Undetermined stroke with an embolic pattern—a common phenotype with high early recurrence risk

Figures & data

Figure 1. Flow chart of the study population from admission to phenotype with frequency of atrial fibrillation (AF) among the subgroups. * The dominant etiology was intracranial atherosclerosis in a relevant artery in the 2 patients with AF.

Figure 2. Change of the patient phenotype with age (chi-square P < 0.001). ESUS = embolic stroke of undetermined source.

Table I. Baseline data of the study population stratified by patient phenotype.

	Non-cardioembolic stroke etiology (n = 229)	Cardioembolic stroke with a known source (n = 184)	Undetermined stroke with an embolic pattern (n = 127)
Demographics
Age, years	67 (59–76)	72 (64–80)	66 (54–75)^b
Male gender	145 (63.3)	105 (57.1)	70 (55.1)
Pre-existing comorbidities
Hypertension	149 (65.1)	126 (68.5)	78 (61.4)
Diabetes	50 (21.8)	37 (20.1)	16 (12.6)^a
Dyslipidemia	142 (62.0)	108 (58.7)	76 (59.8)
Current smoking (n = 470)	61 (28.9)	32 (21.2)	28 (24.8)
Coronary heart disease	30 (13.1)	37 (20.1)	22 (17.3)
Heart failure	11 (4.8)	31 (16.8)	6 (4.7)^b
History of MI	16 (7.0)	25 (13.6)	9 (7.1)
Peripheral arterial disease	13 (5.7)	6 (3.3)	4 (3.1)
Atrial fibrillation	9 (3.9)	134 (72.8)	0^a,b
Persistent/permanent	3 (1.3)	61 (33.2)	0
Paroxysmal	6 (2.6)	73 (39.7)	0
Cancer	19 (8.3)	23 (12.5)	16 (12.6)
History of TIA	18 (7.9)	7 (3.8)	2 (1.6)^a
History of stroke	30 (13.1)	32 (17.4)	15 (11.8)
CHA2DS2-VASc score	2 (1–4)	3 (2–4)	2 (1–4)^b
CHA2DS2-VASc score 0–1	67 (29.3)	31 (16.8)	40 (31.5)
CHA2DS2-VASc score 2–3	84 (36.7)	71 (38.6)	49 (38.6)
CHA2DS2-VASc score ≥ 4	78 (34.1)	82 (44.6)	38 (29.9)
NIHSS score	2 (1–5)	5.5 (2–12)	4 (1–8)^a,b
NIHSS score 0–6 (mild)	194 (84.7)	105 (57.1)	90 (70.9)
NIHSS score 7–14 (moderate)	18 (7.9)	46 (25.0)	24 (18.9)
NIHSS score ≥ 15 (severe)	17 (7.4)	33 (17.9)	13 (10.2)
Acute treatment
Intravenous thrombolysis	38 (16.6)	55 (29.9)	30 (23.6)
Endovascular thrombectomy	2 (0.9)	11 (6.0)	5 (3.9)
Etiologic work-up
Brain CT at least once	221 (96.5)	183 (99.5)	111 (87.4)^a,b
Brain MRI at least once	65 (28.4)	16 (8.7)	52 (40.9)^a,b
Extracranial artery imaging	223 (97.4)	143 (77.7)	120 (94.5)^b
Doppler ultrasound	128 (55.9)	92 (50.0)	70 (55.1)
CT angiography	116 (50.7)	60 (32.6)	38 (29.9)^a,b
MR angiography	42 (18.3)	13 (7.1)	35 (27.6)^a
Intracranial artery imaging	139 (60.7)	67 (36.4)	67 (52.8)^b
CT angiography	115 (50.2)	60 (32.6)	38 (29.9)^a
MR angiography	41 (17.9)	13 (7.1)	34 (26.8)^a,b
Echocardiography ^c	50 (21.8)	37 (20.1)	69 (54.3)^a,b
Left atrium diameter (mm)	36 (31–40)	39 (32–42)	35 (32–40)
Left ventricle end-diastolic diameter (mm)	48 (43–54)	50 (44–55)	49 (47–53)
Left ventricle ejection fraction (%)	60 (51–69)	55 (40–62)	60 (58–66)^d
Repeated ECGs	218 (95.2)	176 (95.7)	124 (97.6)
Any continuous ECG monitoring	77 (33.6)	95 (51.6)	65 (51.2)^a
Continuous ECG monitoring in ER, stroke unit, or emergency care unit	65 (28.4)	86 (46.7)	45 (35.4)^b
48-hour Holter monitoring	22 (9.6)	16 (8.7)	33 (26.0)^a,b
Testing for thrombophilia	19 (8.3)	7 (3.8)	32 (25.2)^a,b
Antithrombotic medication
Antiplatelet	193 (84.6)	39 (21.2)	108 (85.0)^b
Oral anticoagulation	35 (15.5)	133 (72.3)	17 (13.4)^b

Data are median (interquartile range), or n (%).

CT = computed tomography; ER = emergency room; MI = myocardial infarction; MRI = magnetic resonance imaging; NIHSS = National Institutes of Health Stroke Scale.

^aP < 0.05 for undetermined stroke with an embolic pattern versus non-cardioembolic stroke etiology.

^bP < 0.05 for undetermined stroke with an embolic pattern versus cardioembolic stroke with a known source.

^cTransesophageal echocardiography in 2, 1, and 5 patients, respectively.

^dP < 0.01 for cardioembolic stroke with a known source versus each of the other two groups.

Table II. Comparison of patients with undetermined stroke with an embolic pattern meeting the diagnostic criteria for embolic stroke of undetermined source (ESUS) (+), to those not meeting the criteria, ESUS (–).

	ESUS (–) n = 81	ESUS (+) n = 46	P
Demographics
Age	70 (59–76)	62 (47–71)	0.011
Male gender	44 (54.3)	26 (56.5)	0.811
Pre-existing comorbidities
Hypertension	58 (71.6)	20 (43.5)	0.002
Diabetes	13 (16.0)	3 (6.5)	0.120
Dyslipidemia	51 (63.0)	25 (54.3)	0.341
Current smoker (n = 470)	15 (22.1)	13 (28.9)	0.410
Coronary heart disease	16 (19.8)	6 (13.0)	0.337
Heart failure	6 (7.4)	0	0.086
History of MI	3 (3.7)	6 (13.0)	0.071
Peripheral arterial disease	3 (3.7)	1 (2.2)	1.000
Atrial fibrillation	0	0	N.A.
Cancer	14 (17.3)	2 (4.3)	0.035
History of TIA	2 (2.5)	0	0.534
History of stroke	11 (13.6)	4 (8.7)	0.412
CHA2DS2-VASc score	3 (2–4)	2 (2–4)	0.002
CHA2DS2-VASc score 0–1	19 (23.5)	21 (45.7)	N.A.
CHA2DS2-VASc score 2–3	32 (39.5)	17 (37.0)	N.A.
CHA2DS2-VASc score ≥ 4	30 (37.0)	8 (17.4)	N.A.
NIHSS score	3 (2–8)	4 (1–8)	0.868
NIHSS score 0–6 (mild)	59 (72.8)	31 (67.4)	N.A.
NIHSS score 7–14 (moderate)	14 (17.3)	10 (21.7)	N.A.
NIHSS score ≥ 15 (severe)	8 (9.9)	5 (10.9)	N.A.
Acute treatment
Intravenous thrombolysis	10 (12.2)	20 (44.4)	< 0.001
Endovascular thrombectomy	2 (2.4)	3 (6.7)	0.346
Etiologic work-up
Brain CT at least once	69 (85.2)	42 (91.3)	0.318
Brain MRI at least once	32 (39.5)	20 (43.5)	0.662
Extracranial artery imaging	74 (91.4)	46 (100)	0.040
Doppler ultrasound	48 (59.3)	22 (47.8)	0.104
CT angiography	19 (23.5)	19 (41.3)	0.054
MR angiography	24 (29.6)	11 (23.9)	0.378
Intracranial artery imaging	38 (46.9)	29 (63.0)	0.080
CT angiography	19 (23.5)	19 (41.3)	0.119
MR angiography	24 (29.6)	10 (21.7)	0.037
Echocardiography ^a	23 (28.4)	46 (100)	< 0.001
Left atrium diameter (mm)	37 (32–41)	36 (31–40)	0.752
Left ventricle end-diastolic diameter (mm)	50 (44–55)	49 (45–53)	0.947
Left ventricle ejection fraction (%)	60 (50–66)	60 (55–66)	0.733
Repeated ECGs	79 (97.5)	45 (97.8)	0.916
Any continuous ECG monitoring	19 (23.5)	46 (100)	< 0.001
Continuous ECG monitoring in ER, stroke unit, or emergency care unit	17 (21.0)	28 (60.9)	< 0.001
48-hour Holter monitoring	5 (6.2)	28 (60.9)	< 0.001
Testing for thrombophilia	16 (19.8)	16 (34.8)	0.061
Antithrombotic medication
Antiplatelet	70 (85.4)	38 (84.4)	0.889
Oral anticoagulation	9 (11.0)	8 (17.8)	0.282

Data are median (interquartile range), or n (%).

CT = computed tomography; ER = emergency room; MI = myocardial infarction; MRI magnetic resonance imaging; NIHSS = National Institutes of Health Stroke Scale.

^aTransesophageal echocardiography in 3 and 2 patients, respectively.

Figure 3. Kaplan–Meier estimates of the risk of: A: recurrent stroke (log rank P = 0.089), B: death (P = 0.004), and C: composite of recurrent stroke or death (P = 0.008) in the 540 patients stratified by phenotype.