Statement on FDA’s proposal for once a year COVID vaccination

Public Comment by Rob Wallace, PhD, submitted to the FDA’s Vaccines and Related Biological Products Advisory Committee regarding the future vaccination regimens addressing COVID-19.

I’m an evolutionary epidemiologist. I mapped statistical phylogeographies, including molecular evolutionary traces, for HPAI H5N1 under CDC advisor Walter Fitch at UC Irvine ^1-2. I have consulted for the CDC and am commenting here on behalf of the People’s CDC³.

An underlying supposition of the draft VRBPAC recommendations concerning changing the composition and schedule of COVID-19 vaccination is a putative analogy between SARS-CoV-2 and influenza. Metaphors are necessary, but often problematic. Comparing two problems does not make them the same, a distinction well-explored in science⁴ . The FDA Briefing Document on the topic pings back and forth between acknowledging the fallacy and indulging it.

I will speak on a technical issue here, but will note in passing the present administration’s efforts to end the pandemic by fiat than by deed, from the May 2021 recommendation that the vaccinated needn’t mask to the Impact Research memo⁵. These documented efforts place the present matter in a broader political economy. In framing its response to COVID-19 on the influenza model, the advisory committee risks turning the cause, the virus, into something we wish it was: just like influenza.

Influenza vaccination–as half-ass as it is–works in part because we can predict which strains will seed the next year’s pandemic⁶. Fitch and colleagues showed those H3 strains most evolved at 18 codons almost always served as the progenitor of the next year’s pandemic^7-9.

There’s been some work in this direction on SARS-2, but even under optimistic modeling, the present population parameters do not place us near the kind of endemicity associated with seasonal influenza^10-12. Punctuated equilibria on a schedule of the virus’s own making, not that of industrial production or any other broader objective, are the evolutionary order of the day. And we do not have a handle on that schedule, save perhaps, as recent work showed, SARS-2 variants are spreading faster than before upon emergence and the speed of the arrival of various Omicron subvariants was not deterred by the level of population vaccination¹³.

The implications are twofold. Constraining vaccination to once a year would only further handcuff a public health response already limited by an abandonment of nonpharmaceutical interventions. At the clinical level, vaccination alone offers better protection but no guarantee against reinfection and Long COVID^14,15. At the evolutionary level, models by Okamoto et al. show losing NPI likely selects for changes in SARS-2’s reproductive strategy and the emergence of vaccine-resistant strains^16,17.

In short, prevention, keeping people from getting infected in the first place, remains a critical contribution to preserving vaccine effectiveness. In the other direction, reducing the degrees of freedom we have in the types and scheduling of vaccine production only compounds the public health damage dropping NPI campaigns such as mask mandates and contact tracing produces¹⁸.

Secondly, the rationale here behind simplifying the vaccine schedule is all wrong. It isn’t the present scheduling that keeps booster coverage at a little over 16% of eligible Americans¹⁹. It’s the insistence in declaring a still rapidly evolving pandemic over and ending all those programs that would have helped vaccination programs succeed. The success of federal pharmaceutical campaigns depends on the NPI campaigns going door-to-door that two administrations running dropped. Do not operationalize an intervention that’s based on blaming unvaccinated Americans for that fiasco.