Three weeks ago I discussed very preliminary reports of blot clots observed in people receiving the AstraZeneca vaccine, resulting in 13 countries suspending use of the vaccine. Since then the data has been tracked very carefully, and our understanding of the potential risk has evolved, but no definitive answer yet exists. It’s a great example of how scientists tackle such questions in real time, and the difficulty that epidemiological questions can sometimes present.
Initially the concern was the observation of various types of blood clots in the week following receiving the AstraZeneca vaccine for COVID-19. An initial review found that the total number of reported blood clots was within the expected range for the background occurrence. There was also no evidence for a specific link to the vaccine. In any case, the absolute numbers were so low that even if every observed blood clot was caused by the vaccine more lives would be lost by delaying vaccination.
But attention has shifted to one specific type of blood clot, cerebral venous sinus thromboses (CVST). This is a clot that forms in the veins that drain blood from the brain. If serious enough these clots can be fatal. Further, early reports suggest these clots are occurring more in people under 55 years old and more in women. Perhaps most intriguing (and concerning) is the fact that CVST in these patients has been associated with a low platelet count. This could be a sign of an inflammatory auto-immune process, something that can be triggered by a vaccine.
A preprint non-peer-reviewed paper discusses 9 cases showing low platelets with CVST and concluded that these cases are probably “vaccine induced”. However, expert reaction to this paper has been largely negative, indicating that the data is too slim and the research quality too low to draw any conclusions.
One critical aspect of a possible link between the AstraZeneca vaccine (to be clear, there is no concern about blood clots in other COVID vaccines) and CVST is the question – what is the background rate of these clots? It might seem that this should be an easy question to answer, but it isn’t. These are rare events, and so even a difference in reporting of a few cases can dramatically affect the statistics. Further, it is possible, even likely, that COVID itself increases the risk of blood clots, and so the background rate during the pandemic may be significantly higher. The BBC reports:
Germany’s Paul Ehrlich Institute has reported 31 cerebral venous sinus thromboses and nine deaths out of 2.7 million people vaccinated there.
The most recent UK data reported 30 clots linked with low platelet counts and seven deaths out of 18 million people vaccinated.
The European Medicines Agency, which has assessed data from around the world, estimates there is around a one in 100,000 risk of a CVST in people under the age of 60 who have been given the AstraZeneca vaccine.
The estimated background rate varies from 2 to 16 per million people per year. But this number, again, may be higher during the pandemic. While the EMA did not conclude that there was any specific demographic group at risk, most of the cases are in women under 55. This population may have a higher background risk of CVST also.
Amidst this uncertainty, however, the EMA has signaled it will officially report today or tomorrow that they are now ready to conclude that there is a link between the AstraZeneca vaccine and an increased risk of CVST in people under 55. However, they also stress that the benefits of getting the vaccine (the risk of not getting it) is greater than this risk of rare blood clots. That is turning out to be a hard sell, and many people are forgoing the vaccine because of this possible risk.
There are a few points worth emphasizing here. First, we do not know what the ultimate answer is, in terms of if there a genuine increased risk, in which populations, and due to what specific mechanism? Each step of the way scientists are giving the best answer possible based on the currently available evidence, which includes a wide range of uncertainty.
In normal times, the precautionary principle would prevail in such cases. However, during a pandemic the risk of delaying the vaccine will definitely lead to otherwise preventable deaths. If you are dead, it doesn’t matter if you were killed by COVID or a rare vaccine side effect. So we should take whichever course leads to the fewest deaths, which is clearly on the side of getting the vaccine, even in the worst-case scenario in terms of risk of CVST.
An obvious question is – can’t people just take other vaccines that don’t have this risk? Again, in normal times that might be the case, but not in the middle of a pandemic. In Europe COVID is experiencing another surge, likely caused in part by the emergence of new variants, and perhaps also by prematurely relaxing preventive measures and pandemic fatigue. The AstraZeneca vaccine is a critical part of Europe’s strategy to vaccinate its population as quickly as possible. It is also cheaper and easier to administer than other vaccines, and so is critically important to poorer countries. You can also argue that worldwide we are in a race against the emergence of new and deadlier variants, and any delay in vaccination can have significant repercussions. A global risk vs benefit approach is needed here.
Some countries are compromising, and not giving the vaccine to those under 55 or 50, while still giving it to older populations who apparently do not have the higher risk. If the risk is truly only higher in women, than the excluded population can be narrowed further. This still may not optimize risk vs benefit, but it is better than a global ban.
What we need now is more research to learn as much about this possible link as we can, mitigate if possible, and maximize benefit to risk in the face of a still-raging deadly pandemic.