When to Stop

Introduction

In the era in which endovascular treatment (EVT) is established as the standard of care in patients with acute ischemic stroke with large vessel occlusion,¹ many questions still arise regarding which factors are in our hand to improve patients outcomes.

Despite complete recanalization, a substantial proportion of patients do not achieve good clinical and functional outcome. The HERMES (Highly Effective Reperfusion Evaluated in Multiple Endovascular Stroke Trials) meta-analysis is the largest cohort described to date.² Overall, close to 80% of the EVT treated patients achieved recanalization (modified Thrombolysis in Cerebral Infarction [mTICI] 2B–3, considered as more than 50% or 100% of target downstream territory, respectively) and only 46% had a favorable functional at 3 months. Similar rates of good functional outcome were described in extended time window (up to 24 hours from onset) in patients with favorable imaging profile.^3,4

A recent publication describing final mTICI scores in HERMES trials and related outcomes showed that even among patients achieving successful recanalization (mTICI ≥2b), up to 51.1% did not achieve a favorable outcome.⁵ This may be due to an inadequate patient selection but also to factors associated with the procedure. No data about relationship between number of passes needed to achieve recanalization and outcome has been published.

Several studies have investigated the factors associated with functional outcome in patients who achieved recanalization after EVT; older age, proximal site of occlusion, and higher National Institutes of Health Stroke Scale (NIHSS) score were shown to be baseline independent predictors of poor outcome.^6–8 On the contrary, higher degree of recanalization has been repeatedly associated with better clinical outcomes.^9,10 Variations of the original TICI score, including the mTICI score, have been developed, adding mTICI 2c as a near total reperfusion (90%–99%)^11,12 and the expanded TICI score⁵ to improve reperfusion-grading precision after EVT.

Other studies have focused on the number of passes and its relation with clinical outcome. Better outcomes have been found among patients who achieved mTICI 3 in one pass as compared with those in which more passes were needed,¹³ and the detrimental effect of increasing number of device passes on recanalization and outcome has been shown.^14,15 Four passes have been described as the maximum number of passes that should be done: trying more attempts with respect to recanalization and functional outcome has been found futile.¹⁶ Analysis of the STRATIS (Systematic Evaluation of Patients Treated With Neurothrombectomy Devices for Acute Ischemic Stroke) registry has shown that balloon-guided catheter and use of longer stentretrievers are associated with better rates of first-pass recanalization.^17,18

However, the interaction between the final degree of recanalization and the number of passes needed to achieve it remains unknown. Several questions remain unanswered, which should be the minimum degree of recanalization achieved before stopping the procedure? Should we limit the maximum number of attempts if recanalization is not achieved?

We aim to try to give an answer to those questions and further investigate the interaction between number of passes and degree of recanalization and clinical outcome.

Methods

The data that supports the findings of this study are available from the corresponding author on reasonable request. Ethics approval was obtained from the local institutional review board, and written informed consent was obtained from patients.

Patient Selection

We performed a retrospective analysis of a prospectively maintained database that includes all patients undergoing EVT at our tertiary stroke center since 2012 (n=704). Patients were selected if they had an acute ischemic stroke secondary to an occlusion of terminal intracranial carotid artery or middle cerebral artery (MCA) portions M1 and M2 and Alberta Stroke Program Early CT Score (ASPECTS) of 5 or more (n=542). Patients with a baseline modified Rankin Scale score ≥3 (n=23), isolated extracranial ICA or ACA occlusions (n=32 and 6, respectively), posterior circulation strokes (n=61) or incomplete data related to baseline or imaging characteristic (n=40) were excluded. Main analysis was centered on patients who achieved successful recanalization after EVT, considered as mTICI score ≥2b (n=459). Flowchart with rate of complete recanalization (mTICI 2B–3) after every device pass and recanalization degree after finishing the procedure is showed in Figure 1.¹⁹

• Download figure
• Download PowerPoint

Clinical and Imaging Assessment

Baseline characteristics including age, sex, prestroke modified Rankin Scale, cardiovascular risk factors, previous antithrombotic and statin treatment, glycemia, and blood pressure were registered. Patient with no contraindications for intravenous thrombolytic treatment received IV-tPA (intravenous tissue-type plasminogen activator) after neuroimaging confirmed absence of hemorrhage or large ischemic lesion (>1/3 third of MCA territory).

Baseline imaging profile consisted in a cerebral noncontrast computed tomography (CT) and a CT-angiography. ASPECTs score and occlusion site were determined by consensus between a neurorradiologist and the vascular neurologist in charge. All patients were treated by neurointerventionalists with more than >5 years of experience in EVT. Number of passes, device used, use of distal aspiration or balloon-guided catheter, and workflow times were prospectively collected. Degree of recanalization using mTICI score was rated by consensus between the interventionalist and the vascular neurologist immediately after the procedure; for the present analyses, patients with final mTICI 2C were included in the mTICI 3 group.^11,19 Clinical outcome at 24 hours (NIHSS) and 90 days modified Rankin Scale (rated by a neurologist blinded to the acute phase) were also recorded.

A control CT scan was made 24 hours after treatment to assess the presence of ischemic infarct and hemorrhagic transformation. Symptomatic intracerebral hemorrhage was defined based on ECASS-II (European Cooperative Acute Stroke Study) criteria as any intracranial hemorrhage associated with neurological deterioration of ≥4 points on the NIHSS score at 24 hours.²⁰

Primary end points of the study were the association of degree of recanalization (mTICI 2B or 3), number of passes and its interaction with dramatical clinical recovery (DCR, defined as NIHSS score ≤2 or decrease in 8 NIHSS points in 24 hours), good functional outcome (defined modified Rankin Scale score ≤2 at 90 days), and mortality. Variables associated with first pass recanalization (FPR) and degree of recanalization were also studied. Functional outcome of patients who did not achieve recanalization was compared with patients who did.

Statistical Analysis

Descriptive analysis was used to define baseline patients’ characteristics, imaging, and procedure variables. Normality of distributions was assessed by using histograms and the Shapiro-Wilk test. Univariate comparisons were performed by Fisher exact test, Pearson χ², Spearman correlation test, or Mann-Whitney, as appropriate to the type of variable and its distribution, to find variables associated with DCR and good functional outcome. Continuous variables are displayed as mean and SD or median and interquartile range (IQR; if not normally distributed). Categorical variables are displayed by number and frequencies. Multivariate regression models were performed to evaluate association of different variables with DCR and good functional outcome. Variables with a statistical trend (P<0.1) in the univariate analysis or historically associated with primary end points were added to the analysis. Comparisons of the effect of number of passes for each recanalization degree were made. All analysis were performed using SPSS Statistics version 25 (IBM Corporation, Armonk, NY).

Results

From 542 patients with anterior circulation large vessel occlusion, 459 (84,5%) achieved a recanalization grade mTICI ≥2b and were included in the final analysis (Figure 1). Median age was 75 years (IQR, 64–82), 234 (51%) patients were women, and median baseline NIHSS score was 16 (IQR, 11–20). Median ASPECTS was 9 (IQR, 8–10). Intravenous thrombolytic treatment (rtPA) was administered in 214 (47%) patients. Vessel occlusion was terminal intracranial carotid artery in 91 (20%) patients, M1-MCA in 270 (59%) patients, and M2-MCAI in 98 (21%) patients. Tandem intraextracranial (ECA) occlusion was present in 86 (19%) patients, and 46 (10%) of them received acute extracranial carotid stenting. In 435 procedures, (95%) a stent retriever was used (being Solitaire FR the most frequently used n=355, 77%) in combination with a distal aspiration catheter in 290 procedures (63%). First-pass aspiration approach alone (ADAPT) was used in 24 patients (5%). General anesthesia was used in 21 patients (5%). Table 1 shows baseline demographic, imaging, procedural, and outcome variables depending on degree of recanalization and first-pass recanalization.

Recanalization grade mTICI 2B was achieved in 230 patients (51%) while mTICI 3 in 229 patients (49%). FPR occurred in 213 procedures (46%), including FPR-mTICI 3 in 135 (29%). The rate of FPR progressively decreased as the initial location of the occlusion was more proximal (Figure 2), being an independent predictor of FPR (terminal intracranial carotid artery versus MCA-M1 versus MCA-M2, odds ratio [OR] 0.68; 95% CI, 0.49–0.94; P=0.02). There was an association with the need of multiple passes and general anesthesia (FPR: 1.9% versus more than one pass: 6.9%; P<0.05), probably related to more difficult procedures in those patients who needed general anesthesia versus conscious sedation, since in our center we use conscious sedation as a first approach unless low level of consciousness or respiratory failure. Lack of distal aspiration system and higher ASPECTS were associated in the univariate analysis with FPR but did not achieve significance in the multivariate analysis. Patients who needed multiple passes to achieve recanalization had more hemorrhagic transformation compared with patients with first-pass recanalization (31/213 versus 54/243; P=0.01), but no differences were found for symptomatic hemorrhagic transformation or vessel dissection (Table 1).

• Download figure
• Download PowerPoint

Complete mTICI 3 recanalization occurred more frequently with less device passes, achieved after first pass in 135 patients (63%) and after multiple passes in 94 patients (37%). In a logistic regression adjusting with other possible cofounders (site of occlusion, previous treatment with intravenous thrombolysis, tandem or terminal intracranial carotid artery occlusion, use of distal aspiration or balloon-guided catheter, and atrial fibrillation), FPR was the strongest predictor of mTICI 3 after EVT (OR, 2.7; 95% CI, 1.8–4; P<0.005), along with the absence of tandem occlusions (OR, 2.1; 95% CI, 1.2–3.5; P<0.01) and history of atrial fibrillation (OR, 1.66; 95% CI, 1.09–2.52; P=0.018).

Good functional outcome was achieved in 237 patients (51%), and DCR occurred in 225 patients (49%). Mortality at 90 days was 16%. Variables associated (P<0.1) with good functional outcome and DCR in univariate analysis are shown in Table 2. In a logistic regression model, both FPR (OR, 2.5; 95% CI, 1.4–4.4; P<0.005) and mTICI 3 (OR, 2.63; 95% CI, 1.47–4.76; P<0.005) were the strongest treatment-related predictors of good functional outcome. Other independent predictors of functional outcome were age (OR, 0.94; 95% CI, 0.92–0.97; P<0.005), ASPECTS (OR, 1.25; 95% CI, 1.01–1.56; P=0.037), NIHSS (OR, 0.91; 95% CI, 0.85–0.94; P<0.005), intravenous thrombolysis (OR, 1.87; 95% CI, 1.06–3.3; P=0.029), and time to recanalization (OR, 0.997; 95% CI, 0.990–0.999; P=0.005). FPR (OR, 1.6; 95% CI, 1.02–2.6; P=0.042) and mTICI 3 (OR, 1.79; 95% CI, 2.81–1.14; P=0.012) were also found to be independent predictors of DCR.

Complete recanalization (mTICI 2B or 3) rate declined after each pass (Figure 1, P<0.001). Moreover, patients with complete recanalization had worse outcomes for each device pass needed to achieve complete recanalization (P<0.01, Figure 3). Achieving recanalization after 4 passes or less improved outcome compared with patients who did not achieve recanalization (38.5% versus 16.9%; P=0,023). For 5 or more passes, there was a numerical difference, but it was not statistically significant (25.6% versus 16.9%; P=0.176). While the rate of good outcome begins to decline after the first pass for patients with final mTICI 2B, for patients with final mTICI 3 this progressive decline was only seen after the third pass (Figure 4).

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Discussion

Recent studies linked FPR and degree of recanalization with better clinical outcomes.^9,10,13,14 To better understand the possible mechanisms of this association, we performed an analysis on the impact of the total number of passes according to final reperfusion grade. We found that both, FPR and final mTICI 3 were independent predictors of good outcome. Moreover, in patients who achieved mTICI 3, we found a progressive decline in the rate of good outcome only after the third pass. However, in patients who achieved a final mTICI 2B the negative impact on outcome started already with the second pass. This observation may indicate that the detrimental effect of each additional pass may be counterbalanced by the benefits of a complete mTICI 3 recanalization for up to 3 passes; beyond that number of attempts, outcome is affected despite full recanalization. On the contrary, for patients not achieving a mTICI 3 recanalization the impact of additional passes would negatively influence outcome starting from the second attempt.

Number of passes has been one of latest procedural factors associated with outcome. Zaidat et al¹³ described the first pass effect (one pass mTICI 3) as a strong predictor of good clinical outcome as compared with all other treated patients (including those who did not recanalize). To better characterize the benefits of the one-pass effect, we performed an analysis including only those patients who achieved a substantial (mTICI 2B–3) recanalization. Patients achieving a mTICI 3 were twice more likely to achieve a favorable outcome than those who achieved a final mTICI 2B. The highly beneficial effect of full recanalization was also seen for the first pass effect; patients who only needed one pass to achieving a mTICI 2B or 3 were also twice more likely to reach a favorable outcome than those who needed 2 or more passes to achieve the same degree of recanalization.

The theoretical mechanisms by which additional passes could induce worse outcome could be increased clot fragmentation with distal embolization or accumulated endothelial damage.²¹ Our results raise the question about the convenience to pursue additional passes when a substantial recanalization (mTICI 2B) is achieved after first pass (better is the enemy of good). The theoretical benefits of complete recanalization (mTICI 3) may be counterbalanced by the detrimental effect of multiple passes. Future clinical trials should address this question.

On the contrary, in our cohort, even if recanalization (mTICI 2b–3) is achieved after a third pass, more than 40% of the patients had good functional outcome. Moreover, recanalization improved prognosis until up to 4 passes. Patients who achieved recanalization after 5 passes or more had numerically better outcome but did not achieve statistical significance; we hypothesize that due to the low number of patients. The present analysis supports the hypothesis that additional attempts to achieve recanalization should be pursued at least until 4 passes.

Several clinical features, such as initial clot location or stroke cause, were associated with FPR in our study. Distal aspiration was associated with less rate of FPR but did not achieve significance in multivariate analysis. Comparison between patients who were treated with stentretrievers or aspiration alone could not be performed due to low number of patients treated with aspiration devices. Further analyses centered on device type, ancillary catheters, or presumed clot composition should be performed in the future to individualize treatment strategies and determine the best approach to maximize the probabilities of FPR in each patient.^21,22 Future thrombectomy devices should also be developed with the aim to increase first pass success.

Our study has several limitations; it is a retrospective single-center analysis, and adjudication of final mTICI was not blinded. Previous studies have shown that there is an overestimation in the degree of recanalization by operators as compared with core laboratory assessment.²³ mTICI 2C degree of recanalization was only recently introduced in our registries; and therefore, no fine tune assessment with the expanded TICI scale could be performed. In our center, stentrievers are considered as the preferred initial technique (in 95% of cases) by neurointerventionists; and therefore, the results should not be generalized to all thrombectomy techniques such as aspiration alone. Baseline perfusion imaging was not analyzed for the present study.

Conclusions

In patients with anterior circulation large vessel occlusion, FPR and complete recanalization (mTICI 3) are independent predictors of favorable outcome. Additional passes despite a substantial recanalization (mTICI 2B or more) may negatively affect outcome and should be considered with caution. For patients, without substantial recanalization at least 4 passes should be attempted. Above this number, recanalization may improve outcome, but future studies should address this question.

Acknowledgments

Drs García-Tornel, Ribo, Rubiera, and Molina participated in the conception and design of the study. Drs García-Tornel, Ribo, Rubiera, Requena, and Deck analyzed and interpreted the data. Drs García-Tornel, Muchada, Juega, Pagola, Rodriguez-Luna, Rodríguez-Villatoro, Rubiera, Requena, Boned, Olivé-Gadea, Ribo, Tomasello, and Hernández treated and included the patients in the study. Drs García-Tornel and Ribo performed statistical analysis. Dr García-Tornel wrote the article. All authors reviewed and approved the article.

Footnotes