I am sure you all saw the recent MHRA press release and guidance, discussing the issue of a possible link between the AstraZeneca COVID19 vaccine and a rare type of thrombotic condition of Vaccine induced thrombocytopenic thrombosis (VITT). This is a good example of why it is important to use the yellow card reporting system and a reminder to us all why we ideally need phase 4 research studies (post marketing surveillance) in non-pandemic times.
This statement was a helpful conclusion to the developing concerns around thrombosis and vaccination. There have been several anecdotal cases reported of unusual site thrombosis following vaccination in otherwise healthy adults. We have seen a few of these potential cases locally and it is good to get timely access to national data when you have local concerns.
Bringing together all the yellow card reports and known cases, the MHRA describe the risk of serious harm with the vaccine of around 1:250,000. Unfortunately, this risk seems higher in younger patients – reported at 1:100,000 by the BBC on this nice comparative infographic.
Although concerning, the first thing to highlight is that this risk is still very low indeed; younger adults are still twice as likely to die from COVID19 and around three times more likely to die in a road traffic accident than suffer a serious harm due to vaccine side effects.
But what is this serious harm anyway? Well, I don’t think we know exactly. There appears to be a specific syndrome caused by vaccination (AstraZeneca in particular) which presents at 5-28 days and is characterised by thrombocytopenia and unusual site thrombosis. So much so, that the British Society for Haematology have started calling it Vaccine Induced Thrombosis and Thrombocytopenia (VITT). However, others have given it a different name – down under (and in Canada) they appear to prefer the term Vaccine induced Prothrombotic Immune Thrombocytopenia (VIPIT). Just goes to show that we are dealing with uncertainty.
So before we roll our sleeves up and get stuck into the uncertainty, let’s take a look at a similar condition to see if we can learn anything that may help.
Heparin Induced Thrombosis and Thrombocytopenia (HITT)
This condition not only has very similar wording to the above, but some early bench research from Germany has suggested significant similarities. You would be forgiven for never having seen this in ED, but we see a fair few cases every year on the ICU. HITT tends to present about 5-10 days after starting heparin compounds, and presents with worsening thrombosis and thrombocytopenia.
How on earth do you give an anticoagulant that leads to low platelets, which causes thrombosis you say? I hear you. Simply put, heparin binds to platelet factor 4 (on the surface of platelets) to create an antigen complex; this antigen is attacked by IgG antibodies in a dysfunctional immune response. Platelets get activated, clump, release procoagulant microparticles and thrombin generation is triggered, with all the consequent platelet consumption and thrombotic clinical manifestations.
Clots can be found anywhere with HITT; DVT and PE can extend, but you can also see adrenal vein thrombosis, necrotising skin lesions, cerebral venous sinus thrombosis and mesenteric thrombosis.
Treatment is based on cessation of all heparin compounds (including line flushes) and consideration of alternative anticoagulants to treat thrombosis, such as fondaparinux, argatroban, danaparoid or other therapies (DOACs, etc). Platelet transfusions are to be avoided, as you don’t want more clumping. Emerging treatments include intravenous immunoglobulin and plasmapheresis, 2 treatment modalities that always get wheeled out when an antibody reaction is involved (1). The evidence to date is case report based, but these treatments are acknowledged universally as options for refractory HITT.
I am here for VITT, not HITT?
Fair enough. But it is worth revising HITT based on the early data suggesting that Vaccine induced thrombocytopenic thrombosis has a similar pathophysiology. In March, German researchers from the Paul Ehrlich Institute were busy analysing serum from 31 cases of CVST following vaccination and reported HITT mimicry in multiple samples. The German Society of Thrombosis and Haemostasis subsequently released a statement suggesting that the pathophysiology was so similar, that investigation and treatment modalities should be as well. The BSH produced similar recommendations shortly after.
What is the best way to identify this condition in the ED? I would suggest being your usual glorious selves and taking a thorough history (including recent vaccination), performing a thorough examination (checking for pupura and signs of VTE in particular) and bearing this condition in mind throughout. If a patient is presenting shortly after vaccination (4 to 28 days) with significant symptoms, a platelet count would be a useful place to start regarding investigations. This can be followed by tailored imaging to investigate for thrombosis, if you have clinical suspicion. A very high d-dimer and a low fibrinogen would also support the diagnosis. Specific HITT antibody assays can be sent as well to help clarify the diagnosis, but they do not come back quickly…
It is vital to identify Vaccine induced thrombocytopenic thrombosis early, as the treatment is unusual; all heparin based compounds should be avoided (despite the presence of thrombosis), non-heparin based anticoagulants should be commenced and early consideration given to immunomodulatory therapy (immunoglobulin preferentially). Platelet transfusions should be avoided if at all possible. This is all well and good, but my strongest personal recommendation if you are unlucky enough to encounter a case would be to arrange an early MDT involving haematology. This is a relatively unknown condition, with a very limited evidence base. You do not want to be managing it alone.
HOWEVER. It is as rare as hens’ teeth. Therefore be pragmatic and sensible. Think about a platelet count and considered clinical examination in a patient with post vaccine symptoms and a consistent history. I personally would do very little else, unless my clinical suspicions are otherwise raised.
I thought this was all about headache and CVST?
No. It is not.
Both HITT and VITT can present with unusual site thrombosis. The site of this thrombosis can indeed be in the brain (CVST), thus presenting with headache. But it doesn’t have to be. As above, unusual site thrombosis can refer to clots in a variety of locations. Some authors have reported that Vaccine induced thrombocytopenic thrombosis has a predilection for CVST, but I am not sure we can draw such conclusions when we are talking about a case series of 30 odd patients and a reliance on clinician reporting. This does, however, give us an opportunity to revise CVST a bit, think about when to consider it in the patient with atraumatic headache and our approach to diagnosis in the ED.
CVST in isolation is another rare pathology, with an incidence of 1:100,000. It is tricky to diagnose, as it can present with isolated headache only. Like other rare but serious cases of headache, the consequences of missed diagnosis are severe. If CVST is untreated, backpressure from the obstructed venous sinus often leads to intracranial haemorrhage. You are then in a position where the treatment for the index condition is anticoagulation, but the patient has intracranial bleeding. We have a few of these on our neuroICU every year and they don’t get any easier to manage…
There are some things to remember about CVST that can help you in all this. Firstly, it is very rare, as above. Second, there are specific risk factors which make the condition more likely, such as: late stage pregnancy/post partum; known thrombophilia; active malignancy and COCP use. Third, there are some specific clinical examination findings in the context of headache that can increase your pretest probability: papilloedema on fundoscopy for instance; or the presence of otherwise unexplained focal neurology, or seizures.
A d-dimer test can occasionally add value here also. There is a reasonable body of evidence from 2 previous systematic reviews in 2012 and 2015 to suggest that a negative d-dimer test (below your standard cut point for VTE, not age adjusted!!) has a reasonable sensitivity for identification of disease and a good negative predictive value (3,4). There is even an EMJ BestBET on this (5). However, the evidence is not perfect and there are real downsides to indiscriminate use of this test, as we all know. I would caution against routine evaluation and suggest you apply this test in the context of your overall assessment. For me, that means when I have a patient who is GCS 15 with no focal neurology and no other indication to image, but there are some concerns in the history or specific risk factors that make me consider the diagnosis. I find a negative value helpful reassurance then, and a positive value pushes me towards a discussion about imaging.
Of note, the gold standard imaging tests for CVST is an MR venogram, or a CT venogram if you are happy to take the ionising radiation/contrast. These can be challenging to access and get reported, depending on your neuroradiology expertise available. For what it’s worth, even when seriously considering CVST, I tend to start by ordering a plain CT brain. Sometimes this reveals alternative pathology, sometimes it reveals subtle signs of raised intracranial pressure or venous haemorrhage in keeping with the diagnosis of CVST, and sometimes it is completely normal. All of these results affect my assessment of risk. So any answer is a helpful one.
There is a nice open access and recent review of CVST here (6).
A note on pretest probability.
There has been a lot of chat about getting ‘Bayesian’ on social media, when discussing these issues. The insinuation being either that the risk of these conditions presenting to ED is so very low, that no tests other than a definitive one would be useful, or that the risk is just so low we should not investigate at all.
I am not sure I agree with this. Just because something is rare, does not mean it does not exist in the ED, or that a through clinical evaluation is unhelpful. And if a rare condition has catastrophic consequences if missed, then there is a clear onus on us to consider and tailor investigation where appropriate. As an example, subarachnoid haemorrhage is actually very rare – 6:100,000 according to some authors (7). We investigate this particular condition very aggressively, and have done so for a long time.
There is no doubt that this is a new area of research and discovery. Inevitably there will be changes to the evidence base in the next few weeks so keep an eye on the blog and on our twitter feeds to keep up to date. We are also expecting some guidance from RCEM soon.
We would also recommend reading the two papers published in the NEJM today describing case series of patients and the explanation of the VITT mechanisms from Germany (10,11).
I am not saying here that everyone with a post vaccination headache should get a d-dimer measured. And I am not saying that everyone with a post vaccination headache should get a MRV/CTV. And I am definitely not saying that everyone with a post vaccination headache should not get any tests done, because both VITT and CVST are so rare.
I am saying that we should remember to be good clinicians (diagnosticians?) first and foremost. We should evaluate risk based on historical and clinical features. We should inform this risk assessment judiciously, with ancillary tests like a platelet count, fibrinogen, d-dimer when indicated. And if we really think there is a significant risk of VITT/CVST based on this preliminary assessment, then we should pursue definitive investigations and seek expert help.
Like always, the vast majority of headaches attending the ED will be primary headache disorders and need nothing more than diagnosis, effective analgaesia, reassurance and appropriate disposition (8,9).
Emergency physicians are generally excellent at thinking about and looking for needles, in our continually growing haystack. Vaccine induced thrombocytopenic thrombosis is just a further needle. CVST has always been lurking in there. Keep an eye out. Bear them in mind. Remind yourselves what to look for. Maybe your particular haystack won’t have any of these particular needles in it at all. But at least you’ll be looking. And thinking.
Above all, don’t revolutionise your practice based on a risk that is marginally above that of being struck by lightning. The opportunity costs would likely be far greater than any benefits.
UPDATE from @EMManchester on 13/4/21:
The Royal College of Emergency Medicine, in association with the Society for Acute Medicine and Royal College of Physicians, have released a guideline as linked below. A key factor here is ensuring that we don’t put people into the investigation pathway unless they need investigation for VITT, else you might end up doing a lot of FBCs.
1 – Padmanabhan et al. IVIg for Treatment of Severe Refractory Heparin-Induced Thrombocytopenia. Chest 2017 Sep;152(3):478-485
2 – Linkins, LA. Heparin Induced Thrombocytopenia. BMJ 2014;349:g7566
3 – Dentali et al. D-dimer testing in the diagnosis of cerebral vein thrombosis: a systematic review and a meta-analysis of the literature. J Thromb Haemost 2012 Apr;10(4):582-9
4 – Alons et al. D-dimer for the exclusion of cerebral venous thrombosis: a meta-analysis of low risk patients with isolated headache. BMC Neurol. 2015; 15: 118.
5 – Smith andd Kumar. BET 1: Does a normal D-dimer rule out cerebral venous sinus thrombosis (CVST)? EMJ 2018;35:396-397.
6 – Ulivi L, Squitieri M, Cohen H, et a lCerebral venous thrombosis: a practical guide Practical Neurology 2020;20:356-367.
7 – Lin et al. Incidence of subarachnoid haemorrhage. Role of Region, Year, and Rate of Computed Tomography: A Meta-Analysis. Stroke 1996;27:625–629
8 – Locker et al. The utility of clinical features in patients presenting with nontraumatic headache: an investigation of adult patients attending an emergency department. Headache 2006 Jun;46(6):954-61
9 – https://www.nice.org.uk/guidance/cg150
10 – Thrombotic Thrombocytopenia after ChAdOx1 nCov-19 Vaccination https://www.nejm.org/doi/full/10.1056/NEJMoa2104840?query=featured_home
11 – Thrombosis and Thrombocytopenia after ChAdOx1 nCoV-19 Vaccination https://www.nejm.org/doi/full/10.1056/NEJMoa2104882?query=featured_home
The following protocol has been recently proposed as a potential approach to these patients at one UK hospital. Please note, this protocol advises a clinical assessment BEFORE a laboratory assessment. Use your noggins and order further tests only when necessary, following this initial assessment. Do not be fooled into ordering a load of bloods at triage, which will cause you heartache when they all come back mildly raised and you have no idea what to do next.
We do not necessarily endorse this pathway and present it only as an example. We would advise you to discuss and agree your pathway locally, with MDT support.