Minor stroke risk varies beyond NIHSS score, review finds

A low score on the National Institutes of Health Stroke Scale (NIHSS) does not make minor ischemic stroke a uniformly low-risk condition, according to a 2026 narrative review. Patients with similar initial impairment scores can have different arterial lesions, amounts of threatened brain tissue, disabling deficits and causes of stroke, producing distinct risks over different time scales. The review proposes assessing four domains: the functional relevance of symptoms, vascular and hemodynamic status, tissue status and etiologic subtype. Its central proposal is an outcome-first approach that distinguishes early neurological deterioration or infarct growth, recurrent ischemia and incomplete recovery. The framework is intended to organize clinical assessment, not serve as a validated prediction score or treatment algorithm.

Minor stroke is often defined through the NIHSS, which measures initial neurological impairment, but thresholds used in trials and registries do not identify a uniform biological subgroup. “Mild” and “nondisabling” are related but not interchangeable concepts: mild describes measured impairment, whereas nondisabling describes whether a persisting deficit would interfere with basic daily activities or, when relevant, return to work. A patient may score only a few points yet have complete hemianopia, severe aphasia, marked neglect or weakness that prevents sustained antigravity effort. Posterior-circulation strokes are particularly prone to underestimation because the standard scale may give relatively little weight to gait or truncal ataxia, dysphagia, oculomotor abnormalities and visual-field loss. Baseline independence, occupational and caregiving demands, and the patient’s own priorities provide necessary context for judging whether a persisting deficit is disabling.

Imaging can confirm injury but cannot, by itself, declare a minor stroke safe. Diffusion-weighted magnetic resonance imaging frequently identifies acute infarction after transient or minor symptoms, showing that the event was ischemic rather than a transient ischemic attack. Such a lesion confirms tissue injury, although it does not establish whether the deficit is functionally minor, identify its mechanism or show that additional tissue remains vulnerable. Noncontrast computed tomography can exclude hemorrhage and reveal some established infarcts, but it is less sensitive for small, early and posterior-fossa lesions. CT or MR angiography defines arterial abnormalities, while diffusion imaging demonstrates acute cytotoxic injury and may reveal multiplicity, strategic location or patterns that help distinguish embolic and atheromatous mechanisms. Lesion multiplicity carries additional prognostic information, particularly when intracranial atherosclerotic disease is present. None of these measurements alone supplies a reliable prognosis.

The timing of adverse outcomes further complicates assessment. Early neurological deterioration and infarct growth generally emerge within hours or days, recurrent ischemia clusters over subsequent days and weeks, and incomplete recovery may persist for months. An unchanged neurological examination offers only provisional reassurance. Deterioration attributable to the index stroke can reflect thrombus propagation, infarct extension, hemodynamic compromise, edema or perforator-territory involvement, whereas recurrent ischemia is a new event arising from an active cause such as an embolic source or unstable atherosclerotic lesion. Reporting the numerical definition and observation window for deterioration is essential because studies variously define it as an increase of at least two NIHSS points and follow patients for periods ranging from 24 hours to several days or longer. The review therefore separates mechanisms that may drive these outcomes from the outcomes themselves.

Assessment of vascular and hemodynamic status is particularly important when symptoms appear unexpectedly mild for the arterial lesion or perfusion abnormality. Relevant findings include proximal or distal occlusion, stenosis, tandem disease, thrombus length, parent-artery plaque, dissection and collateral circulation. In minor large-vessel occlusion, the occlusion site has been associated with early neurological deterioration, while longer thrombi have been associated with failure of early recanalization. Delayed or incomplete collateral filling can indicate tissue vulnerable to inadequate blood flow, and symptoms that change with posture or blood pressure can provide further clues. A perfusion deficit larger than the established infarct may indicate potentially salvageable tissue, but it is a proxy rather than proof of tissue viability. Multiphase or dynamic CT angiography can reveal collateral delay missed by a single arterial phase. These observations support repeated clinical assessment and, when clinically indicated, repeat evaluation of vessel patency, collateral flow, perfusion or lesion size.

High-risk minor stroke is not limited to large-vessel occlusion. Small subcortical infarcts may reflect intrinsic small-vessel disease, but they can also arise from plaque at a perforator origin or branch atheromatous disease. A proximal subcortical lesion reaching the parent artery favors branch atheromatous disease, whereas a small, distal, rounded lesion is more consistent with intrinsic small-vessel disease; either pattern remains a mechanistic inference to be reconciled with arterial imaging and clinical course. Capsular and pontine warning syndromes consist of recurrent, stereotyped lacunar episodes and can precede completed infarction. In 2025, a systematic review of 67 reports and 13,407 participants found early neurological deterioration in more than one fifth of patients with acute lacunar ischemic stroke, although definitions and observation windows varied. A mild presentation without proximal occlusion therefore cannot be classified as low risk solely because CT or MR angiography shows no large-vessel blockage.

Randomized treatment evidence must also be interpreted against the population and outcome studied. In PRISMS, alteplase did not improve 90-day functional outcome compared with aspirin in selected patients with nondisabling minor stroke, while symptomatic intracranial hemorrhage occurred in five alteplase-treated patients and none receiving aspirin. The trial stopped early, excluded clearly disabling deficits and did not require vascular imaging. ARAMIS found dual antiplatelet therapy noninferior to alteplase for an excellent 90-day outcome in selected nondisabling stroke. TEMPO-2 enrolled patients with NIHSS scores from 0 to 5 and either an intracranial occlusion or focal perfusion abnormality; tenecteplase did not improve return to baseline function, 90-day mortality was 5% versus 1%, and symptomatic intracranial hemorrhage was numerically more frequent at 2% versus less than 1% (risk ratio 4.2, 95% confidence interval 0.9–19.7; p=0.059). These trials do not support routine thrombolysis in the populations studied, but their selection criteria also mean that the results cannot be applied indiscriminately to every low-score presentation.

Endovascular therapy is developing separate randomized evidence for carefully selected patients with mild-deficit large-vessel occlusion. MILD-MT was designed as a 300-patient randomized trial comparing endovascular therapy plus best medical management with best medical management alone in adults with anterior-circulation occlusion, NIHSS below 6, ASPECTS of at least 6, infarct core no greater than 50 milliliters, perfusion mismatch of at least 50 milliliters and presentation within 24 hours. Investigators reported at ESOC 2026 that the endovascular group had a higher proportion with 90-day modified Rankin Scale scores of 0–1 and less early neurological deterioration. The efficacy findings remain conference-reported, with a full peer-reviewed results article pending. They also cannot be directly compared with TEMPO-2 because the intervention, imaging selection, reperfusion mechanism, time window and outcome context differed. The emerging signal therefore applies to a narrow, perfusion-enriched subgroup rather than all patients with low NIHSS scores and large-vessel occlusion.

Prevention must follow the mechanism producing the event rather than the initial severity score. CHANCE and POINT showed fewer early ischemic events with clopidogrel plus aspirin than with aspirin alone in selected patients with minor non-cardioembolic stroke or high-risk transient ischemic attack. THALES found fewer strokes or deaths with ticagrelor plus aspirin, but severe bleeding was more frequent. A systematic review indicates that most benefit from clopidogrel-based dual antiplatelet therapy accrues early, whereas bleeding risk continues with longer exposure. For branch atheromatous disease, STRATEGY enrolled 970 patients in China and found that tirofiban plus aspirin did not significantly reduce its primary composite of early neurological deterioration within seven days or new stroke within 90 days: 17.1% versus 19.6%, with a hazard ratio of 0.88 and 95% confidence interval of 0.65–1.19. A lower 90-day disability outcome favored tirofiban, but current randomized evidence does not support routine use for unselected branch atheromatous disease.

Outcome measurement also matters because conventional disability categories can conceal differences that matter to patients. The modified Rankin Scale remains important, but patients may remain ambulatory and independent while failing to resume premorbid activities or work. Fatigue, cognitive and physical limitations, and depression can affect recovery and quality of life. The Canadian Outcome Scale for Minor Stroke is a seven-point measure that includes limitations affecting hobbies, employment, education, service and caregiving, and may complement the modified Rankin Scale. Its initial reliability study found substantial interrater and almost-perfect intrarater agreement, but prospective validation remains necessary, and it has not replaced established trial endpoints. Future risk models and trials should therefore estimate early deterioration, recurrent ischemia, treatment-related hemorrhage and incomplete recovery separately rather than treating “poor outcome” as a single construct. The 2026 review itself was narrative, did not include meta-analysis or formal risk-of-bias assessment, and its four-domain framework has not been prospectively validated or assigned predictive weights.

Subject of Research: Narrative review of risk assessment and management in minor ischemic stroke

Article Title: Heterogeneity and risk assessment in minor ischemic stroke

Article References: Xing, B., Zhang, L., & Tu, W.-J. (2026). Heterogeneity and risk assessment in minor ischemic stroke. Frontiers in Neurology, 17, 1993509. https://doi.org/10.3389/fneur.2026.1993509

Image Credits: AI Generated

DOI: 10.3389/fneur.2026.1993509

Keywords: minor ischemic stroke, NIHSS, risk stratification, large-vessel occlusion, early neurological deterioration, branch atheromatous disease, perfusion imaging, dual antiplatelet therapy, endovascular therapy, stroke prevention, Heterogeneity, risk