Self & Other · mRNA, and What COVID Taught Us
What the Speed Was
The immune system, doing what it has done since Chapter 1, sees a not-self protein and responds. Dendritic cells gather fragments and travel to lymph nodes. T cells with receptors that fit the spike protein activate. B cells with antibodies against the spike protein activate, divide, and start producing antibodies. After a couple of weeks, the body has memory cells specific to the spike protein.
The mRNA itself, having done its job, is broken down within hours. The cells that briefly made the spike protein stop making it within days. The lipid nanoparticles dissolve. What remains is the immune memory — the same kind of memory we discussed in Chapter 4, produced through the same kind of process that vaccines have always used. The novelty is in the delivery mechanism. The biological response is the body doing what bodies always do.
This is worth being clear about because some of the early misinformation about mRNA vaccines suggested they did something fundamentally new to the body — altering DNA, lingering indefinitely, integrating into the genome. None of this is what happens. mRNA does not enter the cell nucleus, where DNA is kept; it works in the cytoplasm, the cellular space outside the nucleus. The cells that make spike protein from the vaccine make it for a few days and then stop. The vaccinated person, after a few weeks, has the same kind of immune memory they would have from any other successful vaccination. Their cells, their DNA, and their long-term biology are unchanged.
What the Speed Was
The COVID vaccines went from genetic sequence to emergency use authorization in about ten months. This is, by far, the fastest a vaccine has ever been developed. The previous record — mumps, in the 1960s — was about four years.
The speed has been one of the most contested aspects of the vaccines. Some people, knowing how long vaccine development normally takes, concluded that something must have been skipped. The actual story is the opposite. Almost nothing was skipped. What changed was that several normally sequential processes happened in parallel, several normally slow processes happened with extraordinary financial and human investment, and — most importantly — several decades of foundational work that would normally have to happen during vaccine development had already been done.
Karikó and Weissman’s work on modified mRNA dated from 2005. The lipid nanoparticle delivery technology dated from a decade of refinement before that. The ability to synthesize mRNA at industrial scale had been built out gradually over the 2010s. The basic science of coronavirus spike proteins had been studied since the SARS outbreak in 2002 and 2003 and the MERS outbreak in 2012. Vaccine candidates targeting coronavirus spike proteins had already been designed and tested in animals before COVID-19 emerged. When the pandemic began, what would normally be the first several years of vaccine development — figuring out the platform, identifying the right target, learning how to deliver it — was already done.
What still had to happen was: design the specific mRNA, manufacture it, run human safety trials, run efficacy trials, and get regulatory approval. These are the irreducible steps. They normally take five to ten years because they are run sequentially — each step waiting for the previous step to complete — and because vaccine candidates compete for limited resources. In 2020, with billions of dollars committed up front and several governments coordinating at scale, the steps were run in overlap. Manufacturing was scaled up while trials were still running, on the bet that the trials would succeed. (When trials of unsuccessful vaccines came back, the manufactured doses were thrown away. The bet, in some cases, was on losing.) Phase 1, 2, and 3 trials, normally separated by months or years, were run in close succession on candidates that were already showing promise. Regulatory review was conducted on rolling submissions rather than waiting for final reports.
The trials themselves were, by historical standards, large — the Pfizer/BioNTech and Moderna Phase 3 trials each enrolled tens of thousands of participants, randomized to vaccine or placebo, followed for months. The endpoint was symptomatic COVID infection. The initial efficacy data, when it came in late 2020, was striking: about ninety-five percent reduction in symptomatic cases among the vaccinated compared to the placebo group, with severe cases reduced even more dramatically. The vaccines were authorized for emergency use in December 2020.
Whatever else can be said about the COVID vaccines, the speed was not, in fact, the result of cutting corners. It was the result of decades of preparatory science meeting an emergency that finally justified spending what it cost to deploy what was ready.
What the Vaccines Did, and Didn’t
The COVID vaccines work, as the data has now made clear over years of follow-up. They are not perfect, and the public conversation around them has often expected them to be perfect in ways that no vaccine ever has been. A more careful accounting is in order.
The vaccines, in the original 2020 trials and in real-world data through 2021, were extraordinarily effective at preventing infection with the original strains of the virus. Efficacy was around ninety to ninety-five percent against symptomatic disease for the first several months after the second dose.
This effectiveness against infection began to decline in 2021 as the virus mutated. The Delta variant reduced effectiveness against symptomatic infection somewhat. The Omicron variant, emerging in late 2021, reduced it substantially — vaccinated people, even those who had received boosters, could and did get infected with Omicron, often without severe illness, but contagiously.