save us. We are likely to see another 200,000 people die in the last few months of this year and perhaps 500,000 next year as the price of those fictions. Meanwhile, South Korea expects to have its current “outbreak” of 3.6 deaths/day (it peaked at 5!) under control in a matter of days.
Excellent analysis and a needed dose of reality for those who still refuse to comprehend the challenges we face. We need to accept masks as a wardrobe essential.
Probably the most concerning aspect, out of the individual’s control, is the chance of never getting over it, becoming a “long hauler,” not to mention the possible load that will put on health services. Seems likely that Covid 19 survivors will be deemed to have a pre-existing condition.
Dick Lilly, as Ed Yong points out in today’s Atlantic, the multifarious consequences of this virus are more visible to us because of the sheer volume of cases. Many other viruses seem to have effects like this, but they aren’t looked for or studied, either because there haven’t been enough cases to bring them to our attention, or because the virus doesn’t kill enough people to require the kind of diagnostic efforts required to see things such as heart inflammation. Worth reading, as it everything Ed writes for the Atlantic on this issue. https://www.theatlantic.com/health/archive/2020/09/covid-19-heart-pandemic-coronavirus-myocarditis/616420/