Ischemic Stroke and Vaccine-Induced Immune Thrombotic Thrombocytopenia following COVID-19 Vaccine: A Case Report with Systematic Review of the Literature

Abstract

Introduction

Vaccine-induced immune thrombotic thrombocytopenia (VITT) is a prothrombotic syndrome observed after adenoviral vector-based vaccines for severe acute respiratory syndrome coronavirus 2. It is characterized by thrombocytopenia, systemic activation of coagulation, extensive venous thrombosis, and anti-platelet factor 4 antibodies. Arterial thrombosis is less common and mainly affects the aorta, peripheral arteries, heart, and brain. Several cases of ischemic stroke have been reported in VITT, often associated with large vessel occlusion (LVO). Here, we describe a case of ischemic stroke with LVO after Ad26.COV2.S vaccine, then we systematically reviewed the published cases of ischemic stroke and VITT following COVID-19 vaccination.

Methods

We describe a 58-year-old woman who developed a thrombotic thrombocytopenia syndrome with extensive splanchnic vein thrombosis and ischemic stroke due to right middle cerebral artery (MCA) occlusion, 13 days after receiving Ad26.COV2.S vaccination. Then, we performed a systematic review of the literature until December 3, 2021 using PubMed and EMBASE databases. The following keywords were used: (“COVID-19 vaccine”) AND (“stroke”), (“COVID-19 vaccine”) AND (“thrombotic thrombocytopenia”). We have selected all cases of ischemic stroke in VITT.

Results

Our study included 24 patients. The majority of the patients were females (79.2%) and younger than 60 years of age (median age 45.5 years). Almost all patients (96%) received the first dose of an adenoviral vector-based vaccine. Ischemic stroke was the presenting symptom in 18 patients (75%). Splanchnic venous thrombosis was found in 10 patients, and cerebral venous thrombosis in 5 patients (21%). Most patients (87.5%) had an anterior circulation stroke, mainly involving MCA. Seventeen patients (71%) had an intracranial LVO. We found a high prevalence of large intraluminal thrombi (7 patients) and free-floating thrombus (3 patients) in extracranial vessels, such as the carotid artery, in the absence of underlying atherosclerotic disease. Acute reperfusion therapy was performed in 7 of the 17 patients with LVO (41%). One patient with a normal platelet count underwent intravenous thrombolysis with alteplase, while 6 patients underwent mechanical thrombectomy. A malignant infarct occurred in 9 patients and decompressive hemicraniectomy was performed in 7 patients. Five patients died (21%).

Conclusion

Our study points out that, in addition to cerebral venous thrombosis, adenoviral vector-based vaccines also appear to have a cerebral arterial thrombotic risk, and clinicians should be aware that ischemic stroke with LVO, although rare, could represent a clinical presentation of VITT.

Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has been associated with an increased risk of arterial and venous thromboembolic complications, probably mediated by an abnormal hypercoagulative and inflammatory state. Several vaccines against SARS-CoV-2 have been developed and administrated worldwide within a year of the COVID-19 outbreak. These vaccines are divided into two main categories: the messenger RNA (mRNA) − based vaccines (including BNT162b2 [BioNTech/Pfizer] and mRNA-1273 [Moderna]) and the adenoviral vector − based vaccines (including Ad26.COV2.S [Janssen/Johnson & Johnson] and ChAdOx1 nCoV-19 [AstraZeneca − Oxford]). Since the start of vaccination, a prothrombotic syndrome characterized by severe thrombocytopenia, systemic activation of coagulation, and extensive venous thrombosis has been observed especially after ChAdOx1 nCoV-19 vaccine [1, 2, 3] and in a few individuals after the Ad26.COV2.S vaccine [4, 5]. This prothrombotic syndrome is similar to the heparin-induced thrombocytopenia (HIT), with high levels of antibodies to anti-platelet factor 4 (PF4) and has been designated vaccine-induced immune thrombotic thrombocytopenia (VITT) or thrombosis with thrombocytopenia syndrome. Thrombosis occurs at unusual sites including cerebral venous thrombosis (CVT) and splanchnic vein thrombosis (SVT), in addition to deep venous thrombosis (DVT) and pulmonary embolism [6]. Arterial thrombosis appears to be less frequent and mainly affects the aorta, peripheral arteries, heart, and brain. The first cases of ischemic stroke described during VITT were often associated with large vessel occlusions (LVOs) and negative outcomes [7, 8]; since then, several other reports have been published. Here we describe a case of ischemic stroke associated with severe thrombocytopenia and extensive SVT after the Ad26.COV2.S vaccine, and then we performed a systematic review of the literature searching for cases of ischemic stroke and thrombotic thrombocytopenia after COVID-19 vaccination. In this study, we aim to describe the demographic, clinical characteristics, and the outcome of patients with stroke in the context of VITT.

Case Presentation

A 58-year-old white Caucasian obese woman, suffering from hypertension, presented to our hospital having been found unconscious and hypotensive at home. She had received the Ad26.COV2.S vaccine 13 days earlier and started complaining of abdominal pain a few days before admission. A thoraco-abdominal computed tomography (CT) scan showed extensive SVT (involving portal, hepatic, splenic, superior and inferior mesenteric veins), DVT (involving left iliac and femoral veins), spleen infarction, and hemoperitoneum. The patient underwent urgent explorative laparotomy with blood drainage and splenectomy. Blood and platelet transfusion and fresh frozen plasma were administered shortly before the emergency surgery. Several hours later, the patient woke up with left hemiparesis, neglect, and right gaze preference. Non-contrast CT and CT angiography (CTA) revealed occlusion of the M1 segment of the right middle cerebral artery (MCA) with extensive ischemic changes and 7 mm of midline shift due to a datable infarct for at least several hours. An urgent decompressive hemicraniectomy was performed. Blood test revealed severe thrombocytopenia (platelet count 18 × 10⁹/L, range 140–440), high D-dimer 35,000 μg/L fibrinogen equivalent units (FEU) (range 0–550) and low fibrinogen; antithrombin activity was normal. Anti-PF4 IgG antibodies enzyme-linked immunosorbent assay (ELISA) and heparin-induced platelet aggregation assay were negative. Coagulation tests (INR, PT, aPTT), anti-phospholipid antibodies, ADAMTS-13, von Willebrand Factor, and blood smear were normal. The SARS-CoV-2 real-time reverse-transcriptase polymerase-chain-reaction test on a nasopharyngeal swab was negative. Considering the clinical scenario, temporal association, thrombosis location and laboratory findings, VITT was suspected. Therefore, intravenous immunoglobulin (1 g/kg/day for 2 days) and high-dose dexamethasone (40 mg for 4 days) were administered and anticoagulation with argatroban was started 2 days after hemicraniectomy. Transthoracic echocardiography and transcranial doppler with bubble test excluded cardiac embolic sources and paradoxical embolization. Then the patient clinically improved, platelet count increased and returned to normal (range 140–440 × 10⁹/L). After serial imaging re-evaluations (through cranial CT) of the cerebral ischemic lesion excluding its hemorrhagic transformation, fondaparinux treatment (7.5 mg daily) was initiated and, after imbrication, anticoagulation with warfarin continued (Fig. 1).

Fig. 1.

Main key findings from our case report, serial platelet count, and D-dimer course in relation to IVIg, steroid, and anticoagulant treatment.

Methods

Study Selection Criteria

A systematic review was performed, applying the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [9] (shown in Fig. 2, PRISMA checklist available as online suppl. material; for all online suppl. material, see www.karger.com/doi/10.1159/000524290). Full-text articles were selected from a comprehensive search of PubMed and EMBASE databases. The following keywords were combined in each database as follows: (“COVID-19 vaccine”) AND (“stroke”), (“COVID-19 vaccine”) AND (“thrombotic thrombocytopenia”). No filters were applied on the publication date of the articles, and all results from each database were included until December 3, 2021. Data from each article were extracted into the Microsoft Excel software. We identified 751 studies. After the duplicates were removed, 484 studies were screened. All articles were evaluated through a reading of titles and abstracts. Articles in any other language other than English, reviews, guidelines, paper statements, research studies, or records with nonrelevant topics were excluded. Overall, 155 records were subjected to an accurate reading to assess eligibility. Two researchers (A.C.R. and G.G.) independently searched and identified eligible studies. Controversies were resolved by the senior author (E.C.A.). Articles were included if they met the following inclusion criteria: (I) described patients who had an ischemic stroke after COVID-19 vaccination and a thrombotic blood disorder (II) included at least some information related to clinical features, imaging, and laboratory results. After an accurate revision of full manuscripts, data extraction was performed on 20 articles satisfying the inclusion criteria.

Fig. 2.

PRISMA flowchart of this study.

Statistical Analysis

Descriptive statistics were calculated for all included patients. Continuous data are presented as means and standard deviations or medians and ranges (as appropriate), categorical data as counts and proportions (as n/N [%], where n is the number of patients in which that variable is present and N the total number of patients for which that particular variable was reported).

Results

We identified 20 records describing patients with ischemic stroke and thrombotic blood disorder after COVID-19 vaccination. Two records were excluded because the stroke was due to secondary vaccine-induced immune thrombocytopenia (ITP) and thrombotic thrombocytopenic purpura (TTP), respectively [10, 11]. Eighteen articles (n = 23 patients) [3, 7, 8, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26] were included in our review. Baseline demographics, relevant clinical data, stroke characteristics, and outcome of patients described in these studies, including our case, are shown in Tables 1 and 2. Quantitative synthesis of the data is shown in Tables 3 and 4.

Table 1.

Published reports of patients with ischemic stroke following Covid-19 vaccination (including our case)

Author	Case No.	Patient (sex, yr)	SARS-CoV-2 RT-PCR test	Vaccine	Medical history	Time from vaccine to admission, days	Platelet count (nadir) Cells ×109/L	D-dimer (peak), FEU ng/mL	Fibrinogen (nadir), g/L	anti-PF4 (ELISA assay)	Platelet functional assay	Other thromboses diagnosed and/or hemorrhages	Case definition
Ref		–	–	–	–	–	150–400	0–550	1.5–4.0	–	–	–
Scully et al. [3]	1	F, 39	Neg	ChAdOx1 nCoV-19	–	10	57	>5,000	4.4	Pos	ND	–	Definite VITT
	2	M, 21	Neg	ChAdOx1 nCoV-19	–	10	113	22,903	1.0	Pos	ND	–	Definite VITT
D'Agostino et al. [12]	3	F, 54	Neg	ChAdOx1 nCoV-19	Meniere disease	12	Low	High	Normal	ND	ND	CVT, PE, PVT, HVT, coronary artery thrombosis, hemoperitoneum, adrenal, and cerebral hemorrhages	Probable VITT
Blauenfeldt et al. [7]	4	F, 60	NA	ChAdOx1 nCoV-19	Hypertension, dyslipidemia	7	5	106,200	2.31	Pos	Pos	Adrenal and renal hemorrhages	Definite VITT
Al-Mahyani et al. [8]	5	F, 35	NA	ChAdOx1 nCoV-19	–	11	64	11,220	NA	Pos	ND	PVT	Definite VITT
	6	F, 37	NA	ChAdOx1 nCoV-19	–	12	9	34,000	NA	Pos	ND	DVT, PE, CVT, IJVT, HVT	Definite VITT
	7	F, 43	NA	ChAdOx1 nCoV-19	–	21	48	24,000	NA	Pos	ND	–	Definite VITT
Garnieret al. [13]	8	F, 26	NA	ChAdOx1 nCoV-19	–	8	Low	NA	Low	Pos	Pos	PE, PSM, DVT	Probable VITT
Jacob et al. [14]	9	F, 39	Neg	ChAdOx1 nCoV-19	Migraine	9	<50	> 10,000	–	Pos	Neg	–	Definite VITT
Bourguignon et al. [15]	10	M, 69	NA	ChAdOx1 nCoV-19	Diabetes, hypertension, obstructive sleep apnea, prostate cancer	12	29	>20,000	2.0	Pos	Pos	CVT, IJVT, PE, DVT, HVT	Definite VITT
Walter et al. [16]	11	M, 31	Neg	ChAdOx1 nCoV-19	Smoking	8	152	1,100	2.3	Pos	Pos	–	Possible VITT
Tiède et al. [17]	12	F, 67	Neg	ChAdOx1 nCoV-19	Hypercholesterolemia	8	40	>35,200	2.7	Pos	Pos	–	Definite VITT
	13	F, 61	Neg	ChAdOx1 nCoV-19	–	9	25	>35,200	0.9	Pos	Pos	Popliteal artery thrombosis, upper extremity DVT, IJVT	Definite VITT
De Michele et al. [18]	14	F, 57	Neg	ChAdOx1 nCoV-19	–	9	23	>4,318	2.50	Pos	Pos	PE, PVT	Definite VITT
	15	F, 55	Neg	ChAdOx1 nCoV-19	–	10	59	31,646	3.22	Pos	Pos	PE, PVT	Definite VITT
Patriquin et al. [19]	16	F, 45	NA	ChAdOx1 nCoV-19	–	8	<50	>35,200	3.22	Pos	Pos	DVT, PE, CVT, renal infarct, bilateral adrenal, and venous cerebral hemorrhage	Definite VITT
Costentin et al. [20]	17	F, 26	NA	ChAdOx1 nCoV-19	OCT	7	17	NA	0.89	Pos	ND	PE, PVT	Probable VITT
Goereci et al. [21]	18	F, 42	NA	ChAdOx1 nCoV-19	–	9	40	35,000	NA	Pos	Pos	–	Definite VITT
Ceschia et al. [22]	19	F, 73	NA	ChAdOx1 nCoV-19	Hypercholesterolemia, hypertension	14	20	32,559	3.99	Pos	Pos	DVT, PE, CVT, renal vein thrombosis, popliteal artery thrombosis	Definite VITT
Rodriguez-Pardo et al. [23]	20	M, 46	NA	Ad26.COV2.S	Smoking, hypertriglyceridemia	14	125	Normal	Normal	Pos	Neg	–	Probable VITT
Ken da et al. [24]	21	F, 51	Neg	ChAdOx1 nCoV-19	Dyslipidemia	7	54	35,783	mildly elevated	Pos	Pos	–	Definite VITT
Su et al. [25]	22	M, 70	Neg	mRNA-1273	Atrial fibrillation, pancreatic cancer, hypertension, COPD	7	×109/L 17	28,550	(nadir), g/L 0.63 Normal 1.21	(ELISA assay) Pos Pos Neg	functional assay Pos NA Neg	– Upper extremity and femoral DVT PSM, DVT, hemoperitoneum, spleen infarction	Definite VITT Definite VITT Probable VITT
Charidimou et al. [26]	23	F, 37	Neg	Ad26.COV2.S	Migraine, OCT	10	30	80,000
Present case	24	F, 58	Neg	Ad26.COV2.S	Hypertension, obesity	13	18	35,000

COPD, chronic obstructive pulmonary disease; CVT, cerebral venous thrombosis; DVT, deep venous thrombosis; FEU, fibrinogen equivalent units; HVT, hepatic vein thrombosis; IJVT, internal jugular vein thrombosis; NA, not available; ND, not done; Neg, negative; OCT, oral contraceptive therapy; PE, pulmonary embolism; Pos, positive, PSM, porto-spleno-mesenteric venous thrombosis; PVT, portal vein thrombosis; Ref, reference range; VITT, vaccine-induced thrombotic thrombocytopenia; RT-PCR, reverse-transcriptase polymerase-chain-reaction.

Table 2.

Stroke features and outcome of published reports (including our case)

Author	Case No.	Symptoms on admission	Neurological stroke deficit	Vascular territory	LVO	Extracranial vessel findings	Acute reperfusion therapy	Diagnostic workup	Anticoagulation	Complications	Outcome
Scully et al. [3]	1	Stroke	NA	MCA	NA	NA	NA	NA	NA	NA	Alive
	2	Stroke	NA	MCA	NA	NA	NA	NA	NA	NA	Alive
D'Agostino et al. [12]	3	Left side sign5 GCS13	, Worsening coma, GCS 6	PCA, Cerebellar, Pons	Basilar	Aortic arch floating thrombus	–	NA	–	Malignant infarct, brain herniation	Death
Blauenfeldt et al. [7]	4	Abdominal pain	Hemiparesis	MCA	ICA	–	–	NA	Dalteparin	Malignant infarct, hemicraniectomy	Death
Al-Mahyani et al. [8]	5	Stroke	Hemiparesis, gaze preference, drowsiness	MCA	MCA-M1	–	–	NA	Fonda pa rinux	Hemorrhagic transformation, malignant infarct, hemicraniectomy	Death
	6	Stroke, headache	Arm weakness, hemianopsia, confusion	MCA (borderzone in fa rets)	Bilateral ICA	Bilateral ICA occlusion [postbulbous and at origin)		NA	Fonda pa rinux		Alive
	7	Stroke	Dysphasia	MCA	–	–	–	NA	Fonda pa rinux	Hemorrhagic transformation	Alive
Garnier et al. [13]	8	Stroke	Hemiplegia, aphasia (NIHSS 8)	MCA	MCA-M1	–	MT	NA	NA	Hemorrhagic transformation	Alive
Jacob et al. [14]	9	Stroke, headache	Left-sided weakness, confusion (NIHSS 15)	MCA	MCA-M1	Moderate to severe ICA stenosis due to large intraluminal thrombus at origin		TTE	Argatroban, fonda parinux plus aspirin	Hemorrhagic transformation, malignant infarct, hemicraniectomy	Alive
Bourguignon et al. [15]	10	Stroke, headache	Left-sided weakness, confusion	MCA	MCA	ICA thrombosis	–	NA	Fon da pa rinux, rivaroxaban	Hemorrhagic transformation	Alive
Walter et al. [16]	11	Stroke, headache	Hemiplegia, aphasia	MCA	MCA-M1	Carotid bulb floating thrombus	IVT	TEE, TCD-B	Aspirin, danaparoid, phenprocoumon	–	Alive
Tiede et al. [17]	12	Headache, stroke	–	r-MCA, l-PCA (small cortical infarcts)	–	Multiple thrombi in the aortic arch and CCA origin Multiple thrombi and cervical ICA occlusion	–	NA	Argatroban, apixaban	–	Alive
	13	Stroke, headache	Dysarthria, hemiplegia, conjugated gaze palsy (NIHSS 17)	MCA, ACA	ICA + MCA-M1		MT + CAS	NA	Argatroban	Hemorrhagic transformation, malignant infarct, hemicraniectomy	Alive
De Michele et al. [18]	14	Stroke	Hemiplegia, dysarthria, neglect	MCA	MCA-M1	–	MT	TEE, TCD-B	Fondaparinux	Malignant infarct, hemicraniectomy, MCA reocclusion	Alive
	15	Abdominal pain	Transient aphasia and hemiparesis, seizure,	Bilateral MCA	Right ICA, left MCA-M1	–	–	TEE	–	Malignant infarct	Death
			coma
Patriquin et al. [19]	16	Stroke	Confusion, decreased level of consciousness	Cerebellar	VA	Bilateral VA occlusion, left ICA and CCA thrombus	–	–	Argatroban, fondaparinux	–	Alive
Costentin et al. [20]	17	Headache	Hemiplegia, aphasia, NIHSS 15	MCA	MCA-M1	–	MT	TEE, CM	NA	Hemorrhagic transformation	Alive
Goereci et al. [21]	18	Stroke, headache	Visual impairment, transient hemiparesis and aphasia	MCA	ICA	ICA occlusion at bifurcation	–	CM, TEE, CUS	Argatroban, apixaban	–	Alive
Ceschia et al. [22]	19	Lower limb pain	–	PCA	–	–	–	NA	Fondaparinux, warfarin	Adrenal hematomas, anemia	Alive
Rodriguez-Pardo et al. [23]	20	Stroke	Amaurosis fugax, blurry speech, arm weakness	AChA	–	Carotid bulb floating thrombus	–	CM, TTE, TTE-B, CUS	Fondaparinux, rivaroxaban	–	Alive
Kenda et al. [24]	21	Stroke	Aphasia, hemiplegia, hemianopsia NIHSS20	MCA	MCA-M1	Chronic ICA dissection with 10 mm pseudoaneurysm	MT	TTE, 24 h-Holter	Fondaparinux, aspirin	Hemorrhagic transformation, MCA restenosis	Alive
Su et al. [25]	22	Stroke	Left-side weakness	Bilateral MCA (multiple scattered infarcts), PCA	–	–	–	–	Rivaroxaban (Afib prevention)	Respiratory failure, ascites, upper GI bleeding, cardiac arrest	Death
Charidimou et al. [26]	23	Stroke, headache	Hemiparesis, neglect, gaze deviation NIHSS 15	MCA, ACA	ICA-T (MCA-M1, ACA-A1, A2)	–	MT	TTE	Argatroban	Hemorrhagic transformation, malignant infarct, hemicraniectomy	Alive
Present case	24	Abdominal pain	Hemiparesis, gaze deviation, neglect	MCA	MCA-M1	–	–	TTE, TCD-B, CM	Argatroban, fondaparinux, warfarin	Hemorrhagic transformation, malignant infarct, hemicraniectomy	Alive

ACA, anterior cerebral artery; AChA, anterior choroidal artery; CAS, carotid artery stenting; CCA, common carotid artery; CM, cardiac monitoring; CUS, carotid ultrasound; GCS, Glasgow coma scale; ICA, internal carotid artery; IVT, intravenous thrombolysis; MCA, middle cerebral artery; MT, mechanical thrombectomy; NA, not available; ND, not done; NIHSS, National Institute of Health Stroke Scale; PCA, posterior cerebral artery; PLTs, platelets transfusion; TCD-B, transcranial doppler bubble test; TEE, transesophageal echocardiography; TTE, transthoracic echocardiography; VA, vertebral artery.

Table 3.

General characteristics in patients with ischemic stroke and VITT

Demographics	Entire cohort (N = 24) n/N (%)
Age, median (range)	45.50 (21–73)
Female, sex	19/24 (79.2)
Vascular risk factors	9/24 (37.5)
Hypertension	5/24 (20.8)
Dyslipidemia	5/24 (20.8)
Smoking	2/24 (8.3)
Diabetes	1/24 (4.2)
Atrial fibrillation	1/24 (4.2)
Prothrombotic risk factors	4/24 (16.7)
Oral contraceptive therapy	2/24 (8.3)
Cancer	2/24 (8.3)
Vaccine
ChAdOx1 nCoV-19	20/24 (83.3)
Ad26.COV2.S	3/24 (12.5)
mRNA-1273	1/24 (4.2)
Days from vaccination to admission, median (range)	9.50 (7–21)
Platelet count nadir, ×109/L, median (range)	40 (5–152) [22]
D-dimer level peak, ng/mL FEU, median (range)	32,102 (1,100–106,200) [20]
Fibrinogen level, g/L, median (range)	2.30 (0.63–4.40) [14]
Anti-PF4 positive	22/23 (95.7)
Platelet functional assay positive	13/16 (81.3)
Additional venous thrombosis	13/24 (54.2)
SVT	10/13 (77)
Deep vein thrombosis	8/13 (61.5)
Pulmonary embolism	9/13 (69)
Cerebral venous sinus thrombosis	5/24 (20.8)
Arterial thrombosis (other than stroke)	5/24 (20.8)

[N] available data. Categorical variables are given as n/N (%).

Table 4.

Stroke characteristics in patients with VITT

Baseline characteristics	Entire cohort (N = 24) n/N (%)
Stroke as presenting symptom	18/24 (75)
Neurological deficit
Motor	18/24 (75)
Language	10/24 (41.2)
Visual deficit	4/24 (16.7)
Neglect	3/24 (12.5)
Altered state of mind	3/24 (12.5)
Vascular territory
Anterior circulation	21/24 (87.5)
Posterior circulation	5/24 (20.8)
Multi-territory (bilateral or both anterior and posterior)	4/24 (16.7)
LVO	17/22 (77)
MCA − M1	11/17 (64.7)
ICA	6/17 (35.3)
VA, BA	2/17 (11.8)
Extracranial vessel intraluminal thrombus	10/22 (45.4)
Acute reperfusion therapy	7/22 (31.8)
Thrombolysis	1/22 (4.5)
Mechanical thrombectomy	6/22 (27.3)
Malignant infarct	9/22 (41)
Decompressive hemicraniectomy	7/22 (31.8)
Death	5/24 (20.8)

Categorical variables are given as n/N (%).

Demographics

Among all 24 patients, the median age was 45.5 years (range 21–73) and eighteen patients (75%) were younger than 60 years. Nineteen patients (79.2%) were women. All patients were healthy or clinically stable at the time of the event. Cardiovascular risk factors were present in 37.5% (hypertension and dyslipidemia in 5 patients, smoking in 2 patients, diabetes in 1 patient, atrial fibrillation in 1 patient, obstructive sleep apnea in 1 patient). Two patients suffered from migraine. Most patients (83.3%) had no known prothrombotic factors. Among the remaining cases two young women were known to take the oral contraceptive pill [20, 26], a 69-year-old man suffered from prostate cancer [15], and a 70-year-old man who had been diagnosed a metastatic pancreatic cancer during hospitalization [25]. Almost all patients had received the first dose of an adenoviral vector-based vaccine (ChAdOx1 nCoV-19 in 20 patients, Ad26.COV.2.S in 3 patients), only one patient had received a m-RNA-based vaccine (mRNA-1273). The first dose was administered from 8 to 21 days before medical admission (median 9.5 days). None of the patients have had recent symptoms of SARS-CoV-2 infection. SARS-CoV-2 real-time reverse-transcriptase polymerase-chain-reaction test on nasopharyngeal swab was negative in all individuals tested (13 patients). None had been exposed to heparin recently.

Laboratory Results

The majority of patients (23 out of 24, 95.8%) had platelet count <150 × 10⁹/L (median 40, range 5–152) and 58% had severe thrombocytopenia (<50 × 10⁹/L). D-dimer levels (converted to ng/mL FEU for case comparison) were very high (>4,000 ng/mL FEU; range 1,100–106,200) in 90.9% of available cases (22 patients), in one patient D-dimer was normal. Five patients had low fibrinogen levels (median 2.30 g/L, range 1.5–4.0). Anti-PF4 antibodies (ELISA assay) were present in 22 of the 23 tested patients (92%), in our case negative results were obtained. A platelet functional test was performed in 16 patients; the result was negative in 3 patients (19%), including our case. These patients with negative anti-PF4 or platelet functional assay met all other clinical and laboratory criteria for the diagnosis of VITT [27, 28] and received specific treatment.

Thrombosis

Additional thrombosis other than stroke was found in 13 of 24 patients (54.2%) and 11 out of 13 patients had multiple sites of thrombosis (see Table 3). Most of the cases had venous thrombosis at unusual sites; in particular, 77% (10 patients) had SVT (more often affecting the portal circulation). CVT was present in 5 of 24 patients (20.8%), an isolated thrombosis of the internal jugular vein was detected in another patient. In 2 cases, CVT was complicated by secondary intracranial venous hemorrhage. Deep extremity veins and pulmonary arteries were the second most common site of additional venous thrombosis and were affected in 12 of the 24 patients (50%) (8 cases of DVT and 9 cases of pulmonary embolism; 5 patients had both). In addition to ischemic stroke, an arterial thrombosis was found in 5 cases (2 popliteal artery thrombosis, 1 myocardial infarction, 1 renal infarct, and 1 spleen infarction). SVT was complicated by intra-abdominal hemorrhage in 4 patients (3 adrenal hemorrhages, 2 hemoperitoneum, and 1 renal hemorrhage).

Classification

Patients were defined to have VITT according to five criteria (time to onset from vaccination, presence of thrombosis, thrombocytopenia <150,000 × 10⁹/L, elevated D-dimer >4,000 FEU and positive anti-PF4 antibodies on ELISA) as established by several guidelines and studies [28, 29, 30]. People who met all five criteria were judged to have definite VITT, while those who did not meet one, two, or more criteria were defined to have probable, possible, unlikely VITT [30]. Of the 24 patients included in our review, 18 were classified as having definite VITT, 5 as having probable VITT because one of the criteria was not met. One patient was classified as having possible VITT due to D-dimer levels <4,000 FEU and platelet count >150,000 × 10⁹/L [16].

Stroke Features

Ischemic stroke was the reason of medical presentation in 18 patients (75%). Among the remaining cases (6 patients), 3 patients presented to the emergency department complaining of abdominal pain, 1 patient with lower limb pain due to DVT, 1 patient arrived with left-sided signs and a mild comatose state from intracerebral venous hemorrhage, and 1 patient complaining headache due to CVT. In these cases, ischemic stroke occurred after hospital admission, on the same day (3 patients) or the day after (2 patients). In another patient, a cerebral infarct was an accidental finding after brain MRI. Analyzing stroke characteristics (Table 4), the majority of patients had focal deficits. Motor deficit was the most common symptom (75%), followed by language (aphasia and/or dysarthria), visual deficit, and neglect. In 2 patients with posterior stroke circulation and 1 patient with bilateral carotid infarction, the onset was characterized by confusion and a decreased level of consciousness up to severe coma. In 1 case, neurological examination was irrelevant except for headache. Prodromes (such as flu-like symptoms, fatigue, headache) have been frequently reported a few days after vaccination. Most patients (87.5%) had an anterior circulation stroke, mainly involving MCA. Multi-territorial vascular infarction with bilateral anterior or both anterior and posterior circulation was reported in 4 patients (16.7%). Of all the available data, 77% of patients (17/22) had an intracranial LVO involving the MCA-M1 segment (64.7%, 11/17), intracranial ICA (35.3%, 6/17), and posterior circulation (basilar artery and vertebral artery, respectively) (11.8%, 2/17). In 45.4% of available cases (10 out of 22) examination of the cervical vessels (by ultrasonography, CTA, MR- angiography, and/or invasive angiography) revealed a carotid occlusion at the neck due to large or multiple intraluminal thrombi (7 patients) or the presence of carotid or aortic arch free-floating thrombus (3 patients) (Tables 2, 4). In one patient, CTA revealed extreme tortuosity and chronic dissection of the proximal left ICA with a 10 mm pseudoaneurysm; in this case, the stroke was probably due to distal embolization of a thrombus formed within the pseudoaneurysm [24]. Regarding the diagnostic workup, cardiological data were available in only 10 patients (42%) and tests excluded cardiac embolic sources and/or paradoxical embolization.

Stroke Treatment and Outcome

Acute reperfusion therapy was performed in 7 of the 17 patients with LVO (41%). One patient with a normal platelet count underwent intravenous thrombolysis with alteplase, while 6 patients underwent mechanical thrombectomy. In one patient with tandem occlusion, urgent carotid artery stenting, other than MT, was performed. The remaining cases had no indication for acute revascularization due to severe thrombocytopenia, hemorrhage in other sites, large ischemic changes on CT scan, or absence of treatable penumbra on perfusion CT. Re-occlusion and restenosis of the same vessel after mechanical thrombectomy were reported in two patients. Anticoagulation was administered in 90% of patients with available data (n = 20). Non-heparin anticoagulants including argatroban, fondaparinux, danaparoid, and direct oral anticoagulants were used in most patients. Direct oral anticoagulants, including rivaroxaban and apixaban, were used in 25% (5 patients), mainly in the subacute and chronic phases (4 patients). Heparin was administered in only 1 case at the start of the vaccination campaign. Malignant infarct occurred in 9 patients and decompressive hemicraniectomy was performed in 7 patients (31.8%). At the time of publication, 19 patients were alive, but 1 patient was in critical condition, and 5 patients had died (20.8%). Data about final neurological functional status are lacking.

Isolated Ischemic Stroke versus Ischemic Stroke plus Venous Thrombosis

We compared baseline (demographics, vascular and prothrombotic risk factors, laboratory results) and stroke characteristics between patients with isolated arterial ischemic stroke and patients with additional venous thrombosis other than stroke, to examine if these 2 groups of patients are intrinsically the same (online suppl. Table e1). Patients with stroke plus venous thrombosis were older, more often women, had lower platelet count and higher D-dimer levels. Furthermore, strokes were more frequently due to LVO (92.3% vs. 55.5%), often evolving into a malignant infarct.

Discussion

VITT is an autoimmune disorder, characterized by antibodies that directly activate platelets, triggering thrombosis in the arterial and venous circulation. It is characterized by severe thrombocytopenia, systemic activation of coagulation (suggested by disproportionately elevated D-dimer levels and reduced fibrinogen), extensive venous thrombosis present at multiple and unusual sites such as CVT and SVT (including mesenterial, portal, hepatic, splenic, or adrenal veins), and positive anti-PF4 [6]. This syndrome was initially reported after ChAdOx1 nCov-19 vaccine, several cases have been also described following Ad26.COV2.S vaccine [4, 5] with similar clinical and laboratory features. Unlike the Pfizer-BioNTech and Moderna vaccines, which are mRNA-based, the ChAdOx1 nCov-19 and Ad26.COV2.S vaccines are nonreplicating adenovirus vector-based DNA vaccines, both use an adenoviral vector, human and chimpanzee, respectively. The pathophysiology of VITT is still incompletely understood and a major area of uncertainty is how adenoviral vector vaccines trigger this syndrome. This disorder is strongly similar to autoimmune HIT, but anti-PF4 and thrombotic complications develop in the absence of previous exposure to heparin. Several hypotheses have been proposed and it has been supposed that a component of the vaccine might interact directly with PF4 and could take the place of heparin, causing a novel autoimmune humoral response that induces platelet activation and a hypercoagulative state [6, 31]. Some clinical data seem to suggest a different pathophysiology between these two adenoviral vaccines, regarding the hypothesized HIT-like mechanism. In fact, patients receiving Ad26.COV.2.S tend to present CVT less frequently and later, have lower D-dimer and aPTT levels than ChAdOx1 nCoV-19 and while anti-PF4 antibodies were present in most patients with either vaccine, platelet function tests were rarely positive in the Ad26.COV.2.S group [32]. Although anti-PF4 positivity (ELISA assay) is considered a standard criterion for VITT, negative or equivocal results have been obtained in several cases [3, 33]. These results could be likely related to several biases such as different ELISA assays, lack of standardization in the functional platelet tests, sample timing (platelet transfusion or immunotherapy before the blood draw could give false negatives), or cases that may require testing by more than one ELISA assay before anti-PF4 can be proven. In addition, the potential for cases of VITT unrelated to anti-PF4 mediated pathophysiological mechanism should be considered. Arterial thrombosis appears to be a rare event compared to venous thrombosis. In a recent UK report of 220 cases, only 47 patients (21%) had one or more arterial thrombotic events and only 17 cases had a cerebrovascular accident [30]. In a German report, only 9 cases of primary ischemic stroke were retrospectively identified after COVID-19 vaccination, and 5 patients had a high VITT risk grade [34]. At present, far fewer cases of ischemic stroke have been published following COVID-19 vaccination, compared to cases with CVT. In our systematic review of the literature, we found 23 cases of ischemic stroke and VITT. Other published reports of strokes due to different thrombotic disorders, such as ITP, immune thrombocytopenic purpura (TTP), or catastrophic antiphospholipid syndrome, were identified but excluded from this study. Furthermore, we described a case of ischemic stroke with extensive SVT, severe thrombocytopenia and elevated D-dimer within 13 days of Ad26.COV2.S vaccination. Our patient had negative anti-PF4 ELISA and platelet functional assay, but tests were obtained after several platelet transfusions and were not repeated (if we had suspected VITT earlier, we would have treated the patient differently by not giving platelets). Nevertheless, severely reduced platelet count, high D-dimer levels (>4,000 FEU), and low fibrinogen levels were strongly indicative of systemic activation of coagulation. ITP was excluded and a complete differential diagnosis was made for other coagulation disorders. TTP was ruled out by the absence of schistocytes and normal ADAMTS13, the biochemical panel was not compatible with disseminated intravascular coagulation (unremarkable changes in PT, aPTT, antithrombin), catastrophic anti-phospholipid syndrome was considered, but tests for anti-phospholipids antibodies were all negative. A causal relationship between thrombosis with thrombocytopenia syndrome and vaccine was highly suspected and according to the last guidelines [27, 28, 29] all mandatory criteria (uncommon thrombosis location, thrombocytopenia, and laboratory results) were satisfied to confirm at least a probable VITT diagnosis [29]. Furthermore, although the stroke was found a few hours after surgery, the presence of ischemic changes with midline shift on the first CT scan would allow it to be dated several hours earlier, probably misdiagnosed as the patient was unconscious. In our review, all reports of ischemic stroke followed the first dose of an adenoviral vector vaccine, except for 1 patient who received a m-RNA − based vaccine (Moderna). However this patient, despite meeting all the criteria for definite VITT diagnosis, had several cardiovascular and thrombotic risk factors such as atrial fibrillation and metastatic pancreatic cancer that could have triggered a thrombotic phenomenon. A probable case of VITT was yet reported in the USA after Moderna vaccination [35]. In a study from the USA [36] among over 350 million vaccine recipients, only three cases of VITT after mRNA-1273 (Moderna) were identified, with a reporting rate of 0.00855. These reports may complicate hypotheses that implicate adenoviral vectors as the sole cause of VITT. However considering the very low reporting rate for VITT following mRNA-based COVID-19 vaccines, it is difficult to establish a clear causal correlation and these reports may represent a background rate of cases of spontaneous HIT with a different pathophysiological mechanism. Ischemic stroke during VITT mostly affects healthy young adults, especially women without known prothrombotic risk factors. Stroke, when it occurs, is often the initial presenting symptom, may be isolated or associated with thrombosis in other venous or arterial sites. Among the included cases we found a very high prevalence of LVOs (71%) that was higher (92.3%) in patients with additional venous thrombosis. Moreover, in these patients platelet count was lower compared to those with isolated ischemic stroke, thus confirming a more severe thrombotic disorder in this subgroup. Ischemic stroke with LVO usually represents 10–30% of all AISs [37] with this percentage varying according to which vessels are considered. A higher prevalence of LVOs (up to 47%) was also found in stroke patients with SARS-CoV-2 infection, across all age groups, even in the absence of risk factors or comorbidities. Moreover, in young patients prevalence reached up to 68.8% [38, 39]. Another unusual feature of ischemic stroke in VITT is the high prevalence of both large intraluminal thrombi and free-floating thrombi in extracranial vessels, such as carotid artery, in the absence of underlying atherosclerotic disease. Floating carotid thrombus is an unusual finding in clinical practice, present in only 1.5% of all stroke patients [40]. Cases of asymptomatic carotid thrombi or causing a TIA have also been reported after COVID-19 vaccine [41, 42]. These findings emphasize even more the severe immune-mediated hypercoagulative state induced by the vaccine. In fact, the high prevalence of large intraluminal arterial thrombi could be explained by several hypotheses some of which support that some vaccine proteins could damage endothelial cells, triggering inflammation and platelet activation, causing PF4 release and thrombosis [6]. Furthermore, re-occlusion and restenosis of the same vessel were reported in two patients after mechanical thrombectomy; authors hypothesized it was probably due to the PF4/VWF complexes aggregation recognized by anti-PF4 antibodies on the post-thrombectomy injured arterial endothelium [18, 24]. Anticoagulation is mandatory in VITT and was administered in 90% of patients. However, the initiation of anticoagulation shortly after a major ischemic stroke is burdened with an increased risk of hemorrhagic infarction which occurred in 10 patients, with possible consequences in terms of death, prognosis, and functional neurological outcome. Patients with VITT often have a negative prognosis. In the UK cohort of Pavord et al. [30], 23% of the patients died at the time of reporting and people with CVT and severe thrombocytopenia were at a higher risk. Regarding ischemic stroke our results are similar. Five patients died (21%) and in 80% (4 out of 5 patients) the primary cause of death was neurological, due to malignant cerebral infarct. The increased mortality could be explained by a higher prevalence of LVO, reperfusion therapies that have failed or not performed due to delayed or missed diagnosis, by additional thrombosis and medical complications that may have further worsened the prognosis. However, it should also be considered that these data are likely overestimated due to the tendency to report more serious cases.

Conclusion

In VITT, the etiology of ischemic stroke is likely due to an abnormal immune-mediated intravascular coagulation that induces the formation of large arterial thrombi occluding intracranial and extracranial vessels. Considering the massive brain thrombosis and clinical severity, prompt treatment is essential. While intravenous thrombolysis is usually avoided due to severe thrombocytopenia and high bleeding risk, the endovascular treatment seems feasible and safe, even if restenosis or re-occlusion of the revascularized artery may occur because of the hypercoagulative state. Our study points out that, in addition to venous thrombosis, adenoviral-vectored vaccines also appear to have a cerebral arterial thrombotic risk and clinicians should be aware that ischemic stroke with LVO, although rare, could represent a clinical presentation of VITT. However, caution is needed in questioning the safety of adenoviral vector vaccines, considering the rarity of VITT and the low prevalence of ischemic stroke, and keeping in mind that COVID-19 itself carries a much greater risk of arterial thrombotic complications. In conclusion, our case adds and confirms what has already been reported in literature, but the strength of our study, through the systematic review, is to consolidate the phenotype of ischemic stroke in the context of VITT, of which physicians and in particular vascular neurologists should be aware.

Statement of Ethics

Written informed consent was obtained from our patient for publication of the medical details and any accompanying images. An ethics statement is not applicable for the remaining cases because this study is based on published literature.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

We did not receive funding for the present study.

Author Contributions

Angelo Cascio Rizzo concepted and drafted the manuscript, the tables and figures, and acquired data and worked up the clinical case as physician of care. Giuditta Giussani acquired data and provided revision of the manuscript. Elio Clemente Agostoni concepted, supervised and planed the manuscript, and provided revision of the manuscript.

Data Availability Statement

All data generated and analyzed in this study are included in this article and its online supplementary material. Further inquiries can be directed to the corresponding author.