Impact of presenting stroke severity and thrombolysis on outcomes following urgent carotid interventions
Aaron Hayson1, Jeff Burton2, Joseph Allen1, W.C. Sternbergh, III1, Daniel Fort2, Hernan Bazan1
1Ochsner Medical Center, New Orleans, LA;2Ochsner Center for Outcomes Research, New Orleans, LA
Background: Carotid interventions are increasingly performed in select patients following acute stroke. We aimed to determine the effects of presenting stroke severity (National Institutes of Health Stroke Scale, NIHSS) and use of systemic thrombolysis (tPA) on clinical and discharge neurologic outcomes (modified Rankin scale score, mRS) following urgent carotid endarterectomy (uCEA) and urgent carotid artery stenting (uCAS).
Methods:The presenting stroke severity and clinical and neurologic outcomes in a tertiary Comprehensive Stroke Center were determined using records from 9,545 patients in > 50 referring sites from January 1, 2015, through May 1, 2022. Urgent carotid interventions were defined as patients undergoing a CEA or CAS during the index hospitalization after presenting with a transient ischemic attack (TIA) or an acute stroke with > 50% common or internal carotid stenosis based on imaging. Patients were then divided into two cohorts: 1) No thrombolysis (‘uCEA/uCAS only’) and 2) use of thrombolysis prior to intervention (‘tPA + uCEA/uCAS’). The presenting stroke severity was determined by vascular/stroke neurologists using the NIHSS. An NIHSS score ≤4 represents a ‘minor' stroke, 5 - 15 a ‘moderate' stroke, 16 - 20 a ‘moderate-severe' stroke, and 21 - 42 a ‘severe' stroke. Functional neurological outcomes were ascertained using the discharge modified Rankin scale (mRS) score. These scores are used to represent non-disabled neurologic status (mRS ≤ 1), neurologic functional independence (mRS ≤ 2), severe neurologic disability with complete functional dependence (mRS 3 - 5), and death (mRS = 6). An mRS ≤ 2 is regarded as a good outcome, whereas an mRS ≥ 3 will require increasing levels of assistance with activities of daily living. Other clinical outcomes ascertained were 30-day postoperative stroke, death, and myocardial infarction (MI). Patient characteristics, carotid
interventions, presenting stroke severity, and clinical outcomes were compared between groups. Differences in hospital length of stay were assessed via lognormal regression while days from hospital admission to intervention and NIHSS score were assessed via negative binomial regression. All other continuous measures were evaluated using two-sample t-tests and categorical measures using chi-squared tests or Fisher’s exact tests. Next, the ability to predict neurologic outcomes at discharge using the presenting stroke severity was assessed using receiver operating characteristic (ROC) curve. Associations between thrombolysis (tPA) and discharge mRS were assessed by negative binomial regression models and stratified by stroke severity (NIHSS). Finally, associations between stroke severity and clinical outcomes defined by discharge mRS were evaluated by odds ratios (OR) from ordinal logistic regression models and relative risks (RR) from log-binomial regression models stratified by thrombolysis group. Differing levels of stroke severity were chosen a priori (NIHSS 0-4 vs. >4, NIHSS 0-10 vs. >10). The resulting within-group ORs represent odds of a lower discharge mRS score and RRs represent the probabilities of (1) neurologic functional independence (mRS ≤ 2) and (2) neurologic non-disabled status (mRS ≤ 1). Where appropriate, RR was chosen over odds ratio (OR) due to OR tending to inflate estimates of association when the outcome is common.
Results: Over a seven-year period, 238 patients underwent uCEA/uCAS (uCEA/uCAS only, n=185; tPA + uCEA/uCAS, n=52). Patient demographics, presenting stroke severity (NIHSS), and discharge functional neurologic outcomes (mRS) are shown in the Table. There is a higher prevalence of a peripheral artery disease (PAD) diagnosis and prior stroke history in the uCEA/uCAS cohort.
Interventions: 181 uCEAs were performed and 57 uCAS interventions were performed, 2 of which were TCAR. 20 cervical carotid interventions were performed at the time of intra-cranial mechanical endovascular revascularization, MER. There were 10 cases of uCAS where no embolic protection device was used, most often due to concern of tandem distal occlusion or physician deemed too hazardous. Mean time to intervention following stroke was 3.2d, with no significant difference between groups, P = 0.804.
Stroke severity: Mean presenting stroke severity was significantly higher in the tPA + uCEA/uCAS compared to the uCEA/uCAS only cohort (NIHSS = 7.6 vs. 3.8, respectively, P = 0.001) (Figure, Panel A), and more patients in the tPA + uCEA/uCAS cohort presented with moderate to severe strokes (57.7% vs. 30.2% with NIHSS > 4).Clinical outcomes: There were nine strokes (4.8%) and six hemorrhagic conversions (3.2%) in the uCEA/uCAS only cohort. The uCEA/uCAS overall rates of stroke, death, and MI within 30 days of intervention were 8.8%, 0% and 0.5% respectively. For the tPA + uCEA/uCAS cohort there were two strokes (3.8%) and four hemorrhagic conversions (7.7%). The tPA + uCEA/uCAS overall rates of stroke, death, and MI within 30 days of intervention were 11.5%, 9.6% and 1.9% respectively. The 30-day stroke/hemorrhagic conversion rate did not differ with tPA use pre-operatively: uCEA/uCAS (8.1% [15/185]) and tPA + uCEA/uCAS (11.5% [6/52]); P = 0.416; nor did the myocardial infarction rate (MI); P = 0.390. However, in the tPA cohort, five deaths occurred (9.6% [5/52]), and none in the uCEA/uCAS only cohort; P < 0.001. Two deaths were following hemorrhagic conversion, the remaining occurred due to cardiac arrest, sepsis and multisystem organ failure, and pulmonary embolism. The relative risks of these clinical outcomes are seen in Figure, panel B.
Outcomes based on mRS: Importantly, though the presenting stroke severity was higher in the tPA + uCEA/uCAS cohort, there was no difference in neurologic functional outcome between the two cohorts (mean mRS = 2.1 vs. 1.7, respectively P = 0.061). The area under the ROC curve, evaluating NIHSS prediction of functional independence (mRS ≤ 2) for all patients, uCEA/uCAS, and tPA + uCEA/uCAS , was 0.78, 0.76, and 0.79 respectively, (Figure, Panel C). ROC curves were also constructed for NIHSS prediction of non-disabled status (mRS ≤ 1), which showed an AUC of 0.76, 0.77, and 0.78, respectively. We ascertained relative risk (RR) and 95 % confidence intervals (CI) for functional independence (mRS ≤ 2) and non-disabled status (mRS ≤ 1) at different presenting stroke severities. In the entire cohort, the likelihood of discharge functional independence (mRS ≤ 2) was not influenced by thrombolysis. This was noted for both minor strokes (NIHSS ≤ 4 vs. > 4: RR = 1.58 [CI 1.21-2.06] vs. RR = 1.58 [CI 1.01-2.47], no tPA and tPA, respectively P = 0.997) and moderate strokes, (NIHSS ≤ 10 vs. > 10: RR = 2.08 [CI 1.10-3.94] vs. RR = 1.94 [CI 0.93-4.06], P = 0.891). The likelihood for being non-disabled (mRS ≤ 1) was also not influenced by thrombolysis. This was true for minor strokes (NIHSS ≤ 4 vs. > 4: RR = 3.08 [CI 1.84-5.17] vs. RR = 1.70 [CI 1.01-2.88], no tPA and tPA, respectively P = 0.113) and moderate strokes, (NIHSS ≤ 10 vs. > 10: RR = 3.39 [CI 1.20-9.61] vs. RR = 3.24 [CI 1.15-9.00], P = 0.952). For both cohorts (n=238), NIHSS ≤ 10 was associated with an over twice as likely chance of neurologic functional independence (mRS ≤ 2; RR = 2.09 [CI 1.30-3.36]) compared to an NIHSS > 10. In addition, among both cohorts, NIHSS ≤ 10 was associated with an over three times as likely chance of non-disabled status (mRS ≤ 1; RR = 3.19 [CI 1.53-6.65]) compared to an NIHSS > 10. The difference between NIHSS of ≤ 4 and > 4 was less pronounced. NIHSS ≤ 4 was associated with 1.61 times chance of neurologic functional independence (mRS ≤ 2; RR = 1.61 [CI 1.29-2.01]) compared to an NIHSS > 4. NIHSS ≤ 4 was associated with an over two times as likely chance of non-disabled status (mRS ≤ 1; RR = 2.38 [CI 1.66-3.41]) compared to an NIHSS > 4.
Conclusions: Patients with a higher presenting stroke severity had worse functional outcomes (Figure, Panel D). The thrombolysis cohort presented with more severe strokes (mean NIHSS 7.6) and had a higher 30-day death rate (9.6%). Our data suggests that patients presenting with minor strokes, NIHSS ≤ 4 compared to > 4, were more likely to have neurologic functional independence (mRS ≤ 2) and non-disabled status (mRS ≤ 1) on discharge, regardless of receiving tPA or not. This finding was even more pronounced in moderate strokes, where an NIHSS ≤ 10 in patients undergoing urgent carotid interventions is associated with discharge neurologic independence (mRS ≤ 2) and non-disabled status (mRS ≤ 1) independent of thrombolysis. While there was a trend for all patients who received tPA to have more complications, this was not statistically significant. The risk of death was increased in patients receiving thrombolysis but, this could be due to the increased stroke severity in this cohort. Further study with larger groups is needed to confirm this finding. Overall, presenting NIHSS is predictive of discharge neurologic functional independence and is not influenced by the use of thrombolysis. 
Back to 2023 Abstracts