Transcranial doppler detected right-to-left shunt is common but not associated with MRI white matter hyperintensity burden: a cross-sectional study

Introduction Right-to-left shunt (RLS) associated with a patent foramen ovale has been related with ischemic stroke. However, its relationship with MRI white matter hyperintensities (WMHs) remains debated. This cross-sectional, single-centre study investigated the prevalence of RLS detected by transcranial Doppler sonography (TCD) and its association with vascular lesions on MRI. Patients and methods 502 outpatients (mean age 47.8 +/- 13 years; 45% male) with non-specific neurological symptoms underwent brain MRI and TCD with contrast saline. WMH severity was visually graded using the Fazekas scale. Results RLS was detected in 39% of the sample. No difference was found in demographics and clinical variables between those with and without RLS. No association was also found between RLS and MRI lesion load. As expected, a significant (P < .001) positive correlation was identified between age and Fazekas scores (ie, higher scores with increasing age). No effect on lesion load was found for sex, hypercholesterolemia, diabetes, obesity and smoking, while a statistically significant association (P = .016) was present for arterial hypertension (odds ratio 1.68, 95% CI, 1.10-2.56; among those with higher Fazekas scores). Finally, no significant association was found between RLS magnitude, both at rest and during the Valsalva manoeuver and the Fazekas scores. Discussion Although RLS was frequently detected in this cohort, it was not associated with the presence or severity of WMHs, which were instead driven by age and arterial hypertension. These findings support WMHs as MRI marker of cerebral small vessel disease rather than subclinical paradoxical embolism. This also suggests limited utility of routine TCD screening for RLS in patients with incidental WMHs and no history or sign of embolic features. Conclusions In patients with non-specific neurological symptoms, we detected a high occurrence of RLS, although this was not associated with an increased risk or severity of WMHs. As such, paradoxical embolism may not be a major determinant of subclinical WMHs in this population.