**Association between CHADS2, CHA2DS2-VASc, ATRIA, and Essen Stroke Risk Scores and Unsuccessful Recanalization after Endovascular Thrombectomy in Acute Ischemic Stroke Patients**

**Hyung Jun Kim 1, Moo-Seok Park 1, Joonsang Yoo 2, Young Dae Kim 3, Hyungjong Park 4, Byung Moon Kim 5, Oh Young Bang 6, Hyeon Chang Kim 7, Euna Han 8, Dong Joon Kim 5, JoonNyung Heo 3, Jin Kyo Choi 9, Kyung-Yul Lee 10, Hye Sun Lee 11, Dong Hoon Shin 12, Hye-Yeon Choi 13, Sung-Il Sohn 4, Jeong-Ho Hong 4, Jong Yun Lee 14, Jang-Hyun Baek 15, Gyu Sik Kim 16, Woo-Keun Seo 6, Jong-Won Chung 6, Seo Hyun Kim 17, Sang Won Han 18, Joong Hyun Park 18, Jinkwon Kim 3, Yo Han Jung 10, Han-Jin Cho 19, Seong Hwan Ahn 20, Sung Ik Lee 21, Kwon-Duk Seo 16, Yoonkyung Chang 22, Tae-Jin Song 1,\*, Hyo Suk Nam 3,\* and on behalf of the SECRET Study Investigators †**


**Citation:** Kim, H.J.; Park, M.-S.; Yoo, J.; Kim, Y.D.; Park, H.; Kim, B.M.; Bang, O.Y.; Kim, H.C.; Han, E.; Kim, D.J.; et al. Association between CHADS2, CHA2DS2-VASc, ATRIA, and Essen Stroke Risk Scores and Unsuccessful Recanalization after Endovascular Thrombectomy in Acute Ischemic Stroke Patients. *J. Clin. Med.* **2022**, *11*, 274. https:// doi.org/10.3390/jcm11010274

Academic Editor: Aristeidis H. Katsanos

Received: 8 December 2021 Accepted: 31 December 2021 Published: 5 January 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).


**Abstract:** Background: The CHADS2, CHA2DS2-VASc, ATRIA, and Essen scores have been developed for predicting vascular outcomes in stroke patients. We investigated the association between these stroke risk scores and unsuccessful recanalization after endovascular thrombectomy (EVT). Methods: From the nationwide multicenter registry (Selection Criteria in Endovascular Thrombectomy and Thrombolytic therapy (SECRET)) (Clinicaltrials.gov NCT02964052), we consecutively included 501 patients who underwent EVT. We identified pre-admission stroke risk scores in each included patient. Results: Among 501 patients who underwent EVT, 410 (81.8%) patients achieved successful recanalization (mTICI ≥ 2b). Adjusting for body mass index and *p* < 0.1 in univariable analysis revealed the association between all stroke risk scores and unsuccessful recanalization (CHADS2 score: odds ratio (OR) 1.551, 95% confidence interval (CI) 1.198–2.009, *p* = 0.001; CHA2DS2VASc score: OR 1.269, 95% CI 1.080–1.492, *p* = 0.004; ATRIA score: OR 1.089, 95% CI 1.011–1.174, *p* = 0.024; and Essen score: OR 1.469, 95% CI 1.167–1.849, *p* = 0.001). The CHADS2 score had the highest AUC value and differed significantly only from the Essen score (AUC of CHADS2 score; 0.618, 95% CI 0.554–0.681). Conclusion: All stroke risk scores were associated with unsuccessful recanalization after EVT. Our study suggests that these stroke risk scores could be used to predict recanalization in stroke patients undergoing EVT.

**Keywords:** ischemic stroke; stroke risk score; recanalization; thrombectomy

#### **1. Introduction**

Endovascular thrombectomy (EVT) plays a pivotal role in improving the prognosis by recanalizing occluded blood vessels in stroke patients [1]. With the recent success of trials on EVT, the number of patients receiving EVT continues to increase [1–5]. Moreover, the time window for EVT has also expanded [4,5]. Nevertheless, a significant number of patients who underwent EVT did not achieve successful recanalization [6]. As unsuccessful recanalization predictably leads to poor patient prognosis, it is important to identify the factors associated with unsuccessful recanalization. Factors associated with such unsuccessful recanalization include greater age, stroke severity, occlusion due to atherosclerosis, and thrombus burden [7–9]. Nonetheless, further research is still needed to identify the factors involved in unsuccessful recanalization [7–9].

Several stroke risk scores have been developed for predicting the clinical outcome or the occurrence of stroke. The CHADS2 [10], CHA2DS2-VASc [11], and ATRIA scores [12] are mainly used to predict thromboembolic risk and vascular outcome in atrial fibrillation (AF) patients. The Essen stroke risk score predicts vascular events in patients without AF [13]. As these stroke risk scores are mainly composed of risk factors and easily identifiable laboratory findings, they have the advantage of being able to easily predict the occurrence of stroke or prognosis.

We hypothesized that stroke risk scores would be associated with unsuccessful recanalization in patients undergoing EVT. Hence, the purpose of this study was to investigate the association between increased CHADS2, CHA2DS2-VASc, ATRIA, and Essen scores and the results of recanalization after EVT.

#### **2. Methods**

#### *2.1. Study Population*

Our study included patients from the Selection Criteria in Endovascular Thrombectomy and Thrombolytic therapy (SECRET) registry (Clinicaltrials.gov NCT02964052). The selection criteria and the definition of included variables in this registry have been published [14]. In brief, the SECRET registry is a nationwide, multicenter registry that included

patients undergoing reperfusion therapy such as EVT [14]. The SECRET registry did not establish strict inclusion or exclusion criteria for reperfusion therapy and recommended treatment according to the updated guideline at the time of treatment. Furthermore, the doctor of each institution determined whether to administer reperfusion therapy, and all patients who underwent reperfusion therapy were consecutively registered in the SECRET registry. All registered clinical and imaging information was reinvestigated and rechecked by the core laboratory after the anonymization process. The demographic data, risk factors for cardiovascular disease, medication history of prior index stroke, blood and urine laboratory examination results, time parameters for reperfusion therapy, neurologic status including severity, and image findings related to reperfusion therapy were investigated.

Between January 2012 and December 2017, we retrospectively enrolled patients who received reperfusion thrombolysis and were consecutively registered in 15 hospitals. In addition, between November 2016 and December 2017, we prospectively enrolled patients who received reperfusion thrombolysis from 13 hospitals. A total of 1231 patients who underwent reperfusion thrombolysis were included, of which 507 patients underwent EVT. Finally, 501 patients who underwent EVT were included, excluding 6 patients, for whom information about the modified thrombolysis in cerebral infarction (mTICI) grade was not acquired (Figure 1). Written informed consent was obtained from the prospectively included patients or their next caregivers. Our Institutional Review Board approved our study (Yonsei University College of Medicine, 4-2015-1196).

**Figure 1.** Patient selection strategy used in the study. tPA, tissue plasminogen activator; EVT, endovascular thrombectomy; mTICI, modified thrombolysis in cerebral infarction.

Stroke severity was defined using the National Institutes of Health Stroke Scale (NIHSS) score, and the neurologic change after 24 h of EVT was defined as the difference between the initial NIHSS score and the NIHSS score at 24 h (Initial NIHSS score—NIHSS score at 24 h = change in NIHSS score after 24 h). Therefore, if this value was positive, it means neurological improvement, 0 means no improvement, and negative means neurological worsened. Time parameters of EVT were acquired from onset-to-start of EVT (onset to puncture time) and administration of intravenous (IV) thrombolysis (tissue plasminogen activator, tPA) to start of EVT (needle to puncture time) [14]. In case of unclear symptom onset time, the last normal time (LNT) when the patient was asymptomatic was considered

as the time of onset. Computed tomography (CT), CT angiography, magnetic resonance imaging (MRI), MR angiography, and digital subtraction angiography (DSA) images were acquired during the admission period.

Data related to reperfusion therapy, for example, the administration of IV thrombolysis, the total trial number of stent-retriever passes and the types of devices were investigated. Intra-arterial (IA) thrombolysis without IV tPA is defined as first-line therapy with EVT, who are contraindicated for IV tPA. Combined IV/IA thrombolysis is defined as IV tPA administration prior to EVT who could be treated with IV tPA within 4.5 h after symptom onset. The status of reperfusion therapy was investigated in the patients who underwent EVT using the final angiographic findings, including the DSA, and graded based on the mTICI grade. For the outcome parameter, a grade of mTICI 2b or 3 was defined as successful recanalization, and a grade of mTICI 0–2a was defined as unsuccessful recanalization. EVT was performed using a stent-retriever technique, a direct aspiration first pass technique (ADAPT), and the Solumbra technique. The first-line technique is based on the clinical situation of each center and each patient. If the first-line technique is unsuccessful, the second-line technique is used. Stent-retriever alone was defined as using only a stentretriever as a first-line technique and not using ADAPT or the Solumbra technique as the second-line technique. Aspiration alone was defined as using ADAPT as a first-line technique and not using any other device as a second-line technique. The type of device used for each technique was based on operator preference (typically Solitaire FR device, Trevor stent device, and Penumbra).

#### *2.2. The Stroke Risk Scoring Systems*

We identified pre-admission CHADS2, CHA2DS2-VASc, ATRIA, and Essen scores for each patient. The variables included in each scoring system are set according to the existing definition. The CHADS2 and CHA2DS2-VASc scores, congestive heart failure, hypertension, age, diabetes mellitus (DM), previous stroke history, vascular diseases, and sex were included as scoring variables [10,11]. The ATRIA score included age, sex, hypertension, DM, congestive heart failure, presence of proteinuria, and kidney dysfunction (estimated glomerular filtration rate <45 mL/min per 1.73 m2) as scoring parameters [12]. The Essen score included age, hypertension, DM, previous stroke history, myocardial infarction history, peripheral arterial occlusive disease, and other vascular diseases [13].

#### *2.3. Statistical Analyses*

Continuous variables and categorical variables were analyzed using an independent *t*test or Mann–Whitney *U* test and the chi-square test or Fisher's exact test, respectively. Uniand multivariable logistic regression was performed to evaluate factors for unsuccessful recanalization. Body mass index (BMI) and onset to puncture time, which are important cofounders for unsuccessful recanalization, and *p* < 0.1 (excluding age and DM, which are common overlapping variables for all stroke risk scores) from the univariable analysis were entered in multivariable analysis. The results of uni- and multivariable analyses were expressed as odds ratios (ORs) and 95% confidence intervals (CIs). Because the risk of vascular outcome increased as the stroke risk scores increased, the main outcome was defined as unsuccessful recanalization in this study. Subgroup analyses were performed, including demographic data, classical vascular risk factors, and stroke risk scores, and were dichotomized by the median values and the optimal cut off values. The interaction between unsuccessful recanalization and each subgroup was investigated with a two-tailed test in the logistic regression analyses. For the sensitivity analysis, we further analyzed all stroke risk scores for patients with AF-related stroke only.

For evaluating the predictability of CHADS2, CHA2DS2-VASc, ATRIA, and Essen scores, receiver operating characteristic (ROC) curve analysis and area under the curve (AUC) were investigated. The AUC was calculated and the optimal cutoff values of the stroke risk scores were defined at the level with the highest Youden index (sensitivity + specificity − 1). The AUC of each stroke risk score was compared to determine whether there was a difference in the predictability of unsuccessful recanalization among the stroke risk scores. We utilized the multivariable model as the benchmark to assess the role of stroke risk scores in enhancing the risk prediction for unsuccessful recanalization in EVT patients. We compared AUCs to assess model discrimination and calculated net reclassification improvement (NRI) and the integrated discrimination improvement (IDI). All statistical analyses were performed using SPSS (version 25.0, IBM Corp., Chicago, IL, USA) and open-source statistical package R version 3.6.3 (R Project for Statistical Computing, Vienna, Austria). All variables needed a *p* < 0.05 to be considered statistically significant.

#### **3. Results**

#### *3.1. Study Population*

A total of 501 patients were included in this study. Patient demographics and information on risk factors and variables are summarized in Table 1. Of the 501 patients receiving EVT, 234 patients (46.7%) were female, and the mean age was 76.2 ± 13.3 years. The median value of the NIHSS scores of all patients was 15 (10–19, interquartile range (IQR)). IV thrombolysis was administered to 202 patients (40.3%), and the mean value of the onset to needle time was 119.9 ± 97.3 min. In all patients who underwent EVT, the mean value of the onset to puncture time was 354.6 ± 440.0 min, the mean value of the needle to puncture time was 78.7 ± 50.8 min, stent-retriever alone was used in 371 patients (74.0%), aspiration alone in 25 patients (4.9%), and combined stent-retriever and aspiration in 90 patients (17.9%). Among the patients who underwent stent-retriever alone and combined stent-retriever/aspiration, information about the stent device was obtained from 440 patients: the Solitaire FR device was used in 377 (85.6%) patients, the Trevor stent device in 58 (13.1%) patients, and both stent devices in only 5 (1.1%) patients. The mean value of the number of stent-retriever passes was 2.1 ± 1.9. Among the patients who underwent aspiration alone and combined stent-retriever/aspiration, aspiration device information was obtained from 92 patients: the Penumbra aspiration system was used in 55 (59.7%) patients and an intermediate catheter device in 37 (40.2%) patients.

Among all included patients, 410 (81.8%) patients achieved successful recanalization (mTICI ≥ 2b). The onset to recanalization measured only for patients who successfully recanalized (mTICI 2b/3) was 429.5 ± 481.4 min.

#### *3.2. Association of Stroke Risk Scores with Recanalization Status*

In the successful recanalization group, the proportion of patients with DM was lower (53.1% vs. 70.3%, *p* = 0.004), and there were more patients with coronary disease (31.2% vs. 17.5%, *p* = 0.013). Patients in the successful recanalization group had lower initial NIHSS scores (median 15 (IQR 10–19) vs. median 17 (IQR 12–20.5), *p* = 0.020) and the change in NIHSS scores after 24 h was greater (Initial NIHSS score—NIHSS score at 24 h, median 5 (IQR 0–10) vs. median 0 (IQR −2–3), *p* < 0.001) than those in the unsuccessful recanalization group. Combined IA/IV thrombolysis was significantly associated with successful recanalization (*p* = 0.030). In patients who administration of tPA prior to EVT, the time interval of the needle to puncture was significantly shorter in the successful recanalization group (111.8 ± 50.1 vs. 73.6 ± 49.1, *p* < 0.001). The stent-retriever alone was associated with successful recanalization (*p* = 0.035). However, aspiration alone (*p* = 0.035) was associated with unsuccessful recanalization. In patients who received stent-retrievers, the number of stent passes was significantly lower in the successful recanalization group (2.9 ± 2.8 vs. 2.0 ± 1.6, *p* = 0.002). In laboratory tests, both initial glucose level after admission (152.9 ± 54.3 mg/dL vs. 140.2 ± 49.7 mg/dL, *p* = 0.042) and fasting glucose level after admission (148.8 ± 52.9 mg/dL vs. 128.9 ± 46.9 mg/dL, *p* = 0.002) were lower in the successful recanalization group. All stroke risk scores were significantly lower in the successful recanalization group (CHADS2 score; median 2 (IQR 1–3) vs. 3 (IQR 2–3), *p* < 0.001) (CHA2DS2VASc score; median 3 (IQR 2–4] vs. 4 (IQR 3–5], *p* = 0.002) (ATRIA score; median 7 (IQR 3–9) vs. 9 (IQR 6–10), *p* = 0.002) (Essen score; median 3 (IQR 2–4) vs. 4 (IQR 3–4), *p* = 0.034) (Table 1).


**Table 1.** Clinical and imaging characteristics according to the degree of recanalization.

mTICI, modified thrombolysis in cerebral infarction; SD, standard deviation; BMI, body mass index; eGFR, estimated glomerular filtration rate; National Institutes of Health Stroke Scale, NIHSS; IQR, interquartile range; tPA, tissue plasminogen activator; IA, int; IV, intravenous; LNT, last normal time; ICA, internal carotid artery; MCA, middle cerebral artery; ACA, anterior cerebral artery; PCA, posterior cerebral artery; V-B, vertebro-basilar. \* administration of intravenous tissue plasminogen activator prior to endovascular thrombectomy; \*\* cases in which stent-retriever and aspiration were performed simultaneously or sequentially. † The glucose level test was performed at the time of the first admission to the emergency room. ‡ The glucose level test was performed after 8 h of fasting after admission.

In univariable logistic regression analysis, age, DM, coronary disease, initial NIHSS score, combined IA/IV thrombolysis, stent-retriever alone, aspiration alone, number of stent-retriever passes, and stroke risk scores were associated with unsuccessful recanalization, as shown in Table S1. In multivariable logistic regression analysis, all stroke risk scores were predictive of unsuccessful recanalization along with BMI, onset to puncture time, coronary disease, initial NIHSS score, combined IA/IV thrombolysis, stent-retriever alone, aspiration alone, and the number of stent-retriever passes (CHADS2 score: OR 1.551, 95% CI 1.198–2.009, *p* = 0.001; CHA2DS2VASc score: OR 1.269, 95% CI 1.080–1.492, *p* = 0.004; ATRIA score: OR 1.089, 95% CI 1.011–1.174, *p* = 0.024; and Essen score: OR 1.469, 95% CI 1.167–1.849, *p* = 0.001) (Table 2).

**Table 2.** Multivariable analysis for stroke risk score associated with the unsuccessful recanalization among 501 patients with endovascular thrombectomy.


OR, odds ratio; CI, confidence interval; BMI, body mass index; National Institutes of Health Stroke Scale, NIHSS; IV, intravenous; IA, intra-arterial; tPA, tissue plasminogen activator; EVT, endovascular thrombectomy; \* administration of intravenous tissue plasminogen activator prior to endovascular thrombectomy.

In subgroup analysis, the association of unsuccessful recanalization was stratified by age, sex, comorbidities, NIHSS score, treatment factor (LNT, combined IA/IV thrombolysis), and stroke risk scores. A subgroup of patients with DM (*p* for interaction = 0.003), IA alone (*p* for interaction = 0.023), CHA2DS2VASc score ≥ 4 (*p* for interaction = 0.022), ATRIA score ≥ 8 (*p* for interaction = 0.017), CHADS2 score ≥ 3 (*p* for interaction < 0.001), CHA2DS2VASc score ≥ 5 (*p* for interaction < 0.001), ATRIA score ≥ 9 (*p* for interaction = 0.001), and Essen score ≥ 4 (*p* for interaction = 0.009) were significantly associated with unsuccessful recanalization (Figure 2).

**Figure 2.** Forest plots of unadjusted odds ratios for unsuccessful recanalization (mTICI ≤ 2a) in patients with endovascular thrombectomy. BMI, body mass index; NIHSS, National Institutes of Health Stroke Scale; LNT, last normal time; IA, intra-arterial; IV, intra-venous; ORs, odds ratios; mTICI, modified thrombolysis in cerebral infarction.

In the comparison with AF-related stroke, there were significantly fewer patients with DM (72.1% vs. 53.6%, *p* = 0.042) and more patients with stent-retriever alone (51.2% vs. 80.1%, *p* < 0.001) in the successful recanalization group. Moreover, patients in the successful recanalization group had lower initial NIHSS scores (median 15 (IQR 10–19) vs. median 17 (IQR 12–20.5), *p* = 0.020) and the change in NIHSS scores after 24 h was greater (Initial NIHSS score—NIHSS score at 24 h, median 5 (IQR 1–10) vs. median 0 (IQR −1–2), *p* < 0.001) than those in the unsuccessful recanalization group. An increase in the number of stent-retriever passes was associated with unsuccessful recanalization (3.3 ± 3.0 vs. 2.0 ± 1.5, *p* = 0.007). In laboratory tests, fasting glucose after admission (146.24 ± 50.3 mg/dL vs. 128.4 ± 51.9 mg/dL, *p* = 0.044) were lower in the successful recanalization group. The CHADS2, CHA2DS2VASc, ATRIA, and Essen scores were significantly lower in the successful recanalization group (CHADS2 score; median 2 (IQR 2–3) vs. 3 (IQR 2–3), *p* < 0.001) (CHA2DS2VASc score; median 4 (IQR 3–4] vs. 5 (IQR 4–5.5], *p* = 0.003) (ATRIA score; median 8 (IQR 6–9) vs. 9 (IQR 7.5–10), *p* = 0.033) (Essen score; median 4 (IQR 3–4) vs. 3 (IQR 3–4), *p* = 0.043). The above results are summarized in Table S2. In multivariable logistic regression analysis, the CHADS2, CHA2DS2VASc, and Essen scores were associated with unsuccessful recanalization along with BMI, coronary disease, the initial NIHSS score, combined IA/IV thrombolysis, stent-retriever alone, aspiration alone, and number of stent-retriever passes (CHADS2 score: OR 1.787, 95% CI 1.173–2.725, *p* = 0.007; CHA2DS2VASc score: OR 1.354, 95% CI 1.049–1.747, *p* = 0.020; and Essen score: OR 1.635, 95% CI 1.093–2.448, *p* = 0.017) (Table S3).

#### *3.3. Comparison of Stroke Risk Scores for Unsuccessful Recanalization*

Figure 3 shows the ROC curves of all stroke risk scores for unsuccessful recanalization. The AUC, optimal cutoff value, sensitivity, specificity, positive predictive value, and negative predictive value of each stroke risk score are presented in Table 3.

**Figure 3.** Receiver operating characteristic curve analyses of unsuccessful recanalization based on stroke risk scores. (**A**) Univariable ROC analysis (**B**) Multivariable ROC analysis. ROC, receiver operating characteristic.

Among stroke risk scores, the CHADS2 score had the highest AUC value. However, in pairwise comparisons of the AUC, only the CHADS2 and Essen scores were significantly different (AUC of CHADS2 score; 0.618, 95% CI 0.554–0.681 vs. AUC of Essen score; 0.569, 95% CI 0.506–0.632, *p* = 0.002) (Table S4). Similarly, even when ROC curve analysis was performed on AF-related stroke only, the CHADS2 score had the highest AUC value, and only the CHADS2 and Essen scores were significantly different (AUC of CHADS2 score; 0.666, 95% CI 0.570–0.761 vs. AUC of Essen score; 0.595, 95% CI 0.502–0.687, *p* = 0.006) (Table S5).


**Table 3.** Receiver operating characteristic (ROC) curve analysis of risk scores for the probability of an unsuccessful recanalization.

AUC, area under the curve; PPV, positive predictive value; NPV, negative predictive value.

The continuous-based NRI was significantly improved after the addition of each stroke risk score (CHADS2 score: *p* = 0.010, CHA2DS2VASc score: *p* = 0.023, ATRIA score: *p* = 0.035, Essen score: *p* = 0.005). The IDI also showed improved risk classification after the addition of the CHADS2 score (*p* = 0.014) or the ATRIA score (*p* = 0.015). Overall, the best model for prediction of unsuccessful recanalization after EVT was the CHADS2 score, with the addition of the multivariable model (Table S6).

#### **4. Discussion**

The key finding of this study was that the pre-admission CHADS2, CHA2DS2VASc, ATRIA, and Essen scores were associated with unsuccessful recanalization after EVT. The probability of unsuccessful recanalization increased as the stroke risk scores increased. The CHADS2 score had the highest AUC among all stroke risk scores, although the CHADS2 score differed significantly only from the Essen score.

Previous studies have proven the relationship between stroke risk scores and the clinical outcome of stroke. CHADS2, CHA2DS2-VASc, and ATRIA scores are simple to obtain and are useful tools for estimating the thromboembolic risk and clinical outcomes in patients with AF [15–18]. The Essen score is a simple clinical score that was derived to predict the 1-year risk of recurrent ischemic stroke after ischemic stroke based on the presence of prior vascular comorbidities [13,19]. Unlike the purpose for which the stroke risk scores were developed, the stroke risk scores have been used as a predictor of various outcomes in various patient groups [15,20,21]. Our results are meaningful in that they provide additional information that CHADS2, CHA2DS2-VASc, ATRIA, and Essen scores were correlated with unsuccessful recanalization in patients undergoing EVT, as well as thromboembolic risk and clinical outcome. All stroke risk scores were associated with unsuccessful recanalization even in AF-related stroke patients.

Hypertension and DM have a common weighting factor for all stroke risk scores, and DM, in particular, is known to affect recanalization after IV thrombolysis [22,23]. Although DM and fasting hyperglycemia are also known to affect clinical outcomes after EVT [24,25], there is still insufficient evidence that DM and initial and fasting hyperglycemia influence recanalization after EVT [26,27]. Our results showed that DM, initial, and fasting glucose levels were associated with unsuccessful recanalization. Factors other than hypertension and DM were weighted differently for each stroke risk score. Compared with the CHADS2 score, the CHA2DS2-VASc score has a higher weighting for age and includes the components of sex and vascular diseases; the ATRIA score has a higher weighting for age and includes the components of sex and chronic renal disease, while the Essen score also has a higher weighting for age and sex, similarly to the other scores, along with weighting for vascular disease and current smoking. Each additional component of these scores has been reported as a predictor of stroke severity or outcome [28]. The ATRIA and CHA2DS2-VASc scores were reported to outperform the CHADS2 score in predicting stroke outcome in patients with AF [17,28]. However, unlike previous studies that investigated the relationship between the stroke risk scores and stroke outcome, we found that the CHADS2 score shows better performance in predicting recanalization than the CHA2DS2-VASc and ATRIA scores in EVT patients. This may be because sex, chronic kidney disease, and vascular disease

weighted by the CHA2DS2-VASc and ATRIA scores did not differ with recanalization, and there were more patients with coronary disease in the successful recanalization group in our dataset. As the Essen score is a risk-scoring tool for the prediction of recurrent stroke and combined cerebrovascular events in patients with non-AF, there have been a few studies comparing the performance of the Essen score with that of the CHADS2 score [13]. A recent observational study of the prediction for vascular outcome in stroke patients with AF found no significant difference in the performance between the CHADS2 and Essen scores [21]. In contrast, the CHADS2 score showed significantly better performance than the Essen score in our study. Even when only AF-related stroke patients were analyzed, the Essen score could significantly predict unsuccessful recanalization, although its performance was worse than that of the CHADS2 score. Presumably, as in the case of CHA2DS2-VASc and ATRIA scores, this could be attributed to the observation that peripheral artery disease and current smoking weighted by the Essen score did not differ according to recanalization, and there were more patients with coronary disease in the recanalization group in our dataset. Therefore, unlike previous studies on patients with AF stroke, the ATRIA and CHA2DS2-VASc scores likely did not outperform and the Essen score likely underperformed compared to the CHADS2 score. The significance of this result suggests that most of the factors related to unsuccessful recanalization in EVT patients overlap with most factors related to the occurrence of stroke in AF patients. Therefore, these different factors should be taken into account when creating a new scoring system that predicts recanalization after EVT. An existing pre-admission stroke risk score or suitable new scoring system can be used in addition to the current image-based patient's selection system, which can contribute to lower recanalization failure rates by appropriately selecting patients.

#### *Limitations*

First, although some of the patients included in our study were prospectively included and the registry itself consecutively included stroke patients who received reperfusion therapy, we performed a retrospective evaluation. Therefore, there may be selection bias, l and the possibility of a causal relationship cannot be concluded. Second, this registry is a nationwide observational registry that reflects real-world evidence; however, there may be a selection bias because it is not a randomized controlled study. To reduce the selection bias, we consecutively included patients eligible for EVT according to the valid guidelines [1,29,30]. Third, because our registry enrolled only the Korean population, it is difficult to generalize our findings to all races.

#### **5. Conclusions**

The pre-admission CHADS2, CHA2DS2VASc, ATRIA, and Essen scores were associated with unsuccessful recanalization after EVT. Therefore, these results suggest that stroke risk scores, especially the CHADS2 score, could predict recanalization in stroke patients undergoing EVT.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jcm11010274/s1, Table S1: Univariate logistic regression analysis of the risk of an unsuccessful recanalization; Table S2: Clinical and imaging characteristics according to the degree of recanalization (Only atrial fibrillation-related stroke); Table S3: Multivariate analysis for stroke risk score associated with the unsuccessful recanalization among atrial fibrillation-related stroke with endovascular thrombectomy; Table S4: Comparison of the area under curve (AUC) of each stroke risk score by two. (Univariate ROC analysis); Table S5: Comparison of area under the curve (AUC) of each stroke risk score by two in AF-related stroke. (Univariate ROC analysis); Table S6: Receiver-operating characteristics curve analysis (area under curve), net reclassification improvement, and integrated discrimination improvement of predictive models for unsuccessful recanalization in endovascular thrombectomy patients.

**Author Contributions:** Conceptualization, T.-J.S. and H.S.N.; methodology, H.J.K., T.-J.S. and H.S.N.; data acquisition, M.-S.P., J.Y., Y.D.K., H.P., B.M.K., O.Y.B., H.C.K., E.H., D.J.K., J.H., J.K.C., K.-Y.L., H.S.L., D.H.S., H.-Y.C., S.-I.S., J.-H.H., J.Y.L., J.-H.B., G.S.K., W.-K.S., J.-W.C., S.H.K., S.W.H., J.H.P., J.K., Y.H.J., H.-J.C., S.H.A., S.I.L., K.-D.S., Y.C. and on behalf of the SECRET study Investigators; writing—original draft preparation, H.J.K. and T.-J.S.; writing—review and editing, H.J.K., T.-J.S. and H.S.N. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2021R1F1A1048113). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

**Institutional Review Board Statement:** The SECRET study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of the Yonsei University College of Medicine, 4-2015-1196.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the SECRET study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding authors. The data are not publicly available due to privacy.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Article* **Advanced Neuroimaging Preceding Intravenous Thrombolysis in Acute Ischemic Stroke Patients Is Safe and Effective**

**Klearchos Psychogios 1,2,\*, Apostolos Safouris 1,2, Odysseas Kargiotis 1, Georgios Magoufis 1, Athina Andrikopoulou 1, Ermioni Papageorgiou 1, Maria Chondrogianni 1,2, Georgios Papadimitropoulos 1, Eftihia Polyzogopoulou 3, Stavros Spiliopoulos 4, Elias Brountzos 4, Elefterios Stamboulis 1, Sotirios Giannopoulos <sup>2</sup> and Georgios Tsivgoulis <sup>2</sup>**


**Abstract:** Advanced neuroimaging is one of the most important means that we have in the attempt to overcome time constraints and expand the use of intravenous thrombolysis (IVT). We assessed whether, and how, the prior use of advanced neuroimaging (AN), and more specifically CT/MR perfusion post-processed with RAPID software, regardless of time from symptoms onset, affected the outcomes of acute ischemic stroke (AIS) patients who received IVT. Methods. We retrospectively evaluated consecutive AIS patients who received intravenous thrombolysis monotherapy (without endovascular reperfusion) during a six-year period. The study population was divided into two groups according to the neuroimaging protocol used prior to IVT administration in AIS patients (AN+ vs. AN−). Safety outcomes included any intracranial hemorrhage (ICH) and 3-month mortality. Effectiveness outcomes included door-to-needle time, neurological status (NIHSS-score) on discharge, and functional status at three months assessed by the modified Rankin Scale (mRS). Results. The rate of IVT monotherapy increased from ten patients per year (*n* = 29) in the AN− to fifteen patients per year (*n* = 47) in the AN+ group. Although the onset-to-treatment time was longer in the AN+ cohort, the two groups did not differ in door-to-needle time, discharge NIHSS-score, symptomatic ICH, any ICH, 3-month favorable functional outcome (mRS-scores of 0–1), 3-month functional independence (mRS-scores of 0–2), distribution of 3-month mRS-scores, or 3-month mortality. Conclusion. Our pilot observational study showed that the incorporation of advanced neuroimaging in the acute stroke chain pathway in AIS patients increases the yield of IVT administration without affecting the effectiveness and safety of the treatment.

**Keywords:** acute stroke; intravenous thrombolysis; perfusion imaging; CT perfusion; MR perfusion; RAPID

#### **1. Introduction**

Intravenous thrombolysis (IVT) with alteplase in acute ischemic stroke (AIS) administered within the first 4.5 hours following symptom onset remains the mainstay of acute reperfusion therapies [1–3]. Despite tissue plasminogen activator (tPA) effectiveness, only a small number of AIS patients worldwide benefit from IVT [4,5]. Short therapeutic time window, strict inclusion and exclusion criteria of the pivotal randomized controlled clinical trials (RCTs), as well as health care system disparities, such as public awareness on how to

**Citation:** Psychogios, K.; Safouris, A.; Kargiotis, O.; Magoufis, G.; Andrikopoulou, A.; Papageorgiou, E.; Chondrogianni, M.; Papadimitropoulos, G.; Polyzogopoulou, E.; Spiliopoulos, S.; et al. Advanced Neuroimaging Preceding Intravenous Thrombolysis in Acute Ischemic Stroke Patients Is Safe and Effective. *J. Clin. Med.* **2021**, *10*, 2819. https://doi.org/10.3390/ jcm10132819

Academic Editors: Hyo Suk Nam and Byung Moon Kim

Received: 31 May 2021 Accepted: 24 June 2021 Published: 26 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

act in case of stroke symptoms, organization of emergency medical services, and the paucity of organized stroke centers in rural areas [6], have been significant barriers to overcome. Nevertheless, off-label use of IVT [7,8] is increasingly incorporated in the everyday clinical practice of many stroke practitioners.

Advanced neuroimaging may help us overcome time constraints and expand the implementation of acute reperfusion therapies [9]. CT and MR perfusion with automated post-processing software (RAPID, iSchemaView, Menlo Park, CA, USA) have proven effective in recent RCTs, for both mechanical thrombectomy candidates in the late time window (6–24 h) [10,11] and for IVT (4.5–9 h and wake-up patients) [12–14]. Advanced neuroimaging provides a "brain physiology snapshot in time" that can guide decisions for recanalization therapies in clinical practice [15]. Numerous stroke centers and stroke units worldwide have incorporated the use of CT and MR perfusion in their acute therapeutic pathways.

In view of the former considerations, we assessed the differences in the use of IVT monotherapy and the outcomes of the AIS patients with or without the use of advanced neuroimaging.

#### **2. Materials and Methods**

We retrospectively evaluated consecutive AIS patients who received IVT admitted to our European Stroke Organization certified stroke unit. We also participate in the SITS (Safe Implementation of Thrombolysis in Stroke) and RES-Q (Registry of Stroke Care Quality) international registries [16,17]. Patients were included if they fulfilled the following criteria: (1) aged over 18 years old; (2) clinically diagnosed with AIS with a measurable neurologic deficit on the National Institute of Health Stroke Scale (NIHSS) presenting within the 4.5 h window from symptom onset; (3) AIS patients were considered eligible for the extended time window of 4.5–9 h if they presented after 4.5 h and sooner than 9 h from last-seen-well (late window patients), according to the clinical and neuroimaging inclusion criteria of the EXTEND trial [10]; (4) AIS patients who woke up with symptoms of stroke («wake-up stroke») were treated according to the WAKE-UP trial [18] protocol; and (5) AIS patients treated with IVT monotherapy. All patients with large vessel occlusion (LVO) who underwent mechanical thrombectomy were excluded. Transient ischemic attacks and stroke mimics were excluded from the current study based on clinical and neuroimaging criteria.

The study population was divided into two different groups according to the neuroimaging protocol used on admission and prior to IVT administration in AIS patients (with prior Advanced Neuroimaging (AN+) vs. without prior advanced neuroimaging (AN−)). Of note, the neuroimaging protocol was modified in our center on December 2017 after the introduction of perfusion imaging (with RAPID software) and on August 2018 after the publication of the WAKE-UP trial. Patients in the first study group (AN−) underwent baseline emergent neurovascular imaging using either non-contrast-enhanced computed tomography (NCCT), with or without CT angiography (CTA), or magnetic resonance imaging (MRI) with magnetic resonance angiography based on the treating physician's decision. Patients in the second study group (AN+) underwent NCCT/CTA/computed tomography perfusion (CTA/CTP) or magnetic resonance angiography/magnetic resonance perfusion (MRA/MRP) unless they presented certain contraindications (e.g., renal insufficiency, severe allergic reactions to iodinised agents, etc.). CT perfusion was performed using two continuous 2.5 cm slabs, starting at the level of the circle of Willis for most patients, lower for those presenting with symptoms suggesting posterior fossa ischemia, and higher for those presenting with symptoms suggestive of cortical ischemia. Ischemic core (rCBF < 30%), critically hypoperfused ischemic region (Tmax > 6 s), and mismatch volume corresponding to ischemic penumbra, were estimated by using RAPID as previously described [19]. The hyperdense vessel sign (HVS), a highly specific marker of arterial obstruction [20], was identified on non-contrast CT if the lumen of any, non-calcified, intracranial artery appeared denser than adjacent or equivalent contralateral arteries. Clot length was quantified based on CT angiography by using standard methodology [21]. The LVO was defined as the occlusion of the internal carotid artery (ICA), basilar artery (BA), and the first segment of the Middle Cerebral Artery (MCA-M1). CT/MR findings were interpreted and extracted independently by experienced neurologists or neuroradiologists that were blinded to clinical outcomes.

The following parameters were recorded for all included patients: (1) demographic characteristics; (2) history of vascular risk factors (diabetes mellitus, hypertension, current smoking, hypercholesterolemia, coronary artery disease, peripheral artery disease, congestive heart failure, and valvular disease) as previously described [22]; (3) prior history of stroke or Transient Ischemic Attack (TIA); (4) laboratory test values on admission (total platelet count, glucose, and low-density lipoprotein (LDL) levels); and (5) admission systolic and diastolic blood pressures, measured using automated blood pressure cuffs. Stroke severity was assessed with the NIHSS (National Institute of Health Stroke Scale) score at admission, 2 h and 24 h post IVT, and at discharge. Safety outcomes included prevalence of symptomatic intracranial hemorrhage (sICH), prevalence of any intracranial hemorrhage in the 24-h post thrombolysis neuroimaging studies, and 3-month mortality. sICH was defined using standard SITS registry definitions (local or remote parenchymatous hemorrhage type 2 combined with an NIHSS-score increase of >4 points or leading to death\22–36 h) [14]. Any intracranial hemorrhage was recorded according to the ECASS criteria [23]. Effectiveness outcomes included door-to-needle time, neurological improvement at 24 h and on discharge, and functional status at discharge and at 3 months by using the modified Rankin Scale (mRS). Functional independence (FI) and favorable functional outcome (FFO) were defined as an mRS-score of 0–2 or an mRS-score of 0–1 at 3 months, respectively. Stroke severity and functional outcome (mRS) at discharge and at 3 months were assessed by certified vascular neurologists as previously described [24].

All follow-up evaluations occurred at 90 ± 10 days from symptom onset at the Stroke Outpatient Clinic of our institution as previously described [25]. The evaluation of the mRS-score was performed by certified vascular neurologists who were unaware of the neuroimaging protocol that was implemented at baseline.

#### *Statistical Analysis*

All binary variables were presented as percentages, while continuous variables were presented with their corresponding mean values and standard deviations (SDs), in cases of normal distributions, or as medians with interquartile ranges (IQRs) in cases of skewed distributions. Statistical comparisons between the two groups were performed using the unpaired t test, Mann–Whitney U-test, *χ*<sup>2</sup> test, and Fisher exact test, as appropriate. The distribution of the 3-month mRS scores between patients treated before and after RAPID implementation was compared using the Cochran–Mantel–Haenszel test and the univariable/multivariable ordinal logistic regression (shift analysis).

All efficacy and safety outcomes of interest were further assessed in univariable and multivariable binary logistic regression models adjusting for the a priori defined confounders of the age and baseline NIHSS-score. The final variables that were independently associated in the multivariable logistic and the ordinal regression analyses with the outcome of interest, were selected using an alpha value of 0.05 and adjusted associations were provided as odds ratios (ORs) or common odds ratios (cORs), with their corresponding 95% confidence intervals (95% CI).

All statistical analyses were conducted with the Stata Statistical Software Release 13 (StataCorp LP, College Station, TX, USA).

#### **3. Results**

A total of eight hundred and nineteen patients were screened in the setting of an acute stroke code between February 2015 and January 2021. The complete flowchart of our study is shown in Figure 1. Three hundred and seventy-seven patients were screened before December 2017 (AN implementation) and twenty-six received IVT, whereas four hundred and forty-two were screened after December 2017 and fifty patients among them received IVT (three of them were not screened with prior AN due to contraindications).

Our final cohort was comprised of 76 AIS patients who received IVT throughout the entire study period. All patients who received endovascular reperfusion therapy with mechanical thrombectomy were excluded from our analysis (*n* = 71). Twenty-nine patients received IVT without prior advanced neuroimaging (AN−) and forty-seven patients with the use of advanced neuroimaging (AN+). The rate of IVT monotherapy increased from ten patients per year in the AN− to fifteen patients per year in the AN+ group. Baseline characteristics of the two treatment groups are summarized in Table 1. Patients in the AN+ group were significantly (*p* = 0.003) older than patients in the AN− group (mean age 73 years vs. 63 years, respectively). Median admission NIHSS-scores were 4 points (IQR: 2–7) in the AN− group and 5 points (IQR: 4–9) in the AN+ group, a difference that was also significant (*p* = 0.047). The prevalence of large vessel occlusions was 17.2% in the AN− group and 19.1% in the AN+ group (*p* = 0.835). The location of stroke in posterior circulation was more frequent in the AN− group (34.5%) than in the second study group (19.1%). The median elapsed time between symptom onset (or last-seen-well) to initiation of IVT was significantly longer in the second group (198 min (IQR: 151–240)) in AN+ vs. 121 min ((IQR: 130–220) in AN−; *p* < 0.001), whereas the door-to-needle time was almost identical between the two groups (median 44 min (IQR: 36–60)) in AN− vs. 45 min ((IQR: 30–61) in AN+; *p* = 0.956). The rate of patients treated according to the EXTEND trial or WAKE-UP protocol was significantly higher in the second study group (23.4% vs. 3.4%; *p* = 0.020). All patients were treated with alteplase, except for four patients in the AN+ with large vessel occlusions who were treated with tenecteplase.

**Figure 1.** Flowchart of the study population. Acute Ischemic Stroke, AIS; Large vessel occlusion, LVO; intravenous thrombolysis, IVT; Advanced Neuroimaging, AN.


**Table 1.** Baseline characteristics in patients treated before and after the implementation of advanced neuroimaging.

Blood pressure, BP; National Institute of Health Stroke Scale, NIHSS; interquartile range, IQR; Alberta Stroke Program Early CT score, ASPECTS; standard deviation, SD.

The neuroimaging characteristics are summarized in Tables 1 and 2. The median thrombus length tended to be higher in the AN+ group (12 vs. 9 mm, *p* = 0.053). The median ASPECTS score and the presence of a hyperdense vessel sign were similar across

the study groups. MR imaging was performed in 6.9% of AN− and 10.3% of AN+ patients (*p* = 0.584). In patients who underwent perfusion imaging, the mean ischemic core volume was calculated at 2.1 ± 1.2 mL and the mean volume of critical hypoperfusion was 16.3 ± 4.0 mL (Table 2).

**Table 2.** Neuroimaging characteristics of patients treated after the implementation of perfusion imaging.


Table 3 summarizes the effectiveness and the safety outcomes in the two patient groups. There was only one missing 3-month follow-up evaluation in each treatment group. Neurological status assessed by NIHSS at 2 h, 24 h, and at hospital discharge was similar between the two groups. The rates of sICH (3.4% vs. 0%; *p* = 0.2) and any intracranial hemorrhage (6.9% vs. 10.6%; *p* = 0.584) were similar between the two groups. The rates of 3-month favorable functional outcome (75% vs. 78.3%; *p* = 0.746), 3-month functional independence (82.1% vs. 89.1%; *p* = 0.394), and 3-month mortality (0% vs. 4.3%; *p* = 0.263) did not differ between the two groups either. A secondary analysis restricted to the patients in the early time window shows similar results (Supplemental Table S1).



National Institute of Health Stroke Scale, NIHSS. \* mRS-scores of 0–2. \*\* mRS-scores of 0–1. \*\*\* Cochran–Mantel– Haenszel test.

The distribution of 3-month mRS-scores was similar between the two groups (*p* for Cochran–Mantel–Haenszel test: 0.466). Table 4 shows the univariable and multivariable associations of the neuroimaging protocol with safety and efficacy outcomes in multivariable

logistic regression models adjusting for the age and admission NIHSS-score. There was no association between the advanced neuroimaging protocol and any ICH (crude OR 1.60, 95% CI: 0.29–8.88; *p* = 0.586), functional independence at three months (crude OR 1.78, 95% CI: 0.47–6.8; *p* = 0.398), and favorable functional outcome at three months (crude OR 1.20, 95% CI: 0.40–3.63; *p* = 0.747). In adjusted analysis AN was associated with better functional independence at 3 months (adjusted OR 12.89, 95% CI: 1.47–113; *p* = 0.021).

**Table 4.** Univariable and multivariable binary logistic regression analyses evaluating the association of the use of advanced neuroimaging in acute stroke chain pathway with outcomes.


Odds ratio, OR; confidence intervals, CI. \* Adjusted for the age and baseline NIHSS score.

#### **4. Discussion**

Our pilot observational single-center study showed that the shift in our clinical practice, with the incorporation of advanced neuroimaging in AIS patients, increases the yield of IVT administration by approximately 50% without major effectiveness and safety repercussions. On the contrary, all comparisons showed that it is equally safe, and even in a population with more negative prognostic factors (higher admission NIHSS-score, older age, longer thrombus), we documented a trend towards better functional outcomes without any delays in door-to-needle time. Better outcomes in patients with prior AN possibly reflect the comparison between different study periods and the accumulating experience of the stroke team through the years. It might also encompass the more favorable prognosis of patients treated in the extended time window, already proven by large clinical trials [12]. However, this result should be treated with caution given the large confidence intervals due to our small study sample and the fact that it was not demonstrated in the crude analysis as well.

Almost 25% of patients in the advanced neuroimaging group were treated based on neuroimaging criteria (either extended time window 4.5–9 h or wake-up strokes, see Figure 2) and this further substantiates our previous observations [20]. Considering that the extra time needed to perform the CT perfusion and to acquire the RAPID templates is at least ten min, it is striking that the median door-to-needle time was only one min longer in the advanced neuroimaging group compared to the median door-to-needle time in the standard neuroimaging group. This observation reflects the interplay of many other important key factors: the acquired experience of the personnel who are involved in the acute stroke chain, the increased use of perfusion imaging particularly in "borderline" cases (e.g., stroke mimics) [26] that otherwise would necessitate two different imaging modalities (CT and MRI), and the fact that the clinical decision in most cases was made immediately after the non-contrast CT and IVT could be initiated in the radiology department before completion of the perfusion imaging.

**Figure 2.** This is an illustrative case of a patient fulfilling both neuroimaging and clinical EXTEND eligibility criteria who was treated successfully with intravenous thrombolysis in the extended time window. An 80-year-old woman was transferred from an island to the emergency department 5 h after an acute onset of expressive aphasia, mild right facial paresis, and mild right upper arm paresis (NIHSS score 9 points). (**a**) Her CT-perfusion mismatch map post-processed with RAPID software demonstrated a hypoperfused region of 11 mL in the Broca's area (shown in green) and no area of reduced cerebral blood flow, resulting in a 11 mL mismatch difference (infinite mismatch ratio). (**b**,**c**) CT angiogram revealed no large vessel occlusion. The patient fulfilled all EXTEND eligibility criteria; IVT with alteplase started 5 h and 45 min after symptom onset with partial resolution of symptoms at the end of tPA infusion (NIHSS-score of 6 points). (**d**) Repeat MRI at 24 h demonstrated a small insular infarct and another acute infarct in the left temporoparietal region which was captured in the Tmax maps of initial perfusion imaging as Tmax > 4 s prolongation (**c**/arrow). The patient's mRS-score at three months was 0.

The present study investigated the effect of advanced neuroimaging on IVT monotherapy. Patients who received endovascular reperfusion therapy were excluded from our analyses. Consequently, our cohort included predominantly mild to moderate severity strokes with a small ischemic core and penumbra volumes or patients with LVO who responded to IVT with successful reperfusion and did not need further endovascular treatment. This probably induces a selection bias by excluding AIS patients with a more "unfavorable prognosis". Previous studies [27,28] that served as pilot studies for the major MT RCTs, have underscored the feasibility of this physiologic imaging approach in cases with LVO-attributed ischemic stroke. Major RCTs that also used the same approach in the early time window [29,30] showed even greater treatment effects, substantially enhancing the use of this approach in clinical practice.

The use of perfusion imaging in AIS patients who present in the first 4.5 h after symptoms onset is still controversial. In our cohort, patients who did not present with a favorable profile (based on neuroimaging criteria) in the early time window, were still offered tPA according to current recommendations. The majority of these patients (n = 19) had no ischemic core or had only hypoperfusion that did not meet the Tmax > 6 s typical criteria of the penumbra. Some of these patients (4/19, 21%) had a "benign oligemia" profile with Tmax prolongation > 4 s, but with either ongoing clinical symptoms or symptoms in partial resolution. This could be due to technical issues (lesion outside the selected slabs when

CT perfusion was used), lacunar infarcts [31], spontaneous recanalization before imaging, or small lesions in the posterior circulation [32] where CT perfusion has lower sensitivity. However, it may also imply that among the "benign oligemia" regions, there might exist grey zones close to the Tmax 6 s threshold delay that correspond more to critical hypoperfused areas, and which, if left untreated, may lead to permanent neurological deficits. Indeed, the DEFUSE study [33] showed that among patients who did not experience early reperfusion, Tmax > 4 s threshold was more accurate in predicting final infarct volume. Even though Tmax > 6 s has been proven to be the best perfusion measurement marker in predicting clinical outcome [34,35] after successful recanalization, infarct growth is perhaps a more complex process influenced by many clinical and pathological factors.

Based on current knowledge, perfusion imaging may not be critical for therapeutic decisions in the early time window by excluding patients with large ischemic core or those with no or minimal perfusion deficit. For instance, the "too good to treat" pattern [36] of small distal perfusion lesions with no vessel occlusion, needs to be studied in larger populations and with more potent thrombolytic agents, including tenecteplase. Even though time since last-seen-well is a poor proxy for perfusion status, we are far from changing the paradigm of IVT administration and endovascular treatment in the early time window from time-based to imaging-based. Nevertheless, in the era of precision medicine and shared decision-making [37], perfusion imaging may still provide additional support to the clinician: for instance, to communicate the decisions with the patient and the patient proxies, strengthen the diagnostic confidence by excluding stroke mimics, accelerate the processes in fast-progressors, and possibly, predict prognosis.

Certain limitations of the present pilot study need to be acknowledged including the single-center retrospective design and analysis of a prospectively maintained patient database, the relatively small sample size, the lack of randomization, and blinding in the evaluation of clinical outcomes. In addition, a major limitation is the heterogeneity induced by the comparison of data from different time periods where practices and experiences of the involved personnel are changing and protocols are reviewed and updated periodically.

#### **5. Conclusions**

In conclusion, the implementation of advanced neuroimaging in unselected AIS patients receiving reperfusion monotherapy with IVT, results in an increase of tPA administration rates without delaying door-to-needle time and without raising safety or effectiveness concerns.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/jcm10132819/s1, Table S1: Outcomes in patients treated before and after the implementation of AN, confined in patients treated with IVT during the early time window.

**Author Contributions:** Conceptualization, K.P. and G.T.; writing—original draft preparation, K.P. and G.T.; writing—review and editing, K.P., A.S., O.K., G.M., A.A., Ermioni Papageorgiou (E.P.), M.C., G.P., Eftihia Polyzogopoulou (E.P.), S.S., E.B., E.S., S.G. and G.T.; visualization, G.T.; supervision, K.P. and G.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Ethical review and approval were waived for this study as per current Greek law regarding retrospective studies of anonymized standard care data. No Internal Review Board approval and no written consent were required but patients were informed of their participation and offered the possibility to withdraw.

**Informed Consent Statement:** Informed consent was obtained regarding the use of imaging data.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

