Ravinder-Jeet Singh, MBBS, DM
Several studies in the past looked at the factors which would predict outcome in patients with ischemic stroke. The prognostic usefulness of some of the clinical variables is now well established; for example, age, stroke severity, admission glucose, cardiac comorbidity. Recent interest is in investigating imaging predictors of functional outcomes, especially infarct volume and intraparenchymal hemorrhage. Some of these clinical and imaging variables are combined to form prognostic scale to predict functional outcome or mortality. While few of the available scales can be applied to the general acute stroke population (for example, ASTRAL, iScore, VISTA), others are more suited for population treated with intravenous thrombolysis (like DRAGON, SEDAN, SPAN-100) or mechanical thrombectomy (POST). In the JURaSSiC study, a validated tool performed better than clinician in predicting the outcomes.1 Therefore, the evidence suggests the usefulness of a validated prognostic tool to predict outcome and also a need to further improve these prediction tools.
In the study by Rudilosso et al.,2 the authors were interested in the question of if a common clinical variable, the stroke severity, could be used to predict outcome when measured longitudinally. They investigated the association between timing of clinical improvement after mechanical thrombectomy (MT) and the long-term outcome (modified Rankin Scale score at 90 day). The cohort included consecutive patients with acute ischemic stroke having a proximal occlusion in the anterior cerebral circulation and treated with MT using stent retriever or distal aspiration devices. The stroke severity was measured on the National Institutes of Health Stroke Scale (NIHSS) before MT, within 30 minutes from the end of MT (d0), at day 1 after MT (d1), and at day 7 or at discharge if it occurred before day 7. Using absolute difference (ΔNIHSS) and percentage changes (%NIHSS) of NIHSS scores between the pretreatment and at different time points after treatment, the authors derived useful cutoffs of ΔNIHSS and %NIHSS for outcome prediction (good outcome defined as modified Rankin Scale score of 0–2 at 90-day).
The most important finding of the study was that the shorter delay to clinical improvement was associated with higher odds of good functional outcome. Clinical improvement within 30 minutes of MT had the strongest association with better outcome than at later time points (OR 47.4 vs 27.7 on day 1 and 12.6 on day 7). An improvement of >30% in National Institutes of Health Stroke Scale score within 30 minutes of end of MT had 25.79 higher odds (95% CI, 12.92–51.47) of good outcome in the multivariable analysis.
Some limitations are inherent to the clinical tool itself. For example, NIHSS would be less reliable as an accurate measure of clinical deficits in patients who have depressed level of consciousness either due to use of sedation or general anesthesia when used during the thrombectomy procedure. Therefore, very early NIHSS will be unreliable as a prediction variable in these patients. Second, the cohort included only anterior circulation stroke, thus, findings cannot be applied in patients with basilar artery occlusion. Finally, the study needs validation in an independent cohort. The study demonstrates usefulness of a readily available clinical variable to predict functional outcome among patient undergoing endovascular thrombectomy. Incorporation of this information in future scores might further improve accuracy of these tools.
References:
- Saposnik G, Cote R, Mamdani M, Raptis S, Thorpe KE, Fang J, et al. JURaSSiC: Accuracy of clinician vs risk score prediction of ischemic stroke outcomes. Neurology. 2013;81:448–455.
- Rudilosso S, Urra X, Amaro S, Llull L, Renú A, Laredo C, et al. Timing and relevance of clinical improvement after mechanical thrombectomy in patients with acute ischemic stroke. Stroke. 2019;50:1467-1472.
