Doppler flow morphology characteristics of epiaortic arteries in aortic valve pathologies: a retrospective study on a cohort of patients with ischemic stroke.

Background and aims: Neurovascular ultrasound (nvUS) of the epiaortic arteries is an integral part of the etiologic workup in patients with ischemic stroke. Aortic valve disease shares similar vascular risk profiles and therefore not only presents a common comorbidity, but also an etiologic entity. The aim of this study is to investigate the predictive value of specific Doppler curve flow characteristics in epiaortic arteries and the presence of aortic valve disease.

Methods: Retrospective, single-center analysis of ischemic stroke patients, both receiving full nvUS of the extracranial common- (CCA), internal- (ICA) and external carotid artery (ECA) and echocardiography (TTE/TEE) during their inpatient stay. A rater blinded for the TTE/TEE results investigated Doppler flow curves for the following characteristics: 'pulsus tardus et parvus' for aortic valve stenosis (AS) and 'bisferious pulse', 'diastolic reversal', 'zero diastole' and 'no dicrotic notch' for aortic valve regurgitation (AR). Predictive value of these Doppler flow characteristics was investigated using multivariate logistic regression models.

Results: Of 1320 patients with complete examination of Doppler flow curves and TTE/TEE, 75 (5.7%) showed an AS and 482 (36.5%) showed an AR. Sixty-one (4.6%) patients at least showed a moderate-to-severe AS and 100 (7.6%) at least showed a moderate-to-severe AR. After adjustment for age, coronary artery disease, arterial hypertension, diabetes mellitus, smoking, peripheral arterial disease, renal failure and atrial fibrillation, the following flow pattern predicted aortic valve disease: 'pulsus tardus et parvus' in the CCA and ICA was highly predictive for a moderate-to-severe AS (OR 1158.5, 95% CI 364.2-3684.8, p < 0.001). 'No dicrotic notch' (OR 102.1, 95% CI 12.4-839.4, p < 0.001), a 'bisferious pulse' (OR 10.8, 95% CI 3.2-33.9, p < 0.001) and a 'diastolic reversal' (OR 15.4, 95% CI 3.2-74.6, p < 0.001) in the CCA and ICA predicted a moderate-to-severe AR. The inclusion of Doppler flow characteristics of the ECA did not increase predictive value.

Conclusions: Well defined, qualitative Doppler flow characteristics detectable in the CCA and ICA are highly predictive for aortic valve disease. The consideration of these flow characteristics can be useful to streamline diagnostic and therapeutic measures, especially in the outpatient setting.