Different patterns of white matter lesions among patent foramen ovale, atherosclerotic cerebral small vessel disease and cerebral venous thrombosis

None of studies are available on the predictive ability of white matter lesions (WMLs) among patent foramen ovale (PFO), atherosclerotic cerebral small vessel disease (aCSVD) and cerebral venous thrombosis (CVT). Herein, we aimed to uncover the difference of the WML patterns among the three disease entities in a real-world setting to provide clinical references for predicting probable WML etiologies. We retrospectively reviewed data from consecutive patients with imaging-confirmed PFO, aCSVD, or CVT enrolled from 2014 through 2020. WMLs presented on fluid-attenuated inversion recovery (FLAIR) maps were compared among the three groups based on visual evaluation, Fazekas and modified Scheltens scales. Propensity score matching (PSM) was implemented to correct age and hypertension differences among groups. A total of 401 patients were entered into final analysis, including PFO (n = 112, 46.5 ± 12.8 years), aCSVD (n = 177, 61.6 ± 11.8 years) and CVT (n = 112, 37.4 ± 11.4 years) groups. In this study, WMLs occurred in all of the involved patients in the three groups (100%), which were independent to age, symptom onset and disease durations. On visual evaluation, PFO-WMLs were multiple spots distributed asymmetrically around bilateral subcortex and peri-ventricles. aCSVD-WMLs were dots or sheets distributed symmetrically in subcortex and peri-ventricles, and often coexisted with lacunar infarctions. CVT-WMLs were cloud-like around bilateral peri-ventricles, and enabled to attenuate after recanalization. Fazekas and modified Scheltens scores of PFO-WML vs. aCSVD-WML were significantly different even after being matched by 1:2 PSM (all p < 0.05), meaning that the WML burden in aCSVD was considerably heavier than that in PFO. WML patterns induced by PFO, aCSVD and CVT were obviously different, and were therefore of great clinical significance to preliminarily predict and differentiate the three diseases entities.