How Vaccine Safety Can Be Proved

Data, transparency and independent review are crucial for establishing efficacy and building public confidence in science, processes and outcomes.

Kevin Sheth and Serena Spudich report in the Wall Street Journal:

Scientific review and transparent results will be critical. While doing no harm is the bedrock principle of clinical care,  all treatments have unintended consequences. It will be important to have population-level reporting that evaluates events so they can be rigorously reviewed. Federal agencies should track and review events not only by industry sponsors but by disease experts who have access to primary data. Regular reporting should be widely available to the scientific and clinical communities as well as the broader public.

As acute and infectious disease neurologists, we are often asked if a vaccine is safe. Pfizer’s announcement Monday that its Covid-19 vaccine is 90% effective makes the question even more pertinent. Earlier this year, two late-stage clinical trials, from Johnson & Johnson and AstraZeneca, were briefly paused after reports of neurological concerns in individual study participants. Each of these rare instances was carefully scrutinized as an adverse event—a serious clinical observation that requires review by the Data and Safety Monitoring Board.

At least two patients were initially thought to have developed transverse myelitis, a condition in which the patient’s immune system attacks the spinal cord. Similar autoimmune responses can affect the nervous system in conditions that mimic multiple sclerosis or Guillain-Barré syndrome, both of which can cause paralysis. Non-neurological complications such as anaphylaxis are also possible, related also to a vaccine’s misdirected effect on the immune system.

The purpose of listing potential adverse effects is not to create a list of reasons to avoid vaccines. These disorders are rare, and most cases spontaneously occur in otherwise healthy people who haven’t recently been vaccinated. Thus when a subject in a trial develops one of these disorders, it is often unclear whether the vaccine has anything to do with it.

In randomized clinical trials, industry sponsors and trial committees don’t necessarily know if a particular patient has received the vaccine. Patients and doctors often don’t know. Depending on the trial design, they might have received a placebo or an already approved vaccine as part of the control group. That means that in every one of these rare events, researchers must establish whether the afflicted patient received the Covid vaccine, another vaccine or no vaccine at all.

At the same time, a detailed review is required to characterize the event in the first place. Neurological diseases aren’t always straightforward: What looks like a seizure may turn out to be something else. It’s entirely possible that a given event is what it appears to be, and is a side effect of the vaccine. But since these diseases occur in the general population, one has to determine whether the clinical disease was likely related to the study intervention or not.

The broader public can rest assured that randomized trial data will help answer all these questions and provide a sound basis for what to expect when the vaccine is administered to the general population. Many companies have made significant efforts to enroll patients in countries around the world with particular attention to diverse populations. But even the largest of these trials enrolled 50,000 to 100,000 patients. The World Health Organization suggests that the inactivated influenza vaccine is associated with one to two cases of Guillain-Barré per 100,000 patients—a rate of 0.001% to 0.002%. Put that in perspective: If a Covid-19 vaccine were as safe as the widely accepted flu shot, then 100 million vaccinated would correspond to 1,000 cases of Guillain-Barré. But 1,000 cases sounds like a much bigger number than 0.001%.

Whatever the real-world rates of complications turn out to be, they have to be placed in the context of what we know an effective vaccine will do: prevent death, disability and disruption from Covid-19. Recent surveys suggest that 20% to 30% of Americans say that they may not get a Covid-19 vaccine—twice that among minority populations, which have been disproportionately hit by the disease. How can public-health authorities reassure people that the vaccine will be safe and effective?

We will soon learn more about the precise safety and efficacy data from closely watched large-scale trials. Scientific review and transparent results will be critical. While doing no harm is the bedrock principle of clinical care, virtually all treatments have some unintended consequences. No matter how few patients are affected by adverse events, we must acknowledge that they may occur. When they do, it will be important to have a population-level reporting structure that evaluates events carefully so they can be rigorously reviewed. Federal agencies should guide, track and review adverse events. Adverse events should be reviewed not only by industry sponsors but by disease experts who have access to primary data, empowered through public health authorities. Regular reporting should be widely available to the scientific and clinical communities as well as the broader public.

There are many reasons to be optimistic. In the modern era, highly innovative biological approaches to vaccine development such as the use of mRNA platforms are being deployed at an unparalleled scale. It is possible that these newer platforms will provide unprecedented safety and efficacy profiles for vaccines. This innovation is the direct result of years of investment in biomedical research and a bold commercial biopharmaceutical landscape.

It’s likely that this won’t be the last time our society has to do this. Improving our ability to implement a successful scientific advancement in the community will pay lasting dividends for public health.