Detection of Right-to-Left Cardiac Shunt in the Absence of Transcranial Acoustic Bone

PURPOSE: Paradoxical thrombotic embolism via right-to-left cardiac shunt (RLS) is a risk factor of cryptogenic ischemic stroke. Transtemporal Doppler (TTD) is a valid method used in the detection of patent foramen ovale (PFO). Temporal acoustic bone windows are missing with increasing age and in some younger subjects. We studied prospectively whether Doppler ultrasound of the cervical arteries (submandibular internal carotid artery [ICA] and vertebral artery [VA]) is an alternative, when compared to TTD, in the detection and quantification of PFO. MATERIAL AND METHODS: A total of 94 patients with sufficient temporal bone windows suffering from recent ischemic stroke underwent TTD and ICA (n = 51) or TTD and VA (n = 43). After injection of microbubbles, the numbers of artificial high-intensity signals (HITS) were recorded at rest and after Valsalva maneuver. RESULTS: For 47 patients in the ICA group, an RLS was found at rest in 23 patients and after Valsalva in 28 patients. At rest, sensitivity was 100%, specificity 96%, positive predictive value (ppv) 95.6%, and negative predictive value (npv) 100%. After Valsalva, sensitivity was 100%, specificity 95%, ppv 96.4%, npv 100%. For 43 patients in the VA group an RLS was found at rest in 14 patients and after Valsalva in 19 patients. At rest, sensitivity was 71.4%, specificity 100%, ppv 100%, and npv 87.8%. After Valsalva, 94.4%, 96%, 94.4%, and 96%, respectively. Pearson's correlations of the number of HITS between TTD and ICA and between TTD and VA were highly significant. CONCLUSIONS: When transcranial acoustic bone windows are missing, Doppler ultrasound of the cervical submandibular ICA and VAs are valid screening methods to detect RLS due to a PFO.