Article Text

Download PDFPDF

Original research
PROTECT: PRoximal balloon Occlusion TogEther with direCt Thrombus aspiration during stent retriever thrombectomy – evaluation of a double embolic protection approach in endovascular stroke treatment
  1. Christian Maegerlein1,
  2. Sebastian Mönch1,
  3. Tobias Boeckh-Behrens1,
  4. Manuel Lehm1,
  5. Dennis M Hedderich1,
  6. Maria Teresa Berndt1,
  7. Silke Wunderlich2,
  8. Claus Zimmer1,
  9. Johannes Kaesmacher1,3,4,
  10. Benjamin Friedrich1
  1. 1 Department of Diagnostic and Interventional Neuroradiology, Klinikum rechts der Isar, Technical University Munich, Munich, Germany
  2. 2 Department of Neurology, Klinikum rechts der Isar, Technical University Munich, Munich, Germany
  3. 3 Institute of Diagnostic and Interventional Neuroradiology, Inselspital, University of Bern, Bern, Switzerland
  4. 4 Department of Neurology, University of Bern, Inselspital, Bern, Switzerland
  1. Correspondence to Dr Christian Maegerlein, Department of Diagnostic and Interventional Neuroradiology, Klinikum rechts der Isar, Technical University Munich, 81675 Munich, Germany; christian.maegerlein{at}


Purpose Stent retriever-based mechanical thrombectomy (MT) for emergent large vessel occlusions (ELVO) is often complicated by thrombus fragmentation causing distal embolization and embolization to new vascular territories. Well-established embolic protection approaches include proximal flow arrest and distal aspiration techniques during stent retriever maneuvers. Aiming at the reduction of thrombus fragmentation during MT we evaluated a technical approach combining proximal balloon occlusion together with direct thrombus aspiration during MT: the PROTECT technique.

Methods We performed a case-control study comparing the PROTECT technique with sole distal aspiration during MT regarding technical and procedural parameters, n=200 patients with ELVO of either the terminus of the internal carotid artery or the proximal middle artery were included.

Results PROTECT resulted in a shorter procedure time (29 vs 40 min; P=0.002), in a higher rate of successful recanalizations (100% vs 78%; P=0.001) and a higher rate of complete reperfusions (70% vs 39%; P<0.001) compared with sole distal aspiration during MT.

Conclusion The PROTECT technique is a promising new approach to significantly reduce thrombus fragmentation and, hence distal embolization during MT. This safe and efficient technique needs to be validated in larger trials to confirm our results.

  • balloon
  • embolic
  • intervention
  • stroke

Statistics from


Since the international comprehensive implementation of endovascular procedures in stroke treatment due to the overwhelming success of state-of-the-art mechanical thrombectomy (MT) procedures,1–5 increasing effort was invested to technically improve the procedures aiming at complete reperfusions in a minimum of time. Facilitated by new technical innovations and improvements, new approaches and modifications were evaluated during the past months and years. Besides the development of latest-generation, large bore distal access catheters for direct thrombus aspiration,6–8 the concept of proximal flow arrest during the thrombectomy procedure using so-called balloon guide catheters (BGC) has increasingly come into focus in the interventional community.

Several studies proved BGC procedures to be safe and fast. The rates of first-pass successful recanalizations were higher and the rates of thrombus fragmentations and distal embolizations were lower compared with MT procedures without implementation of BGC.9–14 In any case, all former comparative clinical trials dealing with the use of BGC catheters in endovascular treatment of emergent large vessel occlusions (ELVO) of the anterior circulation suffer from one major drawback: the comparison groups treated with non-BGCs have great inherent methodological deficiencies. In these studies, the BGC was simply replaced by a normal guide catheter without the use of adjunctive embolic protection distally. Retrieving the device – loaded with clots – through all vascular curves until reaching the guide catheter without any flow modulation increases the risk of thrombus fragmentation, distal embolization and embolization to new, previously unaffected vascular territories. This is clearly represented by the quite low rate of successful recanalizations (60-70%) when performing stent retriever maneuvers without adjunctive embolic protection.  9–11 It is therefore not surprising that recanalization results in the BGC populations are far better than in the control groups.

We gained good experience over the past years with a different approach in MT using a distal aspiration approach in contrast to the proximal aspiration/flow arrest approach of BGCs. For this purpose, we are using a large bore distal aspiration catheter which is advanced as closely as possible to the thrombus before performing the stent retriever maneuver. During stent retriever retraction, aspiration is applied to the distal aspiration catheter to facilitate opening the occlusion site while avoiding distal embolizations or embolizations to new, previously uninvolved vascular territories. This technique has been described before and was sometimes entitled as ‘SOLUMBRA’.15–17 Using this technique, successful recanalizations of 85% to 93% are achievable.7 18 19

As an advancement of this distal aspiration technique, we here discuss a new technical modification that was first described recently in a case series by Stampfl et al.20 We are suggesting to entitle this technique ‘PROTECT’: PRoximal balloon Occlusion TogEther with direCt Thrombus aspiration. The key point of this modified technique is to additionally insert a large bore aspiration catheter into the BGC and apply both proximal aspiration/flow reversal and distal aspiration during the stent retriever maneuver. This approach became possible as large bore aspiration catheters of up to 6F were able to be inserted into the BGCs (figure 1).

Figure 1

Illustration of the Solumbra technique that was used in the control group (A, B) and the PROTECT technique (C, D) in the case of an occlusion of the middle cerebral artery. After deployment of the stent retriever, suction is applied to the aspiration catheter (*) during retraction of the stent retriever for embolus protection. As antegrade bloodstream is not completely suppressed by the 8F guide catheter (**), there is still a considerable risk of distal thrombus embolization. Using a balloon guide catheter (****) with the balloon inflated (***) during the stent retriever maneuver, the antegrade bloodstream is suppressed. In conjunction with distal aspiration at the large bore aspiration catheter the risk of distal embolization can be reduced effectively.

Our hypothesis is that this technical modification will help to further reduce thrombus fragmentations, finally achieving better technical and clinical results. In the present study, we compare the PROTECT technique to our standard approach with distal aspiration during MT. Although this approach has already been used in various centers, to the best of our knowledge, neither was the PROTECT technique systematically described nor was a BGC-based approach systematically compared with a distal aspiration technique which can be regarded as an alternative standard therapy.

Material and methods

For the present case-control study we performed a retrospective analysis of a prospectively collected database including all consecutive stroke patients treated by endovascular measures. We identified all consecutive patients treated with PROTECT from January 1, 2014 until July 31, 2017. A propensity score-matched analysis including the age and occlusion site with a ratio of 1:4 to patients treated by the SOLUMBRA technique during the same time period was performed.


Patients were eligible for endovascular thrombectomy if CT angiography (CTA) confirmed ELVO of the carotid terminus of the internal cerebral artery (ICA) or the proximal middle cerebral artery (MCA). Parenchymal infarction was limited to an ASPECTS>5. No age or perfusion selection was applied. Groin puncture needed to be achieved within 6 hours after symptom onset and neurological deficits were substantial (National Institute of Health Stroke Scale (NIHSS)>5). After establishing a femoral artery access, in the PROTECT group a BGC (9F Cello, ev3/8F FlowGate,2 Stryker) was introduced into the ICA. In the non-PROTECT group an 8F guide catheter (Vista bright tip, Cordis/Neuron MAX 088, Penumbra) was used respectively. For distal aspiration, a large bore aspiration catheter was introduced into the distal ICA or MCA (5 MAX, Penumbra/ACE 64, Penumbra/ACE 68, Penumbra/SOFIA, MicroVention/Catalyst 6, Stryker). Stent retriever maneuvers were performed in all cases (pREset, Phenox/Trevo XP ProVue, Stryker/Solitaire 2, Medtronic). Therefore, the device was introduced via a microcatheter (Trevo 18, Stryker) and deployed. The microcatheter was carefully removed afterward to increase aspiration applied by an electrical pump (Pump MAX, Penumbra) during the stent retriever retractions. Additional manual aspiration with a large syringe on either the guiding or BG-catheter was performed (figure 1).

Outcome analysis

The neurological deficits prior to the intervention and on the day of discharge were evaluated using the National Institute of Health Stroke Scale (NIHSS) and the modified Rankin Scale by experienced stroke neurologists. ‘Good neurological short-term outcome’ was defined as NIHSS at discharge <5. ‘Substantial neurological improvement’ was defined as either NIHSS at discharge ≤1 or difference between admission and discharge NIHSS ≥8. Reperfusion was graded using the oTICI scale21 by two experienced neuro-interventionalists in consensus (BF, CM). Successful reperfusion was defined as oTICI 2b/3.

Statistical analysis

We performed a propensity score-matched analysis comparing patients treated with PROTECT to patients treated by standard non-BGC at our center in a 1:4 ratio with age and occlusion site as co-variates with nearest neighbor-matching logistic regression analysis with a caliper of 0.2 (SPSS 24, IBM & R 3.3.3). Baseline variables were compared before and after matching to check for reduction of bias. Baseline, technical and outcome parameters between two groups were compared using Mann–Whitney U test or Fisher’s exact test depending on the type of variables analyzed. In order to adjust for potential confounders, a multivariate logistic regression model was applied to assess the independency of associations. The following prespecified factors were included in a model with either successful or complete reperfusion as dependent variable: age; time to groin puncture; pretreatment with tPA; and occlusion site. Results derived from logistic regression analysis are shown as adjusted odds ratio (aOR) and respective 95% confidence interval (95% CI).

Data are generally displayed as n (%) or median, if not indicated otherwise. Statistical significance was assumed at P<0.05.


Forty-six patients were treated with the intention to apply PROTECT. In six patients it was not possible to establish a stable access with the BGC, therefore 40 patients with complete PROTECT were included in the present analysis. These patients were matched in a 1:4 ratio with 160 patients treated with standard guiding catheters (see Materials and methods section). So a total of 200 patients was analyzed in the present study.

After propensity score matching no significant difference between the baseline parameters sex, age, initial NIHSS, usage of rtPA, occluded vessel or time from symptom onset to recanalization between the two treatment modalities could be detected (table 1).

Table 1

Patient characteristics

In a trend, more patients with ICA-terminus occlusions were treated with PROTECT (37% vs 25%) and the patients were slightly more severely affected by their stroke expressed by the NIHSS at admission (16 vs 15). None of those parameters reached statistical significance. Also, there was a trend toward a higher proportion of men treated by PROTECT, without statistical significance. Even though PROTECT might seem technically more challenging and thus potentially more time-consuming at first glance, using PROTECT resulted in a significantly shorter procedure time (29 vs 40 min; P=0.002). As the main finding, the use of PROTECT resulted in a significantly higher rate of successful recanalizations (100% vs 78%; P=0.001) (figure 2). This difference was mainly attributed to the change in the rate of oTICI 3 reperfusion (70% vs 39%; P<0.001) and the lack of insufficient reperfusions with an oTICI score of less than 2b by applying PROTECT (figure 2). To correct for a potential selection bias regarding the six patients in which PROTECT could not successfully be established, we performed an intention-to-treat analysis including all 46 patients in which a BGC was used. In this analysis the difference between PROTECT and the control group regarding technical success remained statistical significant (online supplementary table 1). After adjusting for potential confounders, PROTECT was independently associated with higher rates of complete (oTICI3) reperfusions (aOR 2.3, 95% CI 1.0 to 5.2, P=0.044). The point estimate for PROTECT did not converge for an analysis regarding successful reperfusion. Also by application of PROTECT, the number of maneuvers to achieve recanalization was significantly lower compared with the control group (1.9+/-1.2 vs 2.8+/-2.4; P=0.024) and there was a trend toward a higher rate of first-pass recanalization in the PROTECT group (48% vs 35%; P=0.16).

Supplementary file 1

Figure 2

Distribution of recanalization results comparing the PROTECT technique with the non-BGC distal aspiration in combination with stent retriever thrombectomy (P<0.001).

Aligning with the increase in reperfusion rates, the PROTECT group tended to have a better neurological outcome profile, without reaching statistical significance: good neurological short-term outcome was achieved in 40% using PROTECT versus 31% using simple guiding catheters. Substantial neurological improvement occurred in 55% by applying PROTECT vs 47% by distal aspiration only. Additionally, the ΔNIHSS, the NIHSS at discharge and the mRS at discharge were better in the PROTECT group (table 2). No significant difference in the rate of embolizations into previously unaffected territories could be identified.

Table 2

Procedure characteristics and outcome


After the first promising endovascular approaches with intra-arterial thrombolysis22 the breakthrough in endovascular stroke therapy was not achieved until the introduction of stent retrievers.1 3 5 Although retrieval of clots is highly effective using these devices, there is still a high risk of thrombus fragmentation causing either distal embolization or embolization to new territories if no additional embolic protection system is used.9–12 19 23 24

Well-established approaches for embolic protection during stent retriever maneuvers are, on the one hand, the implementation of proximal flow arrest using BGC and, on the other, the use of distal aspiration catheters.10 11 25 26 Notwithstanding the application of either technique, there is still a considerable risk of thrombus fragmentation during MT which demands further effort to diminish this procedural risk. This endeavor is justified as it could be shown that technical success, ideally resulting in complete reperfusion, is the most crucial modifiable determinant of clinical benefit.27–29

Based on the positive effects of the two aforementioned approaches for embolic protection during MT we are introducing the PROTECT technique that combines both methods to gain more effectivity during the interventional procedures, reducing the risk of thrombus fragmentation. The basis for this technique was the development of large bore BGC that can be loaded with, ideally, 6F aspiration catheters for effective distal aspiration. In the present study, we had significantly better technical results in the PROTECT-group, reaching 100% successful and 70% complete recanalizations as compared with 78% successful and 39% complete recanalizations in the control group with only distal aspiration. This reduction in insufficient reperfusion results with an oTICI score less than TICI 3 can be interpreted as a substantial reduction in distal embolization, compared with the non-BGC group. These results are superior to previously published results dealing with the ‘conventional’ BGC-technique that implicates only proximal flow arrest without additional distal aspiration during stent retriever maneuvers. In a recently published meta-analysis comparing 1083 procedures using BGC with 939 procedures without using BGC, successful recanalizations were reached in 79% (BGC) versus 67% (non-BGC) of the cases. Like in our study, procedure times in this meta-analysis were significantly shorter when using BGC. At first sight, this seems surprising as catheter preparation as well as periprocedural balloon inflation and deflation are further steps that theoretically may lead to increased procedure time. But these potential delays seem to be outweighed in the end by the lower number of maneuvers needed and the higher rate of successful first-pass recanalizations in the previously published BGC studies as well as in the presently introduced PROTECT technique. In contrast to the aforementioned studies we could not identify significant differences in the rate of embolizations into previously unaffected territories. The main reason for that finding might be the application of distal aspiration in the control group in contrast to the literature that mainly describes simple usage of a stent retriever. 

Regarding the unsatisfying technical and clinical results of exclusive stent retriever maneuvers without use of any embolic protection technique as applied in the control groups of the preceded studies dealing with the use of BG-catheters, this technique could be regarded as obsolete9–12 23 24 and should be omitted in the future. This ostensibly easier approach often turns out to be more complicated, as additional salvage maneuvers due to thrombus fragmentations and distal embolizations are required without the certainty of finally reaching a favorable result and additionally increasing the risks of periprocedural complications.

Our study is not without limitations, which need to be considered when interpreting the data. First, it is a retrospective observational study with methodically inherent restrictions. By applying a propensity-score matched analysis, we tried to correct for most of the methodical limitations resulting in a quite homogenous collective regarding the relevant baseline parameters. We used different types of stent retrievers and aspiration catheters at variable positions which should be mentioned as a potential bias. Nevertheless, all stent retrievers we used were latest-generation devices and no older devices like the merci device or the phenox clot retriever were used. In six patients with the intention to use PROTECT, conversion to a non-BGC approach with only distal aspiration was necessary. These conversions were all necessary due to extensive elongations of the vascular access and consecutive failure to establish a stable access with the BGC. This might implicate a certain selection bias, which we tried to correct for by performing an intention-to treat analysis with the same statistical result regarding successful and complete recanalization. Nevertheless, there were no cases of conversions due to procedural failures of the thrombectomy maneuvers itself in the PROTECT group. One possible lesson from those six cases might be to introduce a long sheath into the aortic arch or the common carotid artery to supply more support for the BGC instead of using a short groin sheath.

Another possible limitation is that there is a tendency for higher rates of ICA terminus occlusions in the PROTECT group. But as ICA terminus occlusions are prone to be associated with a higher clot burden, they are often more difficult to re-open and often have devastating clinical outcome,30 this might – if at all – be a disadvantage for PROTECT in comparison to the control group. This slight imbalance in combination with the relatively small sample size of the PROTECT group might explain why we were not able to detect a statistical significant difference in the short-term clinical outcome between our two treatment groups. Nevertheless, there was a clear trend toward a better clinical outcome at discharge expressed by the mRS, the NIHSS, the ΔNIHSS, the ‘Good neurological short-term outcome’ and the ‘Substantial neurological improvement’ shortly missing statistical significance. As the majority of the PROTECT cases were treated very recently, we are not yet able to present clinical data after 90 days (mRS 90) which would be preferable and must be mentioned as another limitation. Although our study fails to provide direct evidence for a superior clinical outcome for patients treated with PROTECT, it seems likely that the increase in reperfusion rates will translate into clinical benefit in larger cohorts.


In this study we present the PROTECT technique: a combination of proximal flow arrest with distal aspiration during stent retriever thrombectomy. This technical approach shows superiority concerning procedural duration time and technical results compared with sole distal aspiration during stent retriever thrombectomy. In our opinion, intracranial stent retriever thrombectomy without any embolic protection should be omitted in the future. As ongoing development in the field of periprocedural embolic protection is necessary, PROTECT seems to be a promising approach which warrants further evaluation.



  • Contributors All authors have provided a substantial contribution to the conception and design of the studies and/or the acquisition and/or the analysis of the data and/or the interpretation of the data; drafted the work or revised it for significant intellectual content; approved the final version of the manuscript; and agree to be accountable for all aspects of the work, including its accuracy and integrity.

  • Competing interests None declared.

  • Ethics approval local ethics committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Correction notice Since this paper was first published online the author name Sebastian Moench has been updated to Sebastian Mönch.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.