Unique Clinicopathologic Subclassifiers of Cryptogenic Cerebral Emboli.

Introduction: Ipsilateral nonstenotic (<50%) internal carotid artery (ICA) plaque, cardiac atriopathy, and patent foramen ovale (PFO) may account for a substantial proportion of embolic stroke of undetermined source (ESUS).

Methods: Consecutive stroke patients at our center (2019-2021) with unilateral, anterior circulation ESUS were categorized into the following mutually exclusive etiologies: (1) nonstenotic ipsilateral ICA plaque (NSP, ≥3mm in maximal axial diameter), (2) sex-adjusted mod-to-severe left atrial enlargement (LAE), (3) PFO, and (4) "occult ESUS" (patients who failed to meet criteria for these 3 groups). Descriptive statistics and multivariable logistic regression were used to model group characteristics.

Results: Of 132 included patients, the median age was 65 (IQR 56-73), 74 (56%) of whom were White, and 54 (41%) were female. Twenty-one patients (16%) had NSP proximal to the infarct territory, 17 (13%) had LAE, 9 (7%) had a PFO, and 85 (64%) had no other mechanism. Patients with LAE were older (p=0.004), and had more frequent intracranial occlusions of the internal carotid and proximal middle cerebral artery (p=0.048), while tobacco use was most commonly found among patients with NSP (75%) when compared to other ESUS groups (p=0.02). Five of 9 patients with LAE who underwent outpatient telemetry had paroxysmal atrial fibrillation (56%), while zero patients with PFO or NSP had paroxysmal atrial fibrillation (p=0.005). Older age (adjusted OR [aOR] 1.05, 95%CI 1.03-1.07), coronary artery disease (aOR 3.22, 95%CI 1.61-6.44) and hypertension (aOR 2.16, 95%CI 1.14-4.06) were independently associated with LAE, while only tobacco use was associated with NSP when compared to other ESUS subclassifiers (OR 3.18, 95%CI 1.08-0.42). Age and tobacco use were both inversely associated with PFO (aOR 0.93, 95%CI 0.88-0.98, and aOR 0.10, 95%CI 0.02-0.90, respectively).

Conclusions: Certain clinical and radiographic features may be useful in predicting the proximal source of occult cerebral emboli, and can be used for cost-effective outpatient diagnostic testing.