(Bloomberg Opinion) — Sam Fazeli, a Bloomberg Opinion contributor who covers the pharmaceutical industry for Bloomberg Intelligence, answered questions about preliminary research from Israel indicating that the Pfizer-BioNTech SE Covid-19 vaccine is highly effective not just in preventing disease but also in curbing the spread of coronavirus infections. The conversation has been edited and condensed.
The news out of Israel sounds very good. Is it as encouraging as it seems?
The holy grail of vaccination is to supress infections. When this happens at a 100% level, it’s called sterilizing immunity and means the virus just cannot infect you. The second-best thing — and what the vast majority of vaccines used against infectious diseases actually do, including the ones approved for Covid-19 — is to prevent disease and, more importantly, serious disease. We now have a claim that the Pfizer-BioNTech vaccine also prevented infections with the Sars-Cov-2 virus by close to 90%. While this does sound encouraging, there are some serious caveats that need to be taken into account before we get too excited and risk causing harm.
How do vaccines prevent infections anyway?
To prevent infections by a respiratory virus, which is what the latest data from Israel is suggesting is happening in the vast majority of vaccinated individuals, you need high levels of antibodies in the nasal passages and the lungs to actually stop the virus from taking hold. Our current Covid-19 vaccines are injected into arm muscles and induce a very strong immune response but not the type that is normally required for preventing an infection in the linings of the respiratory tract. For that — to get technical for a moment — you need antibodies belonging to what’s known as the IgA family, not the IgG family that is induced by the vaccines. High levels of IgG antibodies can curtail an infection by a respiratory virus but not nearly as well as that achieved by IgA antibodies, which reside in the mucous of the respiratory tract and the gut.
What do we know about prevention of infections by Sars-Cov-2 vaccines?
There have been some hints of prevention of infection from vaccine studies. Trials of the AstraZeneca Plc-Oxford University shot were the only ones that required weekly nasal swabs for all participants, and they did show a cut of 67% in asymptomatic infections — that is, evidence of virus in the nasal passages of subjects but no symptoms — in its U.K. phase III trial. There has also been anecdotal evidence, derived from I srael’s Sheba Medical Center’s data, that the Pfizer-BioNTech vaccine can reduce asymptomatic infections (here, estimated at about 50%). This is all very good news, especially when combined with the vaccine’s effectiveness at preventing disease 90 percent of the time.
Why should we be skeptical of this latest data?
The latest data suggesting 90% prevention of asymptomatic infections looks amazing but raises serious questions. First, as of the time of writing this piece, the data have not been published, even in preprint form; some pages were posted on Twitter, that’s all. We believe, given the way the analysis was done, that the study could be exaggerating the efficacy against asymptomatic infections. The comparison shown is between vaccinated and unvaccinated groups that could be easily biased by vaccinated people being from a different demographic that may impact their behavior. An example would be that early adopters of vaccines are likely to be people who are more careful and less likely to contract an infection. Conversely, the “control” group of unvaccinated people used in this study may have a higher risk of infection because they come from a different socioeconomic background. Also, vaccinated individuals are not required to have a Covid test, which again creates bias in the study because testing of unvaccinated people may be driven more by those who get symptoms, contributing to an apparently higher rate of symptomatic disease in unvaccinated people and, conversely, a higher rate of asymptomatic disease in those who have had a vaccine. These observational studies always suffer from such drawbacks and require rigorous analysis and review to make sure these issues are taken into account. I will get excited about this data once it is published in a peer-reviewed journal.
Why does it matter if prevention of infection is 50% or 90%?
The biggest risk with making pronouncements like this is that of an unintended, but likely, consequence that those vaccinated may feel they are at lower risk than they truly are from an infection. Worse still is the perception that they are of lower risk to others who have not yet been vaccinated. It might drive behavioral change such as less mask-wearing and more socializing before we have truly understood the data and the vaccines’ effects. All this risks undermining longer-term confidence in vaccines in the form of a reverse “crying wolf.” Let’s not build up hope until we have done all the checks and balances.
But what if the data turn out to be correct – should we not get excited then?
Even if the data turn out to be right, they may not be of much significance in the long run. It is not surprising that vaccination leads to some reduction in infections. After all, as noted above, we have seen evidence of this in the controlled trials of the vaccines. But the issue is that all these studies are looking at efficacy at a very short time after the second dose of the vaccine, mostly one to two weeks later. That’s when the immune response, in terms of neutralizing antibody levels, is at a very high level. What we really need to know is whether those vaccinated show a similar reduction in the risk of being infected, and by inference being infectious, in three, six or nine months after the second dose, when the initial antibody load has diminished.
So after all this , what should we be looking at?
The key is to remain focused on the number of cases of severe/critical disease, hospitalizations and death. This is what we’re most interested in preventing, and it appears that the vaccines have very high efficacy there. But even here there is a risk of bias in the early reads of data, again driven by demographics. We need more time. I remain excited about the vaccines and their potential to solve our pandemic problem. But it is incumbent on scientists, health authorities and experts to be careful with how they disseminate information and data.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Bloomberg Opinion provides commentary on business, economics, politics, technology and markets.