True first-pass effect in basilar artery occlusions: first-pass complete reperfusion improves clinical outcome in stroke thrombectomy patients

Background Complete reperfusion (mTICI 3) in anterior circulation ischemic stroke patients after a single mechanical thrombectomy (MT) pass has been identified as a predictor of favorable outcome (modified Rankin Score 0-2) and defined as true first-pass effect recently. This effect has not yet been demonstrated in posterior circulation ischemic stroke. We hypothesized a true first-pass effect for the subgroup of acute basilar artery occlusions (BAO). Methods Consecutive patients with acute thromboembolic occlusions in the posterior circulation, treated between 2010 and 2017, were screened and all BAO patients with complete angiographic reperfusion and known symptom onset included for unmatched and matched analysis after adjustment for multiple confounding factors (demographics, time intervals, stroke severity, posterior circulation Alberta Stroke Program early computed tomography Score and comorbidity. The primary objective was outcome at 90 days between matched cohorts of single pass vs. multi pass complete reperfusion patients. Results 90 MTs in BAO were analyzed, yielding 56 patients with known symptom onset, in whom we achieved complete reperfusion (mTICI 3), depending on whether complete reperfusion was achieved after a single pass (n=28) or multiple passes (n=28). Multivariable analysis of 56 non-matched patients revealed a significant association between first-pass complete reperfusion and favorable outcome (p<0.01). In matched cohorts (n=7 vs. n=7), favorable outcome was only seen if complete reperfusion was achieved after a single pass (86% vs. 0%). Conclusion Single pass complete reperfusion in acute basilar artery occlusion is an independent predictor of favorable outcome. Achieving complete reperfusion after multiple passes might impair favorable patient recovery.

Background Complete reperfusion (mTICI 3) in anterior circulation ischemic stroke patients after a single mechanical thrombectomy (MT) pass has been identified as a predictor of favorable outcome (modified Rankin Score 0-2) and defined as true first-pass effect recently. This effect has not yet been demonstrated in posterior circulation ischemic stroke. We hypothesized a true first-pass effect for the subgroup of acute basilar artery occlusions (BAO). Results 90 MTs in BAO were analyzed, yielding 56 patients with known symptom onset, in whom we achieved complete reperfusion (mTICI 3), depending on whether complete reperfusion was achieved after a single pass (n=28) or multiple passes (n=28). Multivariable analysis of 56 non-matched patients revealed a significant association between first-pass complete reperfusion and favorable outcome (p<0.01). In matched cohorts (n=7 vs. n=7), favorable outcome was only seen if complete reperfusion was achieved after a single pass (86% vs. 0%).

Methods
Conclusion Single pass complete reperfusion in acute basilar artery occlusion is an independent predictor of favorable outcome. Achieving complete reperfusion after multiple passes might impair favorable patient recovery.   [18]: Complete reperfusion with the first stent-retriever-based thrombectomy pass resulted significantly more often in favorable outcome at 90 days compared to complete reperfusion after multiple thrombectomy passes in otherwise matched cohorts. This implies that phenomena like microembolizations into the distal vascular territories of the treated proximal vessel, which have been observed in-vitro [19] might be linked to the number of thrombus retrieval attempts in-vivo, impacting functional outcome independent of the elapsed time since symptom onset and the status of collateral flow.
Such a true first-pass effect has not been demonstrated yet in basilar artery occlusions (BAO). We addressed this issue by conducting a study in which we compared the clinical outcome of patients in whom we achieved complete reperfusion (mTICI 3) after a single pass with a matched cohort of patients, in whom we achieved complete reperfusion after ≥ 2 passes.

PATIENT SELECTION
All consecutive patients with AIS in the posterior circulation admitted to our hospital from October 2010 to December 2017 with imaging proven vertebrobasilar LVO were evaluated retrospectively on an intention-to-treat basis. We selected all consecutive patients with thromboembolic BAO, excluding patients where procedural or periprocedural imaging was highly suggestive of an underlying stenosis as etiology of the acute occlusion. Inconclusive cases were also excluded. We collected demographics, pre-existing cardiovascular risk factors and cerebrovascular events and comorbidities.
The National Institutes of Health Stroke Scale (NIHSS) and modified Rankin scale (mRS) scores were assessed by a stroke neurologist.
Our standardized stroke imaging protocol comprises non-enhanced cranial computed tomography (NECT), CT angiography (CTA) and CT perfusion (CTP). All eligible patients receive IVT according to national guidelines. In case of posterior circulation LVO, e.g. BAO, patients are considered for MT regardless of the time window within 24 hours from symptom onset. In such patients with extended time-window and without contraindications for MR-imaging we aim at performing a concise MR stroke imaging protocol before deciding for MT. We also include patients with low posterior circulation Alberta Stroke Program early computed tomography Score (PC ASPECTS) if imaging implies that there is tissue at risk (so called "mismatch concept") due to the fatal prognosis of untreated basilar artery thrombosis [20]. Patients are excluded when imaging reveals intracranial hemorrhage (ICH).
Postinterventional control imaging is routinely performed within 24 h after MT or immediately in case of a new neurological deficit in order to determine the extent of stroke and to exclude postinterventional ICH. Symptomatic ICH (sICH) is defined as any apparent extravascular blood in the brain or within the cranium that is associated with neurological deterioration defined by an increase of ≥ 4 points in the NIHSS score [21].

POSTTREATMENT ANALYSIS
To assess complete reperfusion rates, a reader (board-certified neuroradiologist), who was blinded to clinical and procedural data, re-evaluated all corresponding angiographic images of our prospectively captured database, in which reperfusion was reported as being mTICI 3 or mTICI 2b [24]. Patients with confirmed mTICI 2b or near complete reperfusion (e.g. reperfused BAO with a residual small occlusion of the distal posterior cerebral artery (PCA)) were excluded from analysis.

STATISTICS
We used IQR and Mann-Whitney U tests for data comparison, after testing data distribution for normality with a Shapiro Wilk test. Multivariable analysis was performed with a stepwise logistic regression test indicating odds ratios (OR) and confidence intervals (CI). All indicated CI are 95% confidence intervals. Statistical significance was defined as p ≤ 0.05. All statistical analyses were performed with MedCalc V. 19.04 (MedCalc Software, Ostende, Belgium).

Discussion
Several variables are known to influence the 90 days outcome in AIS. Part of them has been investigated in the pre-thrombectomy era, others after establishing endovascular treatment as the therapeutic mainstay in LVO of the cerebral vasculature. Nowadays "recanalization" (e.g. graded by the Arterial Occlusive Lesion (AOL) scale) is sometimes used interchangeably with the term "reperfusion" (most commonly graded by the TICI scale) in the literature, which involuntarily leads to a newer fundamental principle, which became more evident after the initial futile endovascular treatment approaches based on thrombus separation followed by aspiration without proximal flow control [28]: Creating a perfect local recanalization (AOL 3) does not necessarily lead to a perfect reperfusion of the corresponding downstream territory. In everyday endovascular stroke treating practice the average portion of patients with an mTICI 2b result ranges between 59% (self-reported) and 67.5% (core-lab controlled) in large registries [29,30]. It remains largely unclear how many of these mTICI 2b results are due to thrombus fragmentation and therefore essentially iatrogenic, either constitutes an independent predictor for favorable outcome at 90 days, with 2-3 times higher odds compared to complete reperfusion after multiple passes [18]. Consequently, they coined the term "true first pass effect". Interestingly, with the time interval from symptom onset to final reperfusion being matched and thus taken out of the equation, the true first pass effect seems to have a prognostic value in itself, at least for the anterior cerebral circulation.
The posterior cerebral circulation might be a different story. So far, the evidence on how to optimally treat e.g. BAO is inferior in comparison to the evidence in anterior circulation LVO [33-36], most probably due to the fact that the recently published randomized studies excluded them. This is reflected in the cautious wording of the current guidelines [37]. In our study all outcome analyses showed that there was a significant association between first-pass complete reperfusion of BAO and favorable outcome. As anticipated, procedural time from puncture to reperfusion was always significantly shorter if only one pass was needed for complete reperfusion, as this time span is evidently linked to the number of passes. In accordance with the results of Zaidat et al. in regard to our unmatched analysis results and Nikoubashman et al. for the matched cohorts we found preliminary evidence that a true first-pass effect might also exist in BAO [17,18]. Nonetheless, it cannot be completely ruled out that the results of our non-matched cohort as well as the results of observed effects impacting the outcome not only in the anterior circulation, but also comparably in BAO is intriguing and warrants further studies [38]. The delicate anatomy of the basilar artery with the angiographically barely visible yet clinically paramount pontine and mesencephalic perforators might be especially prone to mechanical thrombus manipulation by various techniques and repeated extraction maneuvers as we are currently able to reach comparable reperfusion rates but still not the same, relatively satisfactory rates of functional outcome we see in anterior circulation LVO [39]. The potential confounder of iatrogenic vessel perforation [40] was low and not significant between groups.
Finally, our study cannot address the question how to best achieve first-pass complete reperfusion.
The evident advantage of flow arrest by a balloon guide catheter in the anterior circulation [41] is difficult if not practically impossible to achieve in all but the rare circumstance of pre-existing nonacute unilateral vertebral artery occlusion. Thus, those endovascular techniques which combine the effectiveness of aspiration with stent-retrievers and yield higher first-pass complete reperfusion results without a mandatory flow arrest might currently offer the best option in BAO [15,16]. Further methodical studies, including possibilities to consistently achieve flow arrest in the posterior circulation are not only necessary but rather urgent, let alone a consistent proof of concept by proper randomization.

LIMITATIONS
This is a retrospective monocentric study design about consecutive BAO cases which were treated by MT based on an individual case by case clinical consensus process. The endovascular techniques were performed by the same team based on institutional principles, but varied nevertheless over time as the search for the best techniques is an ongoing process. We did not re-grade reperfusion results in a core lab, but relied instead on a blinded second reader to improve the validity of imaging interpretation.
Although our case-control design with matched patients has the highest degree of validity when randomization is lacking, the resulting sample size is very low, even though we started with 122 patients. Building on published preliminary results of anterior circulation LVO, we feel confident that our results could be interpreted as initial stimulus for more extensive research about the potentially decisive role of complete first-pass reperfusion in an otherwise devastating clinical condition of BAO.

Conclusion
We found preliminary evidence that first-pass complete reperfusion might constitute a decisive advantage in acute basilar artery occlusion. The choice of endovascular techniques should therefore focus on a successful first pass with a complete reperfusion of the dependent vascular territory without iatrogenic downstream embolization.

Declarations
Ethics approval and consent to participate -The study was approved by the local ethics committee