Don't worry, I can still count. The title of this post comes from one of my teaching aphorisms to remind medical providers that evidence-based medicine isn't some magical entity that can be invoked as a gold standard. In fact, personal opinion is a form of evidence, the lowest form of evidence because it is most subject to error due to known or unknown (explicit or implicit) personal bias. When I see a publication with "evidence-based" in the title I'm immediately wary of the content.
This past week was very busy for me, mostly because of the FDA VRBPAC meeting and its ramifications, all dealing with COVID-19. But first, I wanted to cover one topic unrelated to covid.
Our Shaky Relationship With Animals
Last week CDC published a great Surveillance Summary about enteric disease outbreaks associated with animals. This excludes foodborne outbreaks related to animal products; it deals with direct animal contact so won't help you with your Memorial Day weekend picnics.
The report covers the years 2009 - 2021, so mostly pre-pandemic. The numbers of illnesses and outbreaks didn't show a striking decline in 2020 and 2021.
You won't be surprised that Salmonella was the most common identified pathogen.
I found it very interesting that exposure in private homes was so prominently represented.
The animal type list afforded me a quick lesson in nomenclature - now I know the differences among various reptile categories.
Covid and Vaccines
I spent about 7 hours on Thursday watching, listening, and taking copious notes during the FDA VRBPAC meeting to recommend a strain for the new iteration of covid vaccines. Presentations from FDA, CDC, WHO, and the 3 US covid vaccine manufacturers took up the bulk of the meeting, followed by extensive discussions. The meeting was excellent for the most part, as always I learned a lot. However, a publication from 2 days earlier clearly cast a pall over the deliberations.
The publication, by FDA CBER director Dr. Vinay Prasad and FDA Commissioner Dr. Marty Makary was entitled "An Evidence-Based Approach to Covid-19 Vaccination," ergo the title of this post. It spelled out new FDA criteria for vaccine approval.
It wasn't clear whether just a change in vaccine composition would mandate new trials; influenza vaccine changes strain composition every year without necessitating new trials. Also uncertain was whether the existing vaccine would be available to those outside of the 65+ age group and the younger individuals with at least one risk factor. No mention of any allowance for the common occurrence about family visiting grandma with many risk factors, presumable you can either not visit or let grandma take her chances.
In his opening remarks to the VRBPAC, Dr. Prasad seemed to backtrack a little from the pronouncement, implying that this wasn't a final document and could be modified by further input. As a result, this is now clear as mud.
The appearance of the article just days before the VRBPAC meeting only highlighted how it changed the landscape for FDA regulation of covid vaccines. Here's the usual process for strain selection as shown in one FDA slide from the meeting:
This process occurs continuously at FDA, not just for the VRBPAC meetings. The Prasad/Makarty publication circumvents this process. It was clear from all of the meeting discussion that committee members were concerned that a change in vaccine composition now might trigger requirements for new clinical trials, delaying availability of an updated vaccine or, worst case scenario, being so expensive and difficult that vaccine manufacturers would abandon updating the vaccine.
Here's how one of my many information sources, Dr. Katelyn Jetelina of Your Local Epidemiologist, pictured the process change:
Covid vaccines are the most studied and monitored in the history of vaccinology, all starting with randomized, double-blind, placebo controlled randomized trials. Here's an example from the Pfizer presentation.
I did say at the top that I learned a lot from the meeting, so let's focus on those points. Note that some of the images have web links, but they aren't active in this post.
The #1 take home point from the entire day's meeting, which I didn't hear emphasized in various news reports, is that we are in a really good place for figuring out covid vaccine strain composition for the immediate future! That's because not much has changed, antigenically speaking, with circulating strains in the past year. Everything in the US so far this year is a close cousin of the JN.1 lineage.
In fact, the number of spike protein mutations (the target of all covid vaccines) has been relatively small since early 2024.
What this means antigenically is that, with only minor differences, sera from recipients of last fall's vaccine (mostly a KP.2 strain that is in the JN.1 lineage) neutralizes the newer variants quite well.
Not surprisingly, there are limitations to this information. First, we still don't know what level of neutralizing antibody confers protection against more severe covid illness, so we don't really have a target in mind. Second, the assays themselves aren't standardized across platforms, so we can't compare one lab's results directly with another's. Third, we have virtually no data regarding cell-mediated immune responses - these tests require a lot of blood and are time consuming. Fourth, this is all test tube and animal data, what we really want to know is how it works in the real world. Unfortunately we can't know that until after the vaccines are used widely.
Which brings us to the fifth limitation, which is mostly a good thing. We don't have as many data points to analyze as were available at the height of the pandemic. The good part of this is that we have really do have fewer people with serious infections. The bad side is that at the same time monitoring resources have scaled back considerably, including at the CDC as verified by some at the meeting. Also, individuals appear to be less likely to seek testing when they are ill. The smaller the sample size, the less robust the predictive power for the future.
In particular this is problematic in the pediatric population, a combination of fewer complications (a good thing) coupled with low vaccination rates. Thus, CDC could not provide any estimates of covid vaccine effectiveness in the pediatric population for the past year. Here's the conclusion statement about VE as is currently known:
A bit about seasonality and other covid epidemiology. It was clear during the meeting that terminology used in the past for influenza is infiltrating the world of SARS-CoV-2, particularly with the terms antigenic "drift" and "shift." The former is a minor change in antigenic properties in circulating strains, as we experience most years with flu and currently with covid; the latter is a more extreme change such as with the original Wuhan strain followed by delta and then by omicron.
Covid still doesn't have a predictable seasonality compared to influenza. We've seen recent trends towards surges in late summer/early fall and in winter, but covid never goes away.
Even with the lower levels recently, covid is still a tremendous problem.
Most hospitalizations (and deaths) are in the oldest age groups.
To the FDA's credit, they moved quickly to make a final statement about covid vaccine composition: "...the COVID-19 vaccines for use in the United States beginning in fall 2025 should be monovalent JN.1-lineage-based COVID-19 vaccines (2025-2026 Formula), preferentially using the LP.8.1 strain." In my opinion this accurately represents the consensus of discussion at the meeting. It remains to be seen what the 3 US vaccine manufacturers will do with this information given the vague requirements for new studies if a new variant is used. The mRNA vaccines from Prizer and Moderna can be produced quickly, likely available late summer. Novavax, which does not use the mRNA platform, takes a bit longer, plus FDA has placed new study requirements for them for the current vaccine that might require more studies. The Novavax representatives at the meeting did express confidence they could have vaccine ready by fall.
No crystal ball:
The big picture is that whether we use the existing KP.2-based vaccines going forward or switch to the LP.8.1 variant isn't a big deal - we don't have enough data to suggest one is going to be substantially more effective than the other.
I won't ramble on further about the VRBPAC meeting, though I realize I've used less than 10% of my meeting notes. As always, feel free to ask for more details if you're interested.
The MAHA Report was a HAHA
Or maybe BOOHOO is more appropriate. At 73 pages and over 500 footnotes, one would think there would be some substance to the report, an assessment intended to address chronic disease in children. It's a laudatory goal, but the conclusions seemed to have been made prior to writing the document. Snippets of medical publications were cited but clearly lacked context and many times seemed deliberately misleading to support the conclusions. The apparent authors were the members of the President's Make America Health Again Commission, consisting of cabinet members and other political appointees. Only two of the 14 members had medical backgrounds: the FDA Commissioner and the NIH Director. It most certainly was not a high quality scientific report, perhaps at a middle school student's level of understanding of data analysis. It is a political diatribe, not a balanced assessment of the literature.
The Commission also is charged with submitting a strategy based on this assessment, due in a little less than 3 months. I was saddened to read the assessment and won't comment further.
Memorial Day
It's a lovely holiday day in Maryland, take the opportunity to celebrate with family and friends. I recall that the poem most linked to this holiday, In Flanders Fields, was written by a physician.
