Is a call for more vaccine safety studies an admission that vaccines are unsafe?
A recent perspective piece from some giants in the field is being weaponized for disinformation.
A recent perspective published in the New England Journal of Medicine is being misrepresented by anti-vaccine lobbyist Aaron Siri to claim vindication regarding the lobby’s prior claims of vaccine safety. The perspective piece is paywalled, but I have uploaded it below for anyone to read it if they so desire (despite appearing in a medical journal, it’s not particularly technical) so that they can be confident that I am not misrepresenting what is being stated (though I do intend to provide additional context behind some of the words which I think may be helpful). In this post, I address Siri’s salient claims, as well as offer my thoughts on this perspective.
For those in a hurry, here’s my summary of their key points:
Vaccine hesitancy during the COVID-19 pandemic is a serious problem and suggests the public wants deeper insight into vaccine safety than what has been traditionally available.
Vaccine safety research takes a lot of time to answer important questions, but it offers important insights into how to make new vaccines safer and establish groups who should not get vaccinated for medical considerations.
We lack an understanding of the biology underlying a lot of important adverse events from vaccination.
The National Vaccine Injury Compensation Program (NVICP) is paid for with an excise tax on vaccines and currently has a budget surplus. Using this budget to help fund studies on vaccine safety would be a prudent use of resources that would collectively benefit everyone and also leave total budget unaffected.
Re: Siri’s Claims
You can read Siri’s comments here. I think most can be addressed by simply reading the perspective piece honestly as it’s very clear they are distortions, but I do wish to highlight some blatantly dishonest points.
To begin with, Siri remarks about Stanley Plotkin’s deposition by him in a trial between parents who disagreed about whether or not the child in question should be vaccinated. Stanley Plotkin truly is a giant in the field of vaccines, having invented the rubella vaccine still in use today, worked on a number of other vaccines, and he is one of the editors of the Vaccines textbook that I consider to be an indispensable reference. During this deposition, Plotkin was grilled by Siri on a diverse panoply of subjects relating to vaccines and Plotkin described the experience as being exhausting. As the deposition was recorded, it has been exploited many times over by the anti-vaccine lobby, especially Children’s Health Defense and the ICAN network, to claim an admission that there isn’t proof that vaccines do not cause autism. The ironic part her perhaps is that despite the traction the circulated clips can be said to have gotten, even with the legerdemain of dishonest editing, Siri’s client lost the case. As Dr. Vince Ianelli explains in the hyperlinked article from Vaxopedia, what in fact was said was a comment about the basic fact that in empirical science (as distinct from something that can be described purely theoretically, like math), one cannot definitively prove a negative. However, as Prof. Plotkin explains within that same deposition, as a physician, it is his job to consider the totality of evidence, and when multiple very large, high-quality studies consistently fail to demonstrate any link between vaccination and autism, an intellectually honest person would have to admit that there is no link between the two. Regrettably, when it comes to anti-vaccine lobbyists we are not dealing with intellectually honest people. Nevertheless, Plotkin, after this deposition, sent an email to members of the American Academy of Pediatrics’ Committee on Infectious Diseases and Amanda Cohn of the CDC’s then liaison to the committee, in which he notes that the anti-vaccine lobby is highly organized and well-funded and that to address the specter of vaccine hesitancy, those involved in the delivery of vaccines must respond accordingly. I don’t think anyone who has been involved even peripherally in vaccination could reasonably argue that Stanley Plotkin was at all off-base in his comments.
It should be noted that many of the questions asked of Professor Plotkin are fundamentally bad-faith and impossible to answer honestly in the manner they were posed. For example, Plotkin was asked to comment on the fact that some vaccines are made in cell lines from monkey kidneys and therefore the final vaccine product may contain monkey DNA. The fact of the matter is that as part of the production process, vaccines are treated with enzymes called DNases which would destroy any residual DNA impurities that might be present in the products into tiny fragments. Yes, some residual DNA is present in all biologic products including vaccines, you can’t get rid of it completely, but calling it “monkey DNA” is itself misleading because at the size of the fragments in question it cannot even be identified as “monkey DNA.” To use an analogy: consider for a moment a line from your favorite work of literature. If you hear that line, you can identify that it came from that work of literature. What if you were presented with a single word? Or a single letter? Could you still recognize it as being from that piece of literature? The answer is no- there would be no way to show that the word was taken from that source of literature, so why would we say that it is that literature’s word?
A core premise of Siri’s screed is that this perspective piece is itself an admission that vaccines are unsafe, as the anti-vaccine lobby has been arguing since its inception with smallpox vaccination. I think you’ll find that such an admission exists solely in Siri’s imagination and appears nowhere in the piece. There is in fact mention that the literature on vaccine safety is extensive, even noting that by the time of the IOM’s report on vaccine safety, over a dozen studies had been done to look for an association between vaccines and autism. The perspective does note that it takes too long to do these studies and get them published, and suggests that an accessible funding source for these studies would help. The closest thing you will find to an admission that vaccines are unsafe is a comment in the very beginning that the public’s perception of vaccine safety has been skewed by the number of rare adverse events they hear about, in that when you deliver a vaccine to billions of people (as was done through the COVID-19 pandemic), something that happens to one person out of every few hundred thousand or so will still end up affecting thousands of people, even though is is vanishingly unlikely to occur to any given person. These things are not in contradiction and speak to the main point of the perspective: we should strive to better understand these rare adverse events so that we can avoid them.
Siri also argues that there are no proper safety studies of vaccines. The idea behind this is the argument that the only way to know whether or not a vaccine is truly safe is to test it against saline placebo and to follow an inordinate amount of people for an inordinately long period of time, and we should repeat this for every new vaccine we add to the schedule. Per this claim, no vaccine has ever been subject to such a proper study and therefore the entire basis for all vaccine recommendations lacks basis. The problem with this claim, of course, is that it is unvarnished rubbish. There are detailed guidances for the use of placebo in vaccine trials from the WHO, and the rules are pretty simple:
Placebo use in vaccine trials is clearly acceptable when (a) no efficacious and safe vaccine exists and (b) the vaccine under consideration is intended to benefit the population in which the vaccine is to be tested. In this situation, a placebo-controlled trial addresses the locally relevant question regarding the extent to which the new vaccine is better than nothing, and participants in the placebo arm of the trial are not deprived of the clinical benefits of an existing efficacious vaccine.
Placebo use in vaccine trials is clearly unacceptable when (a) a highly efficacious and safe vaccine exists and is currently accessible in the public health system of the country in which the trial is planned and (b) the risks to participants of delaying or foregoing the available vaccine cannot be adequately minimized or mitigated (e.g. by providing counselling and education on behavioural disease prevention strategies, or ensuring adequate treatment for the condition under study to prevent serious harm). In this situation, a placebo-controlled trial would not address a question that is relevant in the local context, namely how the new vaccine compares to the one that is currently in use, and participants would be exposed to unacceptable levels of risk from delaying or foregoing a safe and effective vaccine that is accessible through the public health system.
The basis for this recommendation goes back to a very simple principle of biomedical ethics: it is never ethical to withhold the standard of care. You cannot knowingly put people in harm’s way by depriving them of a vaccine that works because you want to see whether or not a new vaccine works. If no vaccine exists, there can be genuine uncertainty (known as equipoise) about whether or not it is helpful or harmful, and thus use of a placebo is reasonable. Placebo use, however, is not a requirement to understand vaccine safety. For one thing, at the point that a next-generation vaccine is available, the relevant question is not “Is this product safer than nothing?” but rather “Does this product have a safety profile that is as good as the previous generation vaccine?” It doesn’t make sense to license a vaccine that is inferior to the previous version.
Another factor regarding placebo use has to do with what constitutes a placebo. The definition of a placebo is “an inert comparator drug that lacks the pharmacological effect in question, thus enabling detection of the pharmacological effect of the study drug via the signal of placebo–drug differences in end-point parameters.” Put another way, in the context of a vaccine study, a placebo is a substance that does not induce the immune response that a vaccine induces. Among the anti-vaccine lobby, saline is the only valid placebo, but in fact in the context of a vaccine trial, saline probably isn’t the best choice for placebo. This is because another aspect to vaccine clinical trials is blinding: to see whether a vaccine works you don’t want your clinical trial participants modifying their behavior based on whether or not they got the vaccine, as this could skew results. A saline injection will feel quite different from a vaccine injection because it will not induce the same local or systemic symptoms that a vaccine might. Thus it is not unusual for vaccine studies to use everything other than the antigen for their placebo, or in some cases, use a licensed vaccine against an unrelated disease. Hopefully comparison to an unrelated, licensed vaccine makes sense at this point: it’s a vaccine that has already passed muster and been shown to be safe, and so the new vaccine should have a safety profile that is comparable to it to be acceptable. In the case of using everything other than the vaccine antigen (including things like adjuvant), this achieves the stated goal of the placebo- the measurement of the pharmacological effect of the study drug, and can help preserve blinding because the adjuvants may be locally irritating. Additionally, in some cases, adjuvants can offer some level of nonspecific protection against infection, which theoretically can help protect your placebo group, further enhancing the ethical basis for their use in vaccine trials.
Concerning the duration of study within the clinical trial period: an extreme example of this is sometimes taken as a comment that participants in a study for a hepatitis B vaccine were followed for safety only for a period of 4-5 days. If this were true, it would of course be unacceptable. Beyond that this won’t tell you enough about safety, this isn’t even enough time to tell whether or not a vaccine works as an immune response will only be in its earliest stages at this point. Fortunately, this isn’t actually what happens in vaccine studies and is seemingly based on misreading or misrepresenting the package inserts for vaccines. When a vaccine is studied in a trial, safety monitoring comprises a period of solicited adverse events and unsolicited adverse events. Because the vaccine induces an immune response, this typically means people will experience local symptoms, such as injection site soreness, redness, swelling, etc. and systemic symptoms such as fever and joint aches. It is expected that these symptoms be short-lived, i.e., that they resolve within a period of 4-5 days at the latest (although in most cases they are done well before that). In addition to these solicited adverse events, follow-up for unsolicited adverse events is also done, and this is done for considerably longer than 4-5 days, although the exact duration will vary depending on the vaccine and context. For example, for the Pfizer COVID-19 vaccine, solicited adverse events were monitored for 7 days, unsolicited adverse events were monitored for 1 month, and serious adverse events (basically, anything that was fatal, life-threatening, required hospitalization, resulted in long-term disability, or resulted in a congenital anomaly) were monitored for up to 6 months after the second dose of vaccine.
One thing that always seems to escape the notice of the anti-vaccine lobby is that when it comes to vaccines there is a certain urgency to getting them out to the public. Sure, you could take it cynically and presume that this is about profit for the vaccine manufacturers, but the reality is that vaccine-preventable diseases cause very substantial harms at the population level and are very expensive to deal with. This was perhaps most obvious during the COVID-19 pandemic: at its worst, there were thousands of people dying from COVID-19 every day in the US, and there was very little we could do about that. Each day there wasn’t a vaccine, more people were being taken from us by this disease, as has been occurring with infectious diseases from time immemorial. It is not possible nor reasonable to expect that we would ever know everything about, frankly, anything we put into our body. At a certain point, you have to decide that you know enough, or put another way, that you know that the known and potential benefits of the intervention outweigh the known and potential risks. To that end, there is a very important point that Siri and the anti-vaccine lobby miss when it comes to the safety of vaccines: the safety of licensed vaccines is essentially assured by virtue of the sheer number of individuals who participate in trials.
In pharmacoepidemiology there is a statistical phenomenon known as the rule of three1.
In essence, it says that to have a 95% chance of capturing an adverse event that happens in 1/x people, you need to have 3x people receive that intervention to see that event occur at least once. Equivalently, if an event does not occur in a trial of x people, it can be presumed with 95% confidence that the event occurs at most at an incidence of 3/x. This does assume that the trial participant population is similar to the population of interest you are comparing to, so for example, if an outcome is very skewed to a particular demographic and they are not well represented in the trial population, this will not hold for the risk in *that* demographic. To give a concrete example, to have a 95% chance of seeing an adverse event that happens to 1 in 10,000 recipients of a vaccine, you need to give the vaccine in 30,000 people, and if an event does not occur in a trial where 30,000 people are vaccinated, it at most has an incidence of 3 per 30,000 or 1 in 10,000. In fact, this is fairly typical for the size of a vaccine trial. This was in fact the size of the Moderna COVID-19 vaccine trial, and it is in fact smaller than the trial for the Pfizer COVID-19 vaccine trial. Again though, as the perspective points out, if something happens to 1 in 100,000 people who get the vaccine and you vaccinate 330 million people, you would expect it to happen to ~3300 people. This will not be of any comfort to those who are among those 3300 and in no way does this suggest that those individuals should not receive all appropriate support, but it does mean that the possibility of a rare adverse event from a licensed vaccine that cannot be seen in trials is not particularly important to an individual who is contemplating vaccination because it is extremely unlikely to occur. Furthermore, the vaccines should be considered in the context of the risks from the diseases they prevent.
Siri also claims, with no evidence to support it, the old anti-vaccine talking point that today’s children are the sickest out of any generation and the rise in the incidence of chronic illness in these children must be attributable only to vaccination. This particular claim appears to be from RFK Jr. and colleagues from Children’s Health Defense, an organization whose name is about as honest as “Democratic People’s Republic of North Korea,” but was co-opted by Marianne Williamson, and written about by Dr. David Gorski. The upshot of this is the suggestion that vaccines are responsible for “50% of children having chronic illness today” is really quite ridiculous when one considers how chronic illness is being defined. By far the most common of these conditions is overweight or obesity and it is far more common among the publicly insured than the privately insured (43.2% vs 27.3%). I somehow doubt that vaccines have drastically increased the youth’s calorie consumption, particularly given that the privately insured are far more likely to be up to date on vaccination than those publicly insured. See why post hoc ergo propter hoc fallacy probably isn’t a good foundation for reasoning? There also isn’t anywhere close to 90 vaccines on the childhood schedule which Siri shoehorns into this claim as well, which isn’t true even if you maximally avoid all combination vaccines.
It strikes me as being particularly objectionable that Siri, who presents himself as one who stands up for the rights and wellbeing of those harmed by vaccines, would attempt to suggest something malicious by the mere suggestion of further studies to better understand these rare harms in a more timely manner, which is explicitly stated to have the goal of making vaccination safer.
Additional Thoughts on the Perspective
The perspective’s basic suggestion, that part of the NVICP budget be used to fund safety studies on vaccines, seems eminently reasonable to me, caveated on my admitted lack of expertise in the financial aspects of such an undertaking, and I leave space to those with that expertise to opine as they deem fit. It does not seem to me that anyone should be against research to better understand the safety of vaccines and any such measures will always have my support. I do have some minor quibbles and one major quibble.
Minor Quibbles
The perspective includes a table which describes harms of vaccination that have been observed and whether we understand the biological mechanisms underpinning them.
A few of them are marked “Not understood” but I think that that isn’t entirely fair (particularly if there is a category of “Hypothesized but Uncertain”). Specifically:
Narcolepsy following 2009 Pandemrix influenza vaccine: Narcolepsy is a condition in which individuals experience profound difficulty staying awake, sometimes in response to specific triggers. It has at times appeared as a joke in popular culture, but the condition can be profoundly debilitating and should not be minimized. The Pandemrix vaccine was a vaccine deployed for the 2009 influenza pandemic targeting H1N1 which used the AS03 adjuvant together with inactivated influenza virus. The extent to which this vaccine contributed to excess cases of narcolepsy (that is, cases of narcolepsy that would not have occurred anyway even if the vaccine were never given) is debated. Of further interest is that other 2009 pandemic H1N1 influenza vaccines using the AS03 adjuvant were not associated with narcolepsy, nor was it seen with MF59-adjuvanted vaccines (which were used in the US). Nonetheless, while not definitively proven, there does appear to be a culprit identified: the influenza nucleoprotein. The nucleoprotein is an internal protein of the influenza virus, and it is not the major target of seasonal influenza vaccines, but vaccines that rely on culturing of the flu virus in cells will still have some nucleoprotein remaining. Antibodies against this nucleoprotein can crossreact with a protein called hypocretin receptor 2, whose normal function is to sense the neurohormone hypocretin (aka orexin) and provide signals that tell us to stay awake, resulting in an autoimmune form of narcolepsy known as type I narcolepsy. It is proposed that the Pandemrix vaccine specifically had a higher than normal dose of the virus’s nucleoprotein, and this in turn triggered type 1 narcolepsy in genetically susceptible individuals. At the same time, narcolepsy following influenza infection has also been observed, raising the question of whether the vaccine truly contributed to extra cases of narcolepsy or whether they would have gotten it anyway, and possibly explaining the complexities of the epidemiological data regarding excess cases. Thus I would say the biological mechanism here (if I could propose my own categories) is “probably known but not definitively proven.”
Myocarditis following mRNA COVID-19 vaccines (CVAM): Here the story is more complex. It is true that a satisfactory, definitive explanation is currently lacking. We do have extensive epidemiological data that describes who is at increased risk, how the risk compares to that of COVID-19, whether vaccination affects your risk of myocarditis should you get COVID-19, and even strategies to reduce the risk of myocarditis from mRNA vaccines. Nonetheless, the latest evidence suggests that, in terms of the biological mechanisms, the basis for this reaction to the vaccines is a cytokinopathy: in essence, there is no true autoimmune attack on the heart from the immune system (meaning that the cells of the immune system now interpret the cells of heart as itself being a threat and work to eliminate it), but rather there is an overproduction of particular immune messengers known as cytokines, which immune cells within the heart react to and cause inflammation. This mechanism overall is good news because it suggests that immune-mediated damage to the heart is likely to be transient2, which broadly agrees with the outcomes data we currently have. In contrast, if an autoimmune mechanism were at play, patients experiencing this adverse event would likely need permanent immunosuppression to protect their heart from their immune system, opening them up to massive risks. This is very different from what CVAM today actually looks like, as it can even be managed entirely on an outpatient basis depending on its severity. If I had to classify it for the table by biological understanding, I would say “Incompletely Understood.”
As the Perspective authors outline, the point of understanding these biological mechanisms is to be able to have safer vaccines and to prevent people who would likely experience these events from receiving these vaccines. It is worth mentioning that there is at least one case of such a success story. The oral polio vaccine contains poliovirus which has been adapted to grow poorly inside humans. Thus, the vaccine induces an immune response against polio without carrying the risk of causing paralysis from poliomyelitis- usually. In between 2 to 4 per million individuals, the poliovirus in the vaccine can revert into a form that can cause paralysis, resulting in vaccine-associated paralytic polio (VAPP). This results in other downstream issues for polio eradication initiatives that are beyond the scope of this current post to address. Nonetheless, detailed biological investigation into the requirements for polio to cause paralysis led to the development of nOPV2 (novel oral poliovirus vaccine type 2). This vaccine was designed to have additional mutations in a key region of its genome that polioviruses have to mutate to be able to cause paralysis, resulting in a much higher bar to clear. In the real world, its use has been associated with VAPP at a rate of 7 cases per 100 million individuals, which is substantially lower (although we would all of course like for that incidence to be 03). In the same vein, what we currently know about the other serious harms highlighted in the table also suggest a path forward. For example, the experience with Pandemrix and narcolepsy highlights the importance of minimizing the amount of nucleoprotein in influenza vaccine products or mutating the nucleoprotein of vaccine strain viruses so that it no longer cross-reacts with hypocretin receptor 2. With regard to VITT, the risk seems to arise from an interaction between platelet factor 4 and a region of the adenovirus hexon protein, and thus it is possible that the hexon protein can be engineered to abolish this interaction4.
Major Quibble
I unequivocally support more research to be done into the safety of vaccines, particularly into the biological mechanisms, so that we may have safer vaccine products and effectively protect more people. However, it strikes me as being extremely naive to think that the problem we face with vaccine hesitancy is an information deficit issue, and on the contrary, I think regardless of what these hypothetical studies show, we should expect that they will be distorted into evidence that vaccines are unsafe. As things currently stand now, we have more safety data on COVID-19 vaccines than we do for any other vaccines and it is not close. There is no other intervention where we have studies of as many as >99 million people to look at to examine safety. Further, by every reasonable standard, COVID-19 vaccines have passed with flying colors regarding safety. Risks absolutely do exist, but they are small relative to those posed by COVID-19, and the totality of evidence available to us supports that the benefits of vaccination outweigh the risks for everyone aged 6 months and older (admittedly it is grayer in some groups than others). The problem we face with vaccine hesitancy is not an information deficit, so treating it like one is fundamentally the wrong approach.
It should be noted that despite the psychosocial climate and tainted information ecosystem we experience today, vaccination remains the norm, and it is critical for public health that it remain the norm. The problem, however, is that the disputes about the merits and demerits of vaccines plaguing the landscape today occur on fundamentally unequal footing. A simple majority for many vaccine-preventable diseases is nowhere near adequate for control, let alone elimination and herd immunity. For example, measles requires approximately 95% of people to be vaccinated to terminate transmission chains if the virus is introduced into a community. If the uptake of MMR vaccine is a simple majority, just above 50%, you will see explosive outbreaks of this horrific disease.
Vaccine hesitancy is fundamentally a problem of trust, and they are emblematic of the crisis we face with misinformation. That crisis of trust is fueled overwhelmingly by the highly organized and well-funded anti-vaccine lobby, who make use of highly developed social networks to inculcate their talking points into the public discourse. Many vaccine hesitant individuals are completely unaware that they are presently asking about a carefully constructed talking point that has been presented in bad faith from the leaders of the anti-vaccine lobby, and this is a very intentional effort to help normalize hesitancy. We should not expect any effort to learn more about vaccine safety to have any kind of salubrious effect on vaccine uptake because the data from these studies WILL be distorted by the anti-vaccine lobby into claiming harms that aren't real or exaggerating harms that are in order to allege malice, negligence, and complicity on the part of those who deliver vaccines to the public. If the studies return negative results, they will attempt to discredit them through the genetic fallacy, which becomes particularly facile when studies on vaccine safety would be funded by an excise tax on vaccines through the NVICP.
Vaccine hesitancy should be viewed as a trust problem fueled by misinformation that requires intimate, rapid collaboration between trusted messengers, infodemiologists, agnotologists, and subject matter experts to address. The anti-vaccine lobby should be considered to have a major advantage simply on the basis that they peddle lurid lies. It makes intuitive sense that these ideas would spread far more rapidly than the truth. The truth is boring: vaccine safety is ensured by a group of dedicated professionals who see patients directly, conduct clinical trials, and tirelessly scrutinize .csv files of data and carefully design statistical comparisons to be informative and also accurate while not being given nearly enough funding and not being appreciated nearly enough for their contributions. They are doing what is expected of them. By contrast, any falsity from the anti-vaccine lobby comes with the tacit (although sometimes not) invitation into the anti-vaccine in-group. It comes packaged with the promise that you can be so clever as to not only know more than the experts but that you can evade their malicious attempts to harm you and those you hold dear, and you can demonstrate your membership to this rarefied sect of cognoscenti and protect your fellow man by helping propagate this misinformation. It is a much more exciting premise. It’s a shame that there isn’t an iota of truth to it (although perhaps not, as that would imply grim things about the state of our regulation).
Any new data presented about vaccines should be expected to be subject to distortion and lies to ensure the vaccine comes out looking as bad as possible. As we strive to better understand vaccine safety in pursuit of extremely worthwhile goals that have my absolute support, we would do well to approach the issue of hesitancy proactively, with the mindset that the information ecosystem is adversarial to facts and to expertise itself. Data does not suffice as a compelling defense in that realm, as unfortunate as that admission may be.
In case you’re curious about where the rule of 3 comes from, as it happens, I do have notes on the subject:
Caveated on the initial episode of myocarditis itself not resulting in substantial damage to the heart, which, it seems in the majority of cases, it does not appear to do, thankfully.
The inactivated polio vaccine (IPV) carries no risk of causing VAPP because the polioviruses in it cannot replicate and cause disease. However, IPV has a much weaker effect on transmission than OPV, which means it is generally not believed to be enough for eradication efforts. Beyond that, IPV has to be given by injection, which requires trained injectors to administer the vaccine, whereas OPV requires no special training. Nonetheless, given the rare but real risk of VAPP, the Global Polio Eradication Initiative has called for a shift away from OPV.
Even though this is theoretically true, because VITT is similar in incidence to VAPP from OPV, it is probably not possible to look at this risk with a pre-licensure trial and so any adenovirus-vectored vaccines demonstrating this interaction will likely only show evidence of it after massive, observational studies are conducted. It is also worth noting that this clotting disorder has now also been observed after adenovirus infection.
Thanks Dr N . Your response to Siri is very well argued. Appreciate you taking the time to write this .