Background: Despite continuous improvements in transcatheter aortic valve implantation (TAVI), periprocedural strokes remain a devastating complication. To optimize cerebral embolic protection (CEP), data on affected brain regions and temporal patterns are of particular interest. We aimed to evaluate the regional distribution of clinically apparent periprocedural strokes and their temporal relationships.

Methods: 3,164 consecutive TAVI procedures performed from 2008 to 2019 at a single institution were retrospectively evaluated. Neurological assessment, evaluation of brain imaging and classification of strokes were performed by an experienced vascular neurologist. Cerebral regions were defined as follows: anterior cerebral artery left/right; middle cerebral artery left/right; posterior cerebral artery left/right; basal ganglia left/right; cerebellum left/right; brainstem; central retinal artery. Clinical outcomes were adjudicated according to the Valve Academic Research Consortium-2 criteria.

Results: The overall incidence of stroke was 3.6% at 30 days (disabling: 2.2%, non-disabling: 1.4%). Symptom onset occurred in 81.9% within 72 hours and in 79.1% of patients within 24 hours after TAVI, respectively. 98.4% of strokes were classified as ischemic. 40.9% of patients were found with one and 35.4% with multiple cerebral lesions. Brain regions affected were as follows (see table1): Anterior cerebral artery 7.1% (left) and 8.0% (right); middle cerebral artery 41.6% (left) and 35.4% (right); posterior cerebral artery 18.5% (left) and 16.8% (right); basal ganglia 15.0% (left) and 16.8% (right); cerebellum 23.0% (left) and 20.4% (right); 8.8% brainstem and 1.8% central retinal artery. Overall, 43.2% of patients were exclusively affected in brain regions protected by the Sentinel^TM CEP, the remaining patients were affected in regions only partially protected. There was no association of comorbidities, procedural aspects or temporal patterns with affected brain regions

Conclusion: Periprocedural strokes occurred early after TAVI, were mostly of ischemic origin and affected multiple brain regions that were only partially protected by current CEP devices. Lesion distribution was not affected by procedural aspects or comorbidities. These data may help to improve prevention of cerebrovascular events during TAVI.

Table 1: