Malignant middle cerebral artery syndrome with thrombotic thrombocytopenia following vaccination against SARS-CoV-2

Abstract

Cases of thromboses at unusual sites with thrombocytopenia have been reported following vaccination against Sars-CoV-2. This new syndrome, christened vaccine-induced thrombotic thrombocytopenia (VITT), mainly results in venous thromboses. We report the case of a young woman with a right middle cerebral artery stroke following vaccination with ChAdOx1 nCoV-19. A diagnosis of VITT was made and platelet counts began to recover shortly after commencing treatment with argatroban, intravenous immunoglobulins and corticosteroids. On day 6 following admission, the patient deteriorated neurologically and decision made to proceed with decompressive hemicraniectomy. There were no perioperative complications and anticoagulation with argatroban was reinitiated on the first postoperative day.

VITT is a rare condition resembling auto-immune heparin-induced thrombocytopenia. All critical care staff should be aware of the rare link between vaccination against SARS-CoV-2 and VITT and the need to rapidly commence both anticoagulation, using heparin alternatives, and immunomodulation.

Introduction

Within 15 months of identification of the novel coronavirus (SARS-CoV-2), almost a billion doses of more than a dozen different vaccinations against Covid-19 have been made available globally.^1,2 Recent reports of rare severe adverse reactions involving thrombocytopenia and thrombosis following vaccination with the ChAdOx1 nCoV-19 vaccine have raised questions about the unknown and unpredictable potential of harm for some individuals. As of 31st March 2021, 79 cases of thrombocytopenia and thrombosis following vaccination with ChAdOx1 nCoV-19 have been reported in the UK, 19 of which had a fatal outcome. ³ By 13th April 2021, 6 similar cases have been reported following vaccination with Ad26.COV2 in the U.S.^4,5

On 19th March 2021, the identification of antibodies against platelet factor 4 causing extensive platelet activation in patients presenting with thrombosis and thrombocytopenia following vaccination with ChAdOx1 nCoV-19 was reported. ⁶ The clinical and serological picture of these patients shared features with heparin-induced thrombocytopenia (HIT), however exposure to heparin was uncommon in reported case series.^7,8 Initially, the term “vaccine-induced prothrombotic immune thrombocytopenia” (VIPIT) was proposed, ⁹ and later agreed to “vaccine-induced immune thrombotic thrombocytopenia” (VITT).^7,8

We present a case of a 39-year-old woman who was admitted with a right middle cerebral artery (MCA) territory stroke and thrombocytopenia nine days after vaccination against SARS-CoV-2. This case highlights the unique complexity of managing a patient with a prothrombotic and thrombocytopenic condition in the context of malignant MCA syndrome.

Case summary

A 39-year-old woman presented to her local hospital with acute onset, left-sided weakness, confusion, headache and nausea mid-March 2021 (day 0). She had received her first covid-19 vaccination (ChAdOx1 nCoV-19) nine days previously due to high occupational risk. Prior to admission she was in good health and took no regular medications. She had a history of migraines and had been identified as a BRCA2 carrier. On examination, the patient had a Glasgow Coma Scale (GCS) score of E4V4M6 with a dense left-sided hemiparesis, positive Babinski sign and National Institutes of Health Stroke Scale (NIHSS) Score of 15. Laboratory investigations revealed a thrombocytopaenia of 66 × 10⁹/L and mild hyperbilirubinemia of 50 µmol/L. A computerised tomography (CT) scan of the head was performed, showing a large volume of hypodensity corresponding to the right middle cerebral artery (MCA) territory. There was evidence of early resultant mass-effect with 6 mm midline shift (see Figure 1(a)). CT angiogram revealed complete occlusion of the right M1 segment of the MCA. The patient was given 300 mg of aspirin and subsequently transferred to the Hyper Acute Stroke Unit in the regional tertiary care centre, in accordance with local guidelines for management of patients at risk of developing life-threatening malignant MCA syndrome.

Figure 1.

Panel of serial CT head scans for the patient (a–c). Panel (a) was performed on day 0 of admission, panel (b) on day 2, and panel (c) on day 5. The serial scans show a gradual maturation of the right MCA territory infarct, with increase in the midline shift in correlation with the patient’s deterioration in clinical signs. In addition, development of haemorrhagic transformation of the infarct can be seen in panels (b) and (c).

Collateral history from the patient’s family revealed the patient had been unwell for two days prior to admission and had found herself unable complete her usual 25 minute walk. She had to call family to help her to get home because of pain in her leg. The patient is a non-smoker with no identified cardiovascular risk factors apart of a history of migraines. Initial investigations to identify the aetiology of the stroke were inconclusive (see Table 1). No thrombus was identified on transthoracic echocardiogram. A CT angiogram of the neck on day 2 of admission revealed a large intraluminal thrombus extending superiorly from the origin of the right internal carotid artery, measuring approximately 4 cm and resulting in moderate to severe stenosis. A repeat CT brain on day 2 of admission showed increasing mass effect (see Figure 1(b)).

Table 1.

Overview of investigations.

Test	Day	Result	Reference range
SARS-CoV-2 PCR	D0	Negative
Urine drug screen	D1	Opiates detected
HIT ELISA (Immucor LIFECODES PF4)	D2	Optical density 2.4566	<0.4
HIT chemiluminescent immunoassay (HemosIL® AcuStar HIT)	D2	Negative
Platelet activation assay	D9	Negative
HDL cholesterol	D1	1.11 mmol/L	0.91–2.21 mmol/L
Cholesterol/HDL ratio	D1	4.9	1.0–4.0
D-dimer assay (see Figure 2)	D1	>5000 µg/L	0–243 µg/L
Anticardiolipin G	D1	<1.6 U/mL	0–20 U/mL
Anticardiolipin M	D1	0.5 U/mL	0–20 U/mL
Lupus anticoagulant ratio	D1	1.28
ANA screen	D1	Negative
Serum electrophoresis	D1	Band in gammaglobulin region – IgG (Lambda) abnormality with monoclonal component of 2.2 g/L
HIV 1 & 2 antibody/ p24 antigen	D1	Not detected
Hep B sAg	D1	Not detected
Hep C IgG	D1	Not detected
Syphilis antibody	D1	Not detected
EBV nuclear antigen IgG	D1	Detected
EBV serology	D1	Consistent with past infection

Blood tests showed worsening thrombocytopenia (57 × 10⁹/L). A haematological opinion was sought and further investigations were requested (see Table 1). Pathological findings included a strongly positive HIT screen with ELISA (Immucor LIFECODES PF4; optical density 2.4566, cut off for positive result >0.4) with a negative HIT screen using a chemiluminescent immunoassay (HemosIL® AcuStar HIT). There was no recent exposure to heparin. Electrophoresis showed a band in the gammaglobulin region, which was identified as IgG lambda abnormality with a scanned monoclonal component of 2.2 g/L.

A diagnosis of vaccine-induced thrombotic thrombocytopenia (VITT) was made. Treatment with intravenous immunoglobulins (IVIg, 1 g/kg) and methylprednisolone (100 mg, followed by 50 mg OD) was initiated. The patient was transferred to the Neurosciences Intensive Care Unit for anticoagulation. In anticipation of a potential need for decompressive hemicraniectomy, argatroban was chosen as anticoagulant agent, based on its relatively short elimination half-life of 45 minutes ¹⁰ and efficacy in heparin-induced thrombocytopenia. ¹¹ Following a cautious regimen, dosing was started at a rate of 0.5 microgram/kg/min with a target activated partial thromboplastin time ratio (APTR) range of 1.5–2.5. APTR was measured two-hourly and increased by 0.1 microgram/kg/min whilst APTR was subtherapeutic. After three dose changes, the dosing regimen was adjusted and based on the standard regimen, where the dose was adjusted in increments of 0.5 microgram/kg/min.

Haematological progress

Less than 12 h after initiation of treatment with IVIg and corticosteroids, platelets numbers started to recover. Four days after initiation of treatment, a marked decrease in D-Dimer was noted (see Figure 2). The patient appeared to develop a degree of argatroban resistance, mandating frequent increases in argatroban dosing to achieve the target APTR (see Figure 3).

Figure 2.

Platelet count and D-dimer levels. Arrow indicates time of treatment initiation, asterisk marks time of surgery.

Figure 3.

Argatroban dose and APTR. Asterisk indicates time of surgery.

HIT Screen was repeated after 1 week and remained strongly positive. In a platelet activation assay (HIT Alert, Diapharma) performed on samples taken on day 9, the patient’s serum did not activate healthy donor platelets. Platelets reached normal levels on day 13 of admission and argatroban switched to prophylactic dose fondaparinux (2.5 mg OD) and antiplatelet therapy with low-dose aspirin (75 mg OD) initiated as recommended by current guidance from the British Society for Haematology (BSH). ¹²

Neurological progress

The patient was monitored closely with hourly observations including level of consciousness and limb power/sensation. During the initial period, she remained alert with a GCS Score of E4V4M6. Interval CT imaging showed gradually increasing mass effect and small areas of haemorrhagic transformation. Extensive discussions took place between the stroke, neurocritical care and neurosurgical teams regarding the appropriate course of action given the possible need for decompressive surgery in the context of on-going anticoagulant therapy and reduced platelet count. Five days after admission to hospital, the patient deteriorated clinically and became more difficult to rouse with a GCS score of E3V4M6. An interval CT scan showed an increase in the degree of midline shift and evidence of haemorrhagic transformation of the infarcted region (see Figure 1(c)).

Decision was made to proceed with decompressive hemicraniectomy in light of clear neurological deterioration secondary to malignant MCA syndrome. Argatroban was stopped 1 hour before surgery and 1 unit of platelets transfused to achieve a platelet count >100 10⁹/L perioperatively. A 13 cm bone flap was removed and placed in a subcutaneous abdominal pouch to allow for autologous cranioplasty at a later date.

On day 1 post-operatively, a further CT head showed stable appearances of the haemorrhagic focus. The operative site was full but soft, and a decision was made to re-start the argatroban treatment. She was extubated on post-operative day 2 and was subsequently transferred to the stroke ward for ongoing rehabilitation on post-operative day 8.

She continued to make steady progress towards recovery with active participation in rehabilitation. At the time of transfer to her local stroke unit on day 21, significant left hemiparesis with hemisensory neglect remained. The patient remains in hospital for ongoing rehabilitation.

Discussion

To our knowledge, this is the first reported case of arterial infarction requiring neurosurgical intervention in the context of VITT syndrome. Patients with VITT syndrome typically present with progressive venous thromboses at unusual sites, with a predominance of cerebral venous sinus thrombosis.^12,13 In three recent case series of VITT patients, arterial thrombosis has so far only been reported in two other patients (Patient 1, aortoiliac, Greinacher et al; Patient 2 Scully et al.; Patient 16 in same is the patient reported here).^7,8,14

In this patient, early recognition of probable VITT syndrome and prompt initiation of treatment with IVIg and methylprednisolone allowed successful hemicraniectomy on day 6 following admission. The patient showed a swift response to immunomodulatory treatment, with a 50% increase in platelet count 36 h after treatment initiation. The use of IVIg in VITT syndrome is supported by observations in vitro indicating immunoglobulin-mediated reduction of platelet activation in the presence of serum from VITT patients, ⁷ as well as clinical experience in the treatment of autoimmune HIT ¹⁵ and refractory HIT. ¹⁶

Given the pathophysiology and concerns about aggravating extensive platelet activation, advice is generally to avoid platelet transfusions in VITT syndrome. ¹² Additional thromboses have been observed in patients with VITT syndrome who have received platelet transfusions. ¹⁴ Where transfusions are required (i.e. for urgent neurosurgical interventions mandating platelet counts > 100 × 10⁹/L), platelets should be transfused after or with concomitant IVIg treatment. ¹² In this case, four days had passed since initiation of IVIg treatment when a preoperative platelet transfusion was required. At this point, platelet count had increased fourfold from its nadir (87 × 10⁹/L from 22 × 10⁹/L) and D-Dimer levels were falling (see Figure 2). No adverse events following transfusion were observed.

The benefits and ideal timing of decompressive surgery in space-occupying cerebral infarction have been the subject of several clinical trials.^17–19 A meta-analysis of patients enrolled in these trials confirmed reduced poor clinical outcome and case fatality in patients who underwent surgical intervention within 48 h of stroke. Surgical decompression up to 96 h after ictus reduced mortality, but did not improve clinical outcome. ¹⁷ In this case, surgical intervention was performed on day 6 (approx. 140 h) following stroke. Given her high perioperative risk due to abnormal coagulation, there was consensus that a watch and wait approach with close observation was more appropriate than early surgical intervention. This allowed marked reduction of perioperative risk. At the time of surgery, platelet counts had been rising for five consecutive days and falling D-Dimer levels were indicative of a resolving thrombotic process.

As in previous reported cases, ELISA testing for HIT antibodies (Immucor LIFECODES PF4) was strongly positive, whilst a chemiluminescent immunoassay (HemosIL® AcuStar) was negative. ¹⁴ A platelet activation assay was performed on day 9 and negative. Heparin-independent platelet activation has previously been described in the context of spontaneous HIT syndrome, ²⁰ however these cases typically showed both heparin-dependent and heparin-independent platelet activation. ¹⁵ In contrast, physiologic dose heparin (0.3 U per millilitre) does not increase platelet activation in VITT patients. ¹⁴

Given the high risk of intracranial haemorrhage, the patient was treated with a tighter APTR target range (1.5–2.5) compared to BSH’s HIT guidelines (1.5–3). ²¹ At the time, it was unclear why the patient seemed to develop argatroban resistance, mandating almost fourfold dose increases over a period of 11 days. Interestingly, we have observed similar patterns in another patient treated in our unit with VITT in previous weeks. Since then, it has become apparent that patients with VITT syndrome often present with markedly raised factor VIII levels (in this case 304 u/dL on day 17 of admission), which impairs the correlation between argatroban effects and APTR. The British Expert Haematology Panel therefore currently recommend direct thrombin inhibition assays for monitoring of argatroban in VITT patients. ¹²

A raised paraprotein, as found in this patient, has also been identified in two other cases of VITT syndrome in our centre. A transient paraproteinemia reflective of an inflammatory process is suspected. ²²

During this admission, lupus anticoagulant was mildly elevated. Five years ago, the patient had a negative antiphospholipid screen when investigated for two miscarriages. In one of five reported cases of a recent case series of VITT patients, antiphospholipid antibodies were identified. ⁸ The significance of this is unclear. Underlying thrombophilia and/or a history of thromboses are not thought to be a risk factor for VITT, given the presumed immunological pathomechanism of the disorder. ²³

Conclusion

VITT syndrome is a newly described condition resembling autoimmune HIT. Clinicians must be aware of this syndrome, as it is likely that more patients will present in the following months given the association with vaccinations against SARS-CoV-2. Early recognition is crucial to avoid potential harm associated with heparin treatment or platelet transfusions, as is prompt initiation of treatment with IVIg. In the case presented, early diagnosis and close collaboration between stroke, haematological, neurocritical care and neurosurgical teams allowed successful neurosurgical intervention and a favourable clinical outcome.

More research is needed to gain better understanding of pathogenesis and possible risk factors for developing this disorder, and to identify optimal treatment strategies both in the acute and possibly chronic setting.