1. Efficacy and safety of direct aspiration first pass technique versus stent-retriever thrombectomy in acute basilar artery occlusion—a retrospective single center experience
1.1. Introduction
1.1.1. Acute basilar artery occlusion (BAO) accounts for ap- proximately 10% of all intracranial large vessel ischemic strokes.
1.1.1.1. Successful recanalization of acute BAO prevents from poor clinical outcome and death [1–3]
1.1.2. Despite ad- vances in the treatment of BAO, 70% of the patients re- main disabled or die [4]. A meta-analysis of recent posi- tive randomized stroke trials [5] corroborated the benefi- cial effect of endovascular therapy (EVT) for ischemic stroke due to anterior circulation (AC) major artery occlu- sion on patient outcome.
1.1.3. In acute BAO, it is still unclear whether EVT is superior to standard therapy [1, 4].
1.1.4. Following recent positive EVT trials, stent retrievers emerged as the stan- dard of care for the treatment of stroke due to large artery
1.1.4.1. occlusions in the AC [6, 8].
1.1.5. However, the trials were not designed to demonstrate the superiority of one technique over the other [8]. In vitro studies have shown that stent- retriever thrombectomy (SRT) under continuous distal as- piration and primary aspiration thrombectomy (AT) led to comparable degrees of recanalization [9].
1.1.6. We aimed to compare the efficacy and safety of stent- retriever thrombectomy to first pass aspiration thrombectomy for EVT in patients with BAO.
1.2. Methods
1.2.1. Patients
1.2.1.1. We searched our thrombectomy database to identify all patients with acute BAO who underwent EVT between January 2013 and April 2016
1.2.1.2. We collected the following patient data: age and sex, cerebrovascular risk factors (ar- terial hypertension, diabetes mellitus, atrial fibrillation, dyslipidemia, prior stroke, prior myocardial infarction), National Institutes of Health Stroke Scale (NIHSS) score on admission, time of symptom onset, baseline and follow-up imaging (FU-imaging) modality (CT/CTA or MRI/MRA), etiology group of vascular occlusion accord- ing to the New England Medical Center Posterior Circulation Registry (NEMC-PCR) [14], and if intrave- nous thrombolysis (IVT) was given.
1.2.1.3. Patients fulfilling the following criteria were included in the study: (1) presence of a BAO or an occlusion of the intra- cranialvertebralartery(ICVA,V4-segment)leadingtoaBAO on digital subtraction angiography (DSA) and (2) EVT with SRT or AT. Additional extracranial EVT, e.g., percutaneous transluminal angioplasty (PTA) or stenting for vertebral artery (VA) origin stenosis and IVT prior to EVT, were no exclusion criteria. However, we excluded patients in whom additional intracranial angioplasty or a permanent stent were necessary to maintain the recanalization result attained by thrombectomy.
1.2.2. Image analysis
1.2.2.1. We assessed the following imaging parameters: infarct ex- tent using the posterior circulation Acute Stroke Prognosis EarlyCTscore (pc-ASPECTS)appliedtoCTA-sourceimages (at baseline), non-contrast CT (follow-up), or MRI diffusion weightedimages(baselineandfollow-up)dependingonavail- ability. On FU-imaging, we rated hemorrhagic complications according to the Heidelberg bleeding classification [15]
1.2.2.2. We determined the target arterial occlusion (intracranial vertebral artery (ICVA), proximal, middle and distal basilar artery (BA), and posterior cerebral artery (PCA; P1-segment only)) and extent of the thrombus (posterior circulation clot burden score, see below). Basilar artery subdivision followed the suggestion of Archer et al. [16]: the proximal BA extends from the junction of the two VA to the origin of the anterior inferior cerebellar artery (AICA), the middle BA from the AICA origin to the origin of the superior cerebellar artery (SCA), and the distal BA starts above the SCA and ends at the PCA.
1.2.2.3. We rated the thrombus load with a 10-point posterior cir- culation clot burden score (pc-CBS) built in analogy to the CBS [17]. The pc-CBS allots the intracranial segments of the posterior circulation (PC) arteries 10 points for the pres- enceofcontrastopacificationinCTA orflow-signal onMRA.
1.2.2.4. opacification (CTA) or flow-signal (MRA) in the V4- segments of the VA and the P1-segments of the PCA, and two points each for the proximal, middle, and distal segments of the BA. Partial filling defects suggesting stenosis or non- occlusive thrombus are rated as patent. Aplastic segments of the distalVA and the P1-segmentswillbescoredasabnormal. Apc-CBSof10indicatesabsenceofanocclusion;ascoreof0 indicates occlusion of all major intracranial PC arteries
1.2.3. Endovascular treatment
1.2.3.1. For EVT, we used stent retrievers aided by a distal aspiration catheter at the thrombus and primary AT, also referred to as a direct aspiration first pass technique for stroke thrombectomy (ADAPT) [18]. The treating interventionist chose the se- quence and method of the thrombectomy. Five intervention- ists with 3, 5, 6, 11, and 15 years of practice in neurointerventions treated the patients.
1.2.3.2. The following details were documented: femoral artery puncture time (start of treatment), the most proximal intracra- nial arterial occlusion (the target occlusion [19]), device (stent retriever or aspiration catheter type and size), number of retrieval/aspiration passes, time of recanalization, local recan- alization result according to the arterial occlusive lesion
1.2.3.3. (AOL) scale [19, 20], and procedural complications. We cal- culated the difference (AOLdiff) between the occlusion grad- ing before (AOLpre) and after the intervention (AOLpost) to document treatment effect. Successful recanalization, defined asanAOLdiffgrade of2 or3,was the variableofinterest. SRT or ATwas considered the respective first-line strategy if either was attempted first. We noted any derivation from these tech- niques and whether it was necessary to convert from one to another. Functional
1.2.4. Functional outcome and safety
1.2.4.1. Functionalneurologicaloutcomewasassessedusingthemod- ified RankinScale(mRS)score atdischargefromthehospital. Favorable outcome was defined as mRS scores of ≤3 [1]. Neurologicaloutcome atdischargewasthevariableofinterest regarding clinical efficacy. The type and grade of intracranial hemorrhage [15] and the procedural complications were re- corded as safety variables
1.3. Results
1.3.1. Patients
1.3.1.1. In total, 61 patients (39 (64%) men; median age 65 years, range 25–85 years) with PC stroke were eligible for EVT in ourdepartmentbetweenJanuary2013andApril2016.
1.3.1.2. Forthe purpose of the study, we excluded all patients who had recanalized as demonstrated by the first angiographic series (n = 1), in whom we could not access the target arterial occlu- sion (n = 6), who only received intra-arterial rtPA (n = 2) or PTA (n = 1), or who were treated with AT (n = 10) or SRT (n=8)andanyadditionalintracranialtreatment(supplemental table 1)
1.3.1.3. Finally, we included33patientsinthe study. Thirteen patients were treated with stent-retriever thrombectomy (SRT group) and 20 with aspiration thrombectomy (AT group).
1.3.1.4. Prior to EVT, 23 (70%) patients received IVT, 12 (60%) in the AT, and 11 (85%) in the SRT group (no significant group difference).
1.3.1.5. Baseline characteristics in the two groups differed only regarding the incidence of diabetes mellitus, with more diabetic patients in the AT group (Table 1).
1.3.2. Baseline imaging and target arterial occlusion
1.3.2.1. Before EVT, 32 (96%) patients received a non-contrast CT and CTA and one patient received a MRI and MRA. In two patients, the external CTA was not available at the time of study related image assessment. Baseline imaging parameters (Table 1) did not significantly differ between groups.
1.3.2.2. According to the DSA, the target occlusion was the intra- cranial vertebral artery in 2 patients, the proximal BA in 5 patients, the middle BA in 20 patients, and the distal BA in 6 patients (Table 1). CTA showed a corresponding occlusion in the respective vessel segments in all available images and revealed a patent distal BA-segment in only 2 out of 20 pa- tients with middle BA occlusion proven by DSA.
1.3.3. Endovascular treatment
1.3.3.1. We treated 13 patients with primary SRT
1.3.3.1.1. combined with distal aspiration at the proximal end of the stent retriever via a 4.1F aspiration catheter (041 Reperfusion Catheter, Penumbra Inc., Alameda, CA, USA) in 12 patients and with a 6F-guiding catheter in the V2- segment of the VA in one patient (Table 2).
1.3.3.2. In the AT group, wetreated20arterialocclusionswithfirstpassATusingeither the 041 Reperfusion Catheter withan inner lumen of0.041 in.
1.3.3.2.1. n = 10),a5.4Faspirationcatheterwitha distalinner lumenof 0.060 in. (5MAX ACE Reperfusion Catheter, Penumbra Inc.) (n = 9), or both catheters sequentially (n = 1). Mean time from onset of symptoms to groin puncture (time to treatment, TTT) was 334 min (95% CI 276–391 min) (Table 1). The time of symptom onset was unknown in one patient. TTT did not differ between treatment groups
1.3.4. Recanalization
1.3.4.1. The duration of the intervention
1.3.4.1.1. differed significantly between groups (p = 0.002) as follows: 97 min in the SRT group (95% CI 69–124 min) versus 55 min in the AT group (95% CI 43–66 min) (Table 2)
1.3.4.2. Mean time from symptom onset to recanalization (time to recanalization, TTR) was 407 min (95% CI 348–465 min) (Table 2). TTR did not differ between treatment groups.
1.3.4.2.1. Successful recanalization (AOL 2 and 3) was achieved in 9 (69%) patients treated with SRT and in 17
1.3.4.3. (85%) patients treated with AT (Table 2) but did not signifi- cantly differ between groups.
1.3.4.4. However, complete recanaliza- tion(AOL3)wasachievedmoreoftenwithAT(75%(95%CI 65–85%)) compared to SRT (46% (95% CI 32–60%)).
1.3.4.5. IntwopatientsoftheATgroup,wehadtoconvertfromone device to another. In one patient, AT with a 4.1F catheter failed. After replacing it with a 5MAX ACE, we achieved full recanalization (AOL 3). In another case of unsuccessful AT (4.1F catheter), we changed to SRTwhich was equally unsuc- cessful (AOL 0).
1.3.5. Periprocedural complications
1.3.5.1. As periprocedural complications, we noted a dissection of the VA without hemodynamic impairment in three patients (all SRT group) and an embolization to previously unaffected vessel territories in five patients (three in SRT group, two in AT group).
1.4. Clinical and imaging outcome
1.4.1. A favorable outcome (mRS ≤3) at discharge was observed in 10 patients (30%)
1.4.1.1. For FU-imaging, 14 patients had a non-contrast CTand 19 patientsreceivedaMRI.Mediantime tofirstFU-imagingwas 19 h after the end of the intervention (IQR 14 1/2 h). Final infarct sizes are presented in Table 2. No patient with a pc- ASPECTSof7–10onFU-imagingdieduntildischarge, while all patients who died during their hospital stay showed a pc- ASPECTS of 0–6 on FU-imaging.
1.4.1.2. The proportion of patients with a favorable outcomewashigherinthe ATgroup (notsignificant;Table2). Mortality rate was 24% (eight patients; four deaths per group: 31% in SRT group; 20% in AT group; not significant).
1.4.2. Intracranial hemorrhagic complications occurred in 12 (36%) patients (Table 2).
1.4.3. Two of them had a combination of subarachnoid hemorrhage (SAH) and either hemorrhagic in- farction type 1 (HI1) or parenchymatous hematoma type 1 (PH1). All three patients with a SAH died.
1.4.4. HI1 survived but the one patient with additional SAH. Three offour patients with HI2 deceased as did the patient with PH1 (and simultaneous SAH). Hemorrhagic complications did not differ between groups.
1.5. Discussion
1.5.1. ThesuperiorityofEVTinBAOcouldnotbeshown[1]and isunder further study [4].
1.5.2. However, inthe anteriorcirculation, EVT for ischemic stroke due to large artery occlusion is prov- en in eligible patients (class I grade of recommendation, level of evidence A) [6].
1.5.3. patients (class I grade of recommendation, level of evidence A) [6]. And stent retrievers should be used for thrombectomy (class I, level A). Converging results in effica- cy and safety of SRT and AT makes it reasonable to consider devices other than stent retrievers in selected cases (class IIb, level B-NR) [6]
1.5.4. However, there is no agreement as to which device or technique is best and recent trials were not designed to answer this question [8]. Non-randomized observations re- port non-inferiority of AT compared to SRT. The observed advantages of AT are faster procedures, higher rates of com- plete recanalization, and equal or better outcomes [10, 11, 18, 21].
1.5.5. Two studies compared SRTand AT in the BAO and found comparable procedural and clinical outcomes [12, 13]. One study reported faster procedures and more complete recanali- zations with AT, as reproduced here. However, neither distal aspiration along with a stent retriever nor last generation large bore aspiration catheters were used [12]. The other study did not report procedural details [13].
1.5.6. In vitro studies suggest a con- ceptual advantage when using stent retrievers in combination with distal aspiration [22]
1.5.7. In vivo, this combination shortens the thrombectomy procedure in acute BAO [23]. Anatomically, this technique makes sense in the PC, as a guiding catheter in one VA may only reverse the flow in that arterywithoutaffectingthebasilarthrombusduetocompeting flowfromtheother VA[24].
1.5.8. With distalaspiration,the throm- bus captured in the stent and the aspiration catheter may be removed under controlled suction. Moreover, modern large bore aspiration catheters enable us to even attempt recanaliza- tion without any other device, in both the AC [10, 18, 21] and PC [12].
1.5.9. We had to convert from AT to SRT in only one patient (3%). Typical conversion rates are 44–22% in mainly AC strokes not differentiating between the carotid and vertebrobasilar circulation [21, 25].Reported conversion rates in PC stroke are 0% [12] and 48% [13].
1.5.10. However, if needed, the transition to SRT is easy in first-line ATas the aspiration catheter is already close to the occlusion. Another advantage of AT is its proximal mode of action. Aspiration allows to safely removing the clot without the ne- cessity to first penetrate a diseased vessel segment, as in PC more often than in the AC, there can be an occlusion due to local arteriosclerosis, predominantly in the proximal BA- segments [14].
1.5.11. Occlusions involving the distal BA-segment are mostly embolic
1.5.12. as compared to occlusions in proximal BA-segments [2, 26]
1.5.13. As we excluded patients requiring intracranial EVT other than SRT or AT, a majority of the study patients (80%) suffered from embolic occlusion involving the mid to distal BA (Table 1).
1.5.14. In our study, a rate of 79% AOL 2/3 recanalizations is contrasted by a rate of moderate outcome (mRS ≤3) of 30% at discharge. The in-hospital mortality was 24%.
1.5.14.1. Intracranial hemorrhagic complications occurred in 12 (36%) patients and were associated with six deaths
1.5.15. Two registries of acute BAO treatment [1, 27] reported similar recanalization rates of 72 and 79%, moderate outcome (mRS ≤3) in 32% (at 1 month) and 42% (after 90 days), a mortality rate of 36 and 35%, and hemorrhages in 14 and 6%, respectively. A meta-analysis of SRTstudiesinacuteBAO(15studies,312patients)reporteda 90-day favorable outcome (mRS ≤2) in 42% (95% CI 36–48%), recanalization in 81% (95% CI 73–87%), 30% mortal- ity (95% CI 25–36%), and symptomatic hemorrhages in only 4% (95% CI 2–8%) [26]. Whether this increase in good out- come along with a reduction in hemorrhage truly heralds a
1.5.16. In a recent study, the AOL scale reli- ably classified BAO recanalization, while the mTICI score failed to achieve substantial inter-observer agreement [20].
1.5.17. Consequently, we used the AOL scale as it addresses device efficiencyatthe occlusion [19].Contrarytothe AC[28],there is no validated dichotomization cut-off for the recanalization result in the PC to predict good outcome.
1.5.18. In accordance, the risk of hemorrhagic infarction in stroke due to BAO was described as largely determined by the extension of ischemic changes on admission CT [29, 30]. Of note, there is no classification of bleeding events after ischemic stroke which considers the specifics of the PC [26].
1.6. Untitled
1.6.1. Untitled