The decisions by France, Germany and other countries look baffling. The data we have suggests that numbers of adverse events related to blood clots are the same (and possibly, in fact lower) in vaccinated groups compared to unvaccinated populations. The UK MHRA, WHO and also the International Society on Thrombosis and Haemostasis* have recommended continuing the Oxford AstraZeneca vaccine roll out.

Halting a vaccine roll out during a pandemic has consequences. This results in delays in protecting people, and the potential for increased vaccine hesitancy, as a result of people who have seen the headlines and understandably become concerned. There are no signs yet of any data that really justify these decisions.

*ISTH statement – https://twitter.com/isth/status/1370424157947752452

‍

The news that Germany and other countries have suspended roll out of the AZ vaccine is disappointing. Whilst it is right that any possible links between serious adverse reactions to a vaccines are investigated, it is important to establish the likelihood of causation, as opposed to association, when doing so. The numbers of clotting events associated with the vaccine are very small, and according to the figures from AZ are well below the average that you might expect in any given population – we must remember that many millions of people have been safely vaccinated. Moreover, it is likely that the early stages of the European rollout have primarily comprised elderly and more clinically vulnerable patients, which might be expected to have a much higher incidence of random clotting-related events.

Thus, the concerns raised in my view, whilst valid and worthy of investigation, are probably not proportionate to the response in terms of halting mass vaccination in multiple countries. Since many European countries are currently experiencing another resurgence of SARS-CoV2 infections and yet are lagging behind in terms of roll out, the importance of continuing the vaccination programmes cannot be underestimated, and the harm caused by depriving people of access to a vaccine will likely vastly outweigh even the worst case scenario if any link to the clotting disorders is eventually found. It should also be noted that nationwide gestures such as this are bound to fuel hesitancy, or more extreme anti-vaccine sentiment, further undermining the vaccination effort.

‍

These Phase 4 (post-licensing) events are not unusual when rolling out new vaccines – to tens of millions of people rather than tens of thousands of participants in the earlier clinical trials.

We went through something similar for the 2009 influenza A(H1N1)pdm09 vaccines where one of them was eventually found to have a causal association with narcolepsy:

https://pubmed.ncbi.nlm.nih.gov/24360892/

and earlier multiple studies re-examining the MMR vaccine safety, in view of an earlier study by Andrew Wakefield proposing a link to autism- which has been finally refuted:

https://www.acpjournals.org/doi/10.7326/M18-2101

Most of these studies start by comparing the risk of such adverse events in vaccinated vs. unvaccinated people, compensating for various confounding factors in each group – to assess whether the vaccine increased the risk of any particular event/complication, post-vaccine.

If an increased association is found with vaccination, then attempts to understand the mechanism usually follow – so as to avoid such problems in the future, e.g. if the cause might be due a particular adjuvant or other ingredient in the vaccine:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6413474/

For thromboembolic events, we know that COVID-19 itself increases the risk of such events massively though the exact mechanism is still unclear:

https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(20)30383-7/fulltext

Careful investigations of the various AstraZeneca COVID-19 vaccinees are needed (and ongoing) to separate their natural background risk of developing such complications versus any additional risk (if any) from the vaccine – as has been done for the other examples given above. However, note that no thromboembolic risk was noted in the AstraZeneca clinical Phase 1-3 trials.

Whilst such investigations are ongoing, it is not unexpected that the AstraZeneca vaccine rollout in those affected populations may be temporarily halted.

‍

We are closely reviewing reports but the evidence available does not suggest the vaccine is the cause.

Blood clots can occur naturally and are not uncommon. More than 11 million doses of the AZ vaccine have now been administered across the UK, and the number of blood clots reported after having the vaccine is not greater than the number that would have occurred naturally in the vaccinated population.

We are working closely with international counterparts in understanding the global safety experience of COVID-19 vaccines and on the rapid sharing of safety data and reports.

People should still go and get their COVID-19 vaccine when asked to do so.

Background:

• For further details, see our statement and advice published on 11 March
• The MHRA encourages anyone to report any suspicion or concern they have beyond the known, mild side effects on the Coronavirus Yellow Card site. Reporters do not need to be sure of a link between a vaccine and a suspected side effect but are still encouraged to report.
• For more information on COVID-19 vaccine adverse reactions, see our weekly report

‍

Blood clotting, or thrombosis, occurs for a variety of reasons, the lead up to an even can often be undetected in the general population, and venous thrombosis is relatively common affecting 1 to 2 in 1000.  Risk of thrombosis increases with age, although it can occur at any age, and there are additional risks for some underlying health conditions.

The Paul Erlich Institute in Germany have announced that they are investigating events of Sinus Vein Thrombosis as a possible adverse reaction to the AstraZeneca vaccine, which is notable. Cranial Sinus Vein Thrombosis (CVST) is a rare type of thrombosis affecting 5 in one million patients and causes a stroke, and generally takes place in patients with abnormal platelets which may become hyper-activated in certain circumstances. The specific circumstances in Germany are still not clear here and they have provided cautionary advice that anyone with severe and persistent headaches or pin-prick skin bleeding seek medical attention as a matter of priority.

‍

The report from the Paul Ehrlich Institute gives a diagnosis for recent reports of adverse events following administration of the Oxford AtraZeneca vaccine, Cerebral Venous Sinus Thrombosis (CVT). In the briefing it was stated that seven cases per have been identified in 1.6 million vaccinated (about 4.5/million post vaccination).

While it has been reported that background incidence of CVT is only 2 to 5 cases per million per year and was considered to be very rare, more recent papers have suggested it is more common than previously thought. A study from Australia found  15.7 cases per million per year (Devasagayam S, Wyatt B, Leyden J, Kleinig T. Cerebral venous sinus thrombosis incidence is higher than previously thought: a retrospective population-based study. Stroke. 2016 Sep;47(9):2180-2.) and one from the Netherlands 13.2 cases per million per year (Coutinho JM, Zuurbier SM, Aramideh M, Stam J. The incidence of cerebral venous thrombosis: a cross-sectional study. Stroke. 2012 Dec;43(12):3375-7). The reported incidence was actually higher in people under 50 than in people over 50 years old. So the estimated background incidence of is only 2 to 5 cases per million per year would appear to be underestimated by about 4 to 8 fold.

So we would normally expect to see about 1 or 2 cases per month per million people in people under 50. So it does appear that CVT has been reported more commonly than expected at least in Europe. However, one of the key problems with identifying association between rare/uncommon events is that it is not just one type of possible event that is being looked for. When you consider that there are quite a few possible diseases then the probability of such a statistical clustering for one outcome is actually quite high even when no cause and effect exists.

Against this, the infection mortality rate in men in their mid-40s from COVID is of the order 0.1% or about 1000 deaths per million infections substantially greater than the risk of CVT (O’Driscoll M, Dos Santos GR, Wang L, Cummings DA, Azman AS, Paireau J, Fontanet A, Cauchemez S, Salje H. Age-specific mortality and immunity patterns of SARS-CoV-2. Nature. 2020).

Clearly this possible association needs to be thoroughly investigated, but we do need to consider the real harm from delays in immunization campaigns at a time when the incidence of COVID is still increasing in several European counties when deciding whether or not to pause vaccination campaigns.

‍

The decision of the MHRA to continue to keep a close observation on the evolving evidence on clots having clarified that there has been no signal in the UK to suggest any increase, seems pragmatic and sensible. When it comes to events that may be as low as single figures in a million, it can be difficult to clarify what are random fluctuations in normal background levels vs real effects. Rare events in this statistical territory will not however change the overwhelming risk/benefit case for having a vaccine, but it remains important that the public are made aware of any new risks should they emerge and be confirmed.