Vaccine-induced immune thrombotic thrombocytopenia (VITT)

Question 3 from the second paper of 2021 asked about VITT in some considerable detail. This was an unexpected and welcome departure from the routine practice, where the appearance of the SAQ lags the trends in evidence and prevalence by about a year or so. For this SAQ from 2021, the most effective revision resource would have to be the TSANZ statement on VITT from August 2021. The UpToDate article is also good, made all the better by the uncharacteristically socialist decision to make it free to access. Greinacher et al (2021) and Schultz et al (2021) are also free (courtesy of NEJM). As it is not clear whether NEJM and UpToDate will continue to extend this courtesy after the kudos-rich pandemic period is behind us, Chen (2021) or Macris et al (2021) are offered to the reader as alternatives.

Mechanism of pathogenesis

At a fundamental level, this is an immune-mediated thrombocytopenia. Schultz et al (2021) described it as "a rare vaccine-related variant of spontaneous heparin-induced thrombocytopenia". The mechanism and pathophysiology can be described as follows:

• An adenoviral vector vaccine is administered
• This vaccine contains vast amounts of different original viral molecules
• Some of these molecules (perhaps viral RNA) probably bind to PF4, a small 70-amino-acid protein on the surface of platelets, which binds with high affinity to polyanionic molecules like heparin (its role is to neutralise naturally occurring heparin within the vascular endothelium)
• The complexes between PF4 and these vaccine molecules act as a neoantigen.
• This results in the formation of autoantibodies to the PF4-vaccine complexes.
• These are usually IgG, and their epitope is different from the HITT epitope (i.e. they bind to a different part of the PF4 molecule).
• The result of binding of antibodies to these neoantigens produces the clinically important. combination of thrombocytopenia and thrombosis (because it both activates the platelets, and decreases their circulating lifespan).

Why only the adenoviral vaccines? Why anything. However, it is clear that the mRNA vaccines are not likely to cause this, because of the hundreds of millions of doses, there have been no case reports of thrombocytopenia and thrombosis, apart from a single case report with the Moderna vaccine, which could also have been spontaneous HITT. McGonagle et al (2021) dive deep into the immunology of this thing, and their article is full of some interesting theories. An unforgivably glib reinterpretation would focus on the promiscuity of PF4 molecules, which can interact with all kinds of negatively charged substrates, among which is DNA and RNA. It is thought that the adenoviral nucleotides bind to this receptor and - very rarely - break self-tolerance, resulting in an autoimmune reaction.

Risk factors for VITT

Exposure to the two implicated adenoviral vector vaccines is the main risk factor for VITT.

These are:

• AstraZeneca (ChAdOx1 nCov-19) vaccine
• Johnson & Johnson (Ad26.COV2. S) vaccine

In addition to this, "female sex and younger age" were proposed as risk factors; some case series reported mainly female and mainly younger patients ( younger than 55 years of age). Older and male-er patients are also being reported, and because the case numbers are so low, we may never know whether there is a real trend, or just some kind of play chance.

Clinical presentation

Question 3 from the second paper of 2021 specifically asked for an outline of the clinical presentation, which suggests that they were mainly looking for the constellation of signs and symptoms which would be found on the initial assessment of the patient, rather than laboratory data or the response to therapy. However, sometimes, when the college examiners say "clinical presentation", they also mean "biochemistry and imaging". 

Thus:

• History:
  • AstraZeneca (ChAdOx1 nCov-19) or Johnson & Johnson (Ad26.COV2. S) vaccine within 4-42 days of the presentation
  • Headache 
  • Abdominal pain
  • Decreased level of consciousness
  • Visual changes
  • Seizures
  • Leg pain or swelling
  • Unexplained bleeding, eg. GI bleeding or retroperitoneal
• Examination:
  • Clinical features of DVT or PE
  • Cranial nerve signs supportive of cavernous sinus thrombosis
  • Features of stroke
  • Ascites, features of liver failure
  • Petechii, ecchymoses
• Laboratory data:
  • Thrombocytopenia
  • Elevated D-dimer
  • Evidence of organ dysfunction (eg. deranged LFTs)
• Imaging
  • Finding of thrombosis (CVT, portal vein, or even arterial)

This is based on TSANZ, who give the following list of Ts and Rs:

• Timing: Onset days 4-42 after vaccination (this is based on case report data rather than some sort of scientific physiological reasoning; the peak period seems to be 6-16 days after the vaccination)
• Thrombosis: cerebral venous sinus thrombosis or something splanchnic, like a portal vein thrombosis
• Thrombus markers:  the D-dimer is usually very high
• Thrombocytopenia: no specific values are given, but a rapid fall in platelets is an essential feature
• Refractory to standard anti-coagulation
• Response to IVIG is sometimes seen

TSANZ is not the only body making these guidelines and recommendations. For example, the American Society of Hematology have this list of criteria, all five of which must be satisfied:

1. COVID vaccine 4 to 42 days prior to symptom onset
2. Any venous or arterial thrombosis (often cerebral or abdominal)
3. Thrombocytopenia (platelet count < 150 × 10⁹/L)
4. Positive PF4 “HIT” (heparin-induced thrombocytopenia) ELISA
5. Markedly elevated D-dimer (> 4 times upper limit of normal)

Differential diagnosis of VITT

VITT is rare, whereas thrombocytopenia in general is common and has numerous differentials. Any discussion of VITT would have to include some mention of the mechanisms one might use to exclude the much more common differentials. Statistically, the most likely explanation for all these clinical features is just some other cause of thrombocytopenia in a person who also happens to have received a vaccination in the last forty days. The rapid escalation of vaccine delivery around the world has massively increased the size of the susceptible population, and apart from the vanishingly small risk of VITT all of these people have brought with them all of their pre-existing risk factors for thrombocytopenia, their huge spleens and haematological malignancies and their alcoholism. In short, for exam purposes, you need to demonstrate that you will not jump on the zebra, thoughtfully considering the other possibilities. 

• It's just COVID. Venous thrombosis is a common occurrence with this disease. A highly cited analysis by Malas (2020) found a total rate of 20% or so for DVTs among hospitalised patients (28% if you end up in ICU), which would be about double the usual rate. Anti-anti-vaxxer counter-propaganda by the RAGP often trumpets that the risk profile greatly favours vaccination, as the risk of thrombosis for community COVID is ten times greater with COVID infection than it is with the vaccine. This is based on studies such as Taquet et al (2021), who found a massively increased risk of cerebral venous sinus thrombosis and portal vein thrombosis with COVID-19 infection (42.8 per million for CVT, which still sounds rare, but is in fact much greater than the population average, which is usually reported as 2-5 per million per year).
• It's just a common cause of thrombocytopenia- something else has produced a fall in platelets, and the thrombosis is a separate and unrelated phenomenon (because people develop thrombi in the ICU all the time, for a variety of non-vaccine-associated reasons). The possible causes of thrombocytopenia are detailed elsewhere.
• Antiphospholipid syndrome with thrombocytopenia, or any other disorder which promotes thrombosis (eg. a previously unrecognised Factor V Leiden mutation or a protein C or S deficiency) needs to be considered
• HIT, TTP or ITP - the incidence of vaccination and the incidence of COVID infection in the community may be so great that the chances of encountering a vaccinated person who has developed TTP or ITP for some unrelated reason is very high. Similarly, heparin use is common (basically mandatory) for hospital in-patients, which places HIT high on the list of differentials. These more common differentials need to be excluded. From the helpful UpToDate table, the following useful distinguishing characteristics can discriminate between these differentials:
  • VITT causes macro thrombosis, whereas the others don't tend to.
  • D-dimer is massively increased only in VITT (whereas it might be modestly elevated in the others)
  • Only TTP has a measurable ADAMTS13 deficiency, and only VITT has characteristic anti-PF4 ELISA assay findings
  • VITT does not usually have the petechii or purpura of ITP or TTP, and typically does not present with the "pentad" of confusion, fever, renal failure, etc (though to be fair nor does TTP)

From these differentials, it follows that a fairly broad range of tests might need to be launched. Obviously not everyone needs an ADAMTS13 level, but the list of investigations needs to reflect the need to exclude all the usual suspects. Thus:

Investigations to confirm VITT and exclude differentials

• Screening tests, for when VITT is suspected, are:
  • FBC for platelet count, to confirm that platelets are less than 150×10⁹. TSANZ are not convinced if the platelets are higher than this, but if the other screening tests are suspicious, repeating the platelet count on the next morning is recommended. 
  • D-dimer (to demonstrate that it is elevated at least five times the upper limit of normal)
  • Fibrinogen level, to demonstrate that it is reduced
  • Imaging for thrombosis, eg. a CT venogram of the brain or abdomen, an ultrasound of the portal vein, CTPA, etc.
  • If thrombosis is revealed by the screening tests, VITT is probable; in the sense that whatever this is, TSANZ wants you to treat it like VITT. 
• Specific VITT testing:
  • Antigen-based VITT immune assay needs to be performed. This is not something your local pathology technician should be expected to do on their own. Four clot tubes and four citrated tubes need to be collected and sent to the kind of haematology service that could perform a functional antibody test and a specific ELISA for VITT. The test is sufficiently unique that in NSW, at the time of writing in November of 2021, it cannot be ordered on any sort of standard pathology stationary or ordering system, and a specific request form needs to be filled out. 
• Exclusion of differentials will depend on the scenario, and could include:
  • A repeat FBC in a citrated tube, looking for pseudothrombocytopenia
  • ADAMTS13, if there is enough material to suspect TTP (eg. renal failure, confusion or seizures)
  • HIT anti-PF4 ELISA immunoassay if the patient otherwise fulfils the criteria for HIT
  • HIT antiPF4 antibodies
  • COVID swab, obviously
  • Spleen ultrasound, looking for splenomegaly
  • Lupus anticoagulant, looking for APLS

Supportive and specific management

TSANZ subclassifies the VITT presentations into possible and probable, reflecting that confirmatory tests need to be sent away and that a definitive diagnosis may take some days. The rationale for this is that one should not waste time, and start treatment while waiting for these tests (as a delay in treatment may be criticised, considering this condition can have a 40% mortality).

• Management of possible  VITT (i.e. there is no thrombosis, only suspicion)
  • "Anticoagulation with a non-heparin anticoagulant should be considered" was the TSANZ take on this, which leaves the decision in the hands of the clinician. 
  • If you do give anticoagulation, it should only be until the VITT ELISA is back (and if it is negative, you can cease the anticoagulation). 
  • VTE prophylaxis should be with a non-heparin agent, eg. fondaparinux
  • Platelet transfusion should be avoided
• Management of probable VITT (i.e where thrombosis is confirmed)
  • These patients need some sort of anticoagulation, as they do have a thrombosis. TSANZ recommend anticoagulation with a non-heparin agent, and they don't recommend any specific agent (fondaparinux, argatroban, bivalirudin, etc). The total duration of anticoagulation should probably be something like 3-6 months.
  • IV immunoglobulin (1-2g/kg, split over 2 days) is recommended
  • High dose methylprednisolone and/or plasma exchange is mentioned as something optional, to be considered if the first-line therapy is not effective

The management of probable VITT is basically the same as the management of confirmed VITT; with the exception of anticoagulation drug choices. If VITT is not confirmed, i.e. if the ELISA comes back negative, one could theoretically continue to manage the thrombus conventionally, i.e with heparin.

Choice of anticoagulant

Why do we want to avoid heparin? There is a belief that in "true" VITT, the use of heparin can produce enhanced platelet activation. This comes from Greinacher et al (2021), who recommended people use non-heparin agents, "given the parallels with autoimmune heparin-induced thrombocytopenia". The implication is that all subsequent guidelines have been recommending the use of non-heparin anticoagulants mainly because this condition resembles HIT, rather than any specific data. Certainly, if you accidentally happen to treat VITT with heparin, the patient does not seem to come to any heparin-attributable harm. UpToDate authors point to a 220-patient case series by Pavord et al (2021), who reported no major difference in outcomes for VITT patients who received heparin while the diagnosis had not been established or even considered. There was a mortality difference (20 vs 16%), but this could be attributed to other things, and none of the heparinised patients seemed to develop any major complications, which suggests that heparin might in fact be perfectly fine for VITT. Still, local guidelines reflect the opinions of thrombosis gurus, who hold that on the balance of things it is safer to give a non-heparin product (considering that, at the point of making that decision, often HIT is also one of the differentials).

Immunoglobulin and plasmapheresis

According to Lentz (2021), the use of polyclonal immunoglobulin for VITT is also recommended on the basis of the rationale that VITT and HIT are very similar, and IVIG seems to work in HIT. Theoretically this makes sense. Some abnormal adenoviral RNA fragments are causing all the problems, and there's a lot of COVID in the community, which means there should be plenty of antibodies in donated blood which should specifically target those RNA fragments. Wiping them out of the bloodstream should theoretically prevent them from binding to the PF4 protein and therefore prevent worsening thrombocytopenia and thrombosis.  Beyond that, IVIG can also block platelet activation by anti-PF4 antibodies by competing for binding to Fcγ receptors, which is the platelet surface receptor responsible for platelet activation in VITT. For this competitive effect to be useful, a fairly large amount of immunoglobulin needs to be given - 1-2g/kg over two days is the recommended dose. Logically, this therapy should follow plasmapheresis. 

Post-acute management

Though follow-up and long term management of these patients in the community is really not the province of intensivists, they should probably still be aware of what tends to be the standard. Specifically, they should be aware that nobody really knows how long the anti-PF4 antibodies can hang around, which means the patient may be at some hard-to-estimate risk of recurrence for a prolonged period of time. Local guidelines recommend surveillance for thrombocytopenia and repeat ELISA and functional testing at 6 weeks, 3 months and 6 months.