Andy Larsen: Why the data shows we need booster shots, and what comes next

U.S. is poised to begin administering next month.

(Francisco Kjolseth | The Salt Lake Tribune) People are vaccinated against COVID-19 at the former Provo High School gym at 1125 N. University Ave. on Wednesday, July 28, 2021.

Editor’s note This story is available to Salt Lake Tribune subscribers only. Thank you for supporting local journalism.

Let’s talk about booster shots.

For a while, I was relatively optimistic that most folks wouldn’t need any doses of the COVID-19 vaccine behind the one or two they initially received. Why was I optimistic, beyond my usual sunny disposition? Well, for a couple of reasons: First, those who were infected in the original SARS outbreak in 2003 still had antibodies 12 years later. Second, those who were infected with the new coronavirus also tended to keep antibody levels pretty high for months after their recovery. Early results on vaccine antibody levels were promising, too.

But the delta variant has changed the game. Due to its worldwide spread and high virulency, we’re seeing vaccine efficacy decrease month over month, and it’s that decrease that has caused the opinions of experts and policymakers to shift.

In case you missed it: The Centers for Disease Control and Prevention announced Wednesday that it will offer booster shots to people starting Sept. 20, or eight months after they received their second shot, whichever comes later.

What led the CDC to make this decision? Let’s dig in to what it was seeing.

Numerous studies showing efficacy decline

First came Israel’s studies. As one of the first countries to get a significant portion of the population vaccinated, it was in an ideal position to evaluate the efficacy of the vaccines. But as delta came, Israel’s cases spiked — and it noticed a trend in who was more likely to be infected.

An initial study found that those who were vaccinated in January or February were 53% more likely to be infected by COVID than those who were vaccinated in March or April. But a second look, weeks later, found even more significant changes in vaccine efficacy:

From the Israel Ministry of Health, on coronavirus cases and hospitalizations from June 20 to July 17.

Those who got their second dose in January were just 16% protected against infection by the delta variant, while those who got it in April were 75% protected. Meanwhile, protection against hospitalization remained quite high: 82% for the vaccinated in January group, compared to 88% overall.

Obviously, that’s just one country, with a relatively small sample size. But in the past week, we’ve now gotten multiple studies with literally millions of participants to prove definitively that vaccine efficacy is decreasing.

Let’s first look at one from the United Kingdom. It examined 3 million PCR tests from British people and found a no-question decrease in the Pfizer vaccine’s efficacy over time.

The U.K. study didn’t look at hospitalizations but a U.S. study did. In particular, a CDC study released this week from New York state data also found declining vaccine efficacy — even more than you’d expect from just the introduction of the delta variant alone — but relatively stable hospitalizations.

That was verified by a second CDC study (in which Intermountain Medical Center in Murray took part, by the way) that showed the vaccine’s efficacy in preventing hospitalization has stayed pretty high — except for the immunocompromised.

Data from 21 hospitals shows quite good and continued vaccine efficacy against hospitalization — except that the immunocompromised have seen more significant losses. (https://www.cdc.gov/mmwr/volumes/70/wr/mm7034e2.htm)

Note that the Israel and U.K. data is all for the Pfizer vaccine, while the CDC data doesn’t specify if the vaccine belongs to Pfizer or Moderna. There actually are some indications that Moderna’s vaccine is doing better than Pfizer’s at lasting longer, as antibody levels and vaccine efficacy stay higher for those vaccinated with Moderna when compared head-to-head. That’s potentially due to a larger initial dose size (Pfizer’s dose is 30mcg, while Moderna’s is 100mcg) but we’re not positive that’s why.

How much better is Moderna than Pfizer, and will the “Mod Squad” need booster shots, too? And how about those with Johnson & Johnson’s vaccine? That’s yet to be decided on a large scale. Watch for more research and direction to come here.

Why the decline?

So what’s going on? Why is vaccine efficacy declining at all, and why do the vaccines seem to last longer in preventing delta hospitalization than delta infection?

Former White House COVID-19 adviser Andy Slavitt wraps up one hypothesis here. Essentially, the vaccines teach the body about the shape of the coronavirus particle, telling your immune system to be on the lookout for a spike protein that looks like the one on the coronavirus. That triggers two processes: the creation of antibodies, which deal with infections immediately, and the creation of memory T and B cells, which act as the blueprint for creating more antibodies in the future.

But your body can’t keep huge numbers of antibodies against every intruder forever, and there’s no way to get your body’s immune system to read the newspaper and understand that the coronavirus in particular remains a serious ongoing threat. So antibody levels naturally decline somewhat.

That wasn’t a big deal against the old version of the coronavirus, but the delta version is a lot more virulent. In particular, it creates viral loads that are 1,260 times larger than those that were created by the original form of the coronavirus. And because those loads are so large, they may overwhelm the remaining antibodies — making you sick.

The good news is, because vaccinated people have those memory T and B cells, they have the blueprints to make more antibodies at a moment’s notice. Once your body does that, the antibodies can usually, but not always, overwhelm the infection. Meanwhile, unvaccinated people are still in uncharted waters, at higher risk of drowning in the virus.

Do boosters help?

Israel, looking at its declining vaccine efficacy and spiking cases, decided to give older citizens — at first 60 and over, but now expanded to 50-year-olds, too — booster shots, right at the end of July. Impressively, the country has already inoculated over a million people in that time.

The results? So far, so good.

Here are two graphs, with data from Israeli scientists Barak Raveh and Dvir Aran. The first shows the vaccine’s efficacy over the past couple of weeks by age group, and the second shows the number of coronavirus cases per 100,000 people by how many doses of the vaccine they’ve gotten.

Graph 1 shows vaccine efficacy against infections by age group in Israel in recent weeks. Graph 2 shows how those who have received three doses are seeing fewer cases than those with two doses or the unvaccinated. (https://twitter.com/EricTopol/status/1427342828921769998)

Again, this is early days. But with booster shots scheduled to start Sept. 20 in the U.S., we have plenty of time to pull the plug if the concept begins to fail elsewhere.

Are we going to have to boost forever?

Now I know what you’re thinking: Isn’t the efficacy of the third dose going to go down, too? Are we going to need to boost ourselves every few months forever?

There’s data in both directions. Here’s a slide from Pfizer, which tends to point to the third boost being sufficient.

Pfizer's in-house data on antibody levels generated by their vaccine, in 18-55 year-olds and 65-85 year-olds. The bars, representing antibody levels, decrease in the first month after the second dose. But after the third dose, they actually increase in the ensuing month. Does this indicate a longer-lasting immune response?

There’s a lot going on in that graph, most of which you can ignore. But look at the bars in each graph in the “Post Dose 2″ columns: See how they’re declining from day seven, to one month after, to eight months after? Antibody levels decline. Now look at the bars in the “Post Dose 3″ columns. See how the antibody levels actually increase a bit from day seven to the end of the first month? That’s giving some hope that the antibodies might be longer lasting.

The theory is that, essentially, the first two doses were placed close enough together that they didn’t trigger much of a secondary immune response that might be longer lasting.

Indeed, there are vaccine dosing sequences that we do right now that are three doses, like those for hepatitis B, or even polio. You just haven’t thought about those vaccines since you were a baby, and, well, you probably weren’t thinking about vaccines. Three doses isn’t crazy.

On the other hand, the U.K. study referenced above explicitly looked at whether dosing timing mattered. In particular, some people in Britain were dosed on the normal three- to four-week schedule, and others had two to three months in between their first two doses. Researchers found zero difference in efficacy over time between the two groups. If two to three months weren’t enough to trigger this secondary immune response, do we know that eight months would be enough?

There’s enough uncertainty that the World Health Organization actually argues against booster shots. Its argument, and it’s a reasonable one, begins with the fact that most of the world remains totally unvaccinated. While vaccines are a limited resource, WHO says these doses should go to poorer countries to protect those people against hospitalization, rather than going to rich countries to protect their people from infection. This also has the side benefit of reducing the likelihood that worse variants will develop, since it’s thought likely that variants are more likely to originate in long-lasting, severe illnesses.

This U.S. administration, though, chose to embrace the booster, making it available to Americans in a month’s time. Will the vaccinated embrace it back? That, like so much else with the virus, is still to be determined.

Andy Larsen, one of The Salt Lake Tribune’s Utah Jazz beat writers, doubles as a data columnist. You can reach him at alarsen@sltrib.com.