Clinically low-risk features do not exclude diffusion-weighted imaging-positive lesions in suspected transient ischemic attack

Diffusion-weighted imaging (DWI) is essential for distinguishing tissue-negative transient ischemic attack (TIA) from MRI-defined minor stroke. When MRI capacity is limited, clinical features could help estimate DWI lesion yield. While features associated with DWI positivity are well established, it remains uncertain whether clinically low-risk presentations identify patients with a sufficiently low probability of DWI-positive lesions. We investigated clinical features associated with DWI-positive lesions in patients presenting with suspected TIA and assessed whether clinically low-risk constellations reliably excluded DWI-positive lesions.

We retrospectively analyzed 418 patients presenting with suspected TIA who underwent a standardized MRI protocol within 6 h of admission. The study endpoint was an acute DWI-positive lesion, defined by DWI hyperintensity with corresponding apparent diffusion coefficient reduction. Univariate analyses and exploratory multivariable logistic regression analyses were used to examine clinical, neurological, and cardiovascular features associated with DWI positivity, with particular emphasis on clinically low-risk presentations.

DWI-positive lesions were detected in 118/418 patients (28.2%). Motor symptoms (adjusted odds ratio [aOR] 2.683, p < 0.001), NIHSS 1-5 (aOR 1.940, p = 0.007), and acute symptom onset (aOR 1.999, p = 0.025) were associated with DWI positivity in the clinical-neurological analysis, while cardiovascular risk factors showed the strongest association in the extended analysis (aOR 3.959, p < 0.001). Conversely, subacute onset, non-specific accompanying symptoms, NIHSS 0, and dizziness were associated with lower odds of DWI positivity. Importantly, DWI-positive lesions remained present in 13/133 patients (9.8%) with ≥2 and 2/48 (4.2%) with ≥3 low-risk features.

Clinically low-risk presentations did not reliably exclude DWI-positive lesions, supporting MRI-based evaluation. Clinical and cardiovascular features may help estimate the likelihood of DWI positivity; whether they can inform MRI prioritization requires prospective validation.