Article Text

Download PDFPDF

Original research
Impact of reperfusion pump power on technical and clinical outcomes after direct aspiration thrombectomy (ADAPT)
  1. Ali Alawieh1,2,
  2. Reda Chalhoub1,
  3. Clayton J Korson1,
  4. Mohammad Anadani3,
  5. Jonathan Lena1,
  6. Alejandro Spiotta1
  1. 1 Neurosurgery, Medical University of South Carolina, Charleston, South Carolina, USA
  2. 2 Neurosurgery, Emory University, Atlanta, Georgia, USA
  3. 3 Neurology, Medical University of South Carolina, Charleston, South Carolina, USA
  1. Correspondence to Dr Alejandro Spiotta, Neurosurgery, Medical University of South Carolina, Charleston, South Carolina, United States; spiotta{at}


Introduction Higher aspiration forces using larger bore catheters in direct aspiration thrombectomy (ADAPT) have been associated with shorter procedure time and better outcomes in patients treated for acute ischemic stroke (AIS). However, the effect of using reperfusion pumps of higher aspiration power on outcomes of ADAPT has not been investigated. We evaluated the effect of using pumps with different aspiration powers on technical and clinical outcomes after ADAPT.

Methods We reviewed a retrospective database of AIS cases between January 2018 and February 2019, while comparing technical and clinical outcomes between patients undergoing ADAPT using the MAX pump (28.5 inHg power) vs the ENGINE pump(29.2 inHg power).

Results Among 194 patients (48% females, age 69±15 years) included in the study, 73 patients undergoing ADAPT using the ENGINE pump were age-, gender-, comorbidities-, and operator-matched to 118 patients treated using the MAX pump. The ENGINE group had shorter procedure time (20±17 vs 27±21 mins, p=0.017), lower number of aspiration attempts (2.2±1.6 vs 2.8±1.9, p=0.047), and similar rates of favorable 90 day modified Rankin Scale. Using multivariate linear regression, the use of the ENGINE reperfusion pump with higher vacuum power was independently and inversely correlated with procedure time (coefficient −2.23, p=0.027). While controlling for confounders, there was a trend toward an inverse correlation between use of the ENGINE pump and the number of attempts on linear regression (coefficient −1.04, p=0.09) and lower odds of PH2/intracranial (ICH) hemorrhages on logistic regression (OR 0.227, p=0.075).

Conclusion Our findings suggest that the use of the ENGINE reperfusion pump of higher aspiration power during ADAPT decreases procedure time, without increasing complications and post-procedural hemorrhage rates.

  • large vessel occlusion
  • stroke
  • ADAPT thrombectomy
  • aspiration thrombectomy
  • procedure time

Statistics from


Endovascular mechanical thrombectomy (MT) is the current standard of care in patients with acute ischemic stroke (AIS) due to large vessel occlusion.1–8 Following the success of initial thrombectomy trials using stent retrievers to achieve vessel recanalization, frontline aspiration thrombectomy using a direct aspiration at first pass technique (ADAPT) has shown in randomized trials to be a non-inferior, fast and effective alternative.6 9 This led to further expansion in the arsenal of devices available for the treatment of AIS.

Following its implementation, ongoing advances in the ADAPT technique and the devices used for aspiration have led to improved technical and clinical outcomes,10 11 leading to shortened procedure times and better safety profiles. Compared with smaller catheters, the use of higher caliber catheters (eg, 0.068 inch 5MAX 068ACE) with greater aspiration forces was associated with shorter procedure time, increased rate of first pass success and improved clinical outcomes compared with smaller caliber devices.10 However, beyond the 0.068 inch catheters, the use of alternative similar or higher caliber devices was not associated with further improvement in rates of first pass success, procedure time, or clinical outcomes.12 Alternatively, aspiration power could be further enhanced by using aspiration pumps with higher aspiration forces which have recently become available to neurointerventional surgeons. However, the safety and efficiency of using more powerful pumps have not been previously examined. The aim of this study is to examine the effect of using two different aspiration pumps (the ENGINE pump with 29.2 inHg power vs the MAX pump with 28.5 inHg) on technical and clinical outcomes and rates of hemorrhage following ADAPT thrombectomy for large vessel occlusion stroke at a high-volume comprehensive stroke center.


Patient selection

We retrospectively reviewed a prospectively maintained database of all AIS patients who underwent mechanical thrombectomy using ADAPT between January 2018 and February 2019 at the Medical University of South Carolina (MUSC). The institutional review board at MUSC approved this study without patient consent, as no identifying material had been included. Patients were selected for mechanical thrombectomy based on CT perfusion imaging, independent of the time of onset. They were included in the study irrespective of age, type of occluded vessel, onset to groin time, ASPECT score on CT imaging, or preprocedural administration of intravenous tissue plasminogen activator (IV-tPA). Patients undergoing thrombectomy with either the MAX pump (Penumbra Inc) or the newer version ENGINE pump (Penumbra Inc) were included. Patients undergoing ADAPT using the ENGINE pump were matched by age, gender, admission National Institute of Health Stroke Scale (NIHSS), comorbidities and location of occlusion, and operating interventionalist, to the most recent patients undergoing ADAPT using the MAX pump using propensity score matching with a 1:1.5 ratio. To limit the impact of the operator on outcomes, procedures performed by the same two interventionalists were included in the study.

Adapt thrombectomy

ADAPT has been previously described.10 13 The largest aspiration catheter that the vessel could accommodate was selected for direct aspiration. Larger bore catheters were used for M1 occlusion while smaller bore catheters were used for more distal occlusions. The catheters used include: 5MAX ACE 60 (inner diameter (ID) 0.060 inch), ACE 64 (ID 0.064 inch), ACE 68 (ID 0.068 inch), Jet 7 (ID 0.072 inch), 3MAX (ID 0.035 inch), and 4MAX (ID 0.041 inch). The proximal end of the catheters was attached to one of two pumps: the ENGINE pump and the MAX pump operated at vacuum pressures of 29.2 inHg and 28.5 inHg, respectively. The aspiration pressure was applied differently between the two pumps. Whereas the MAX pump ramps the aspiration pressure to maximum while engaging the clot, the ENGINE pump allows vacuum to fill up in the pump before engaging the clot at a maximum and steady aspiration force. Three to four attempts of direct aspiration with ADAPT were done before an alternative rescue technique with a stent retriever thrombectomy was employed at the discretion of the operating physician. All patients mainly received conscious sedation throughout the procedure.

Data collection

Patient data were collected by reviewing the patients’ charts, procedure notes and radiology reports, for previously selected procedural and preprocedural variables and outcomes. Preprocedural variables included age, gender, race, relevant comorbidities, the NIHSS at admission, and the pre-stroke modified Rankin Scale (mRS). Procedural variables captured encompassed procedure time, total number of attempts, number of attempts with ADAPT only, the type of catheters used in the procedure, recanalization level as defined by the modified Thrombolysis in Cerebral Infarction (mTICI) score, and direct procedural complications. Procedure time was defined as the duration between groin puncture and recanalization. In case of failed thrombectomy or procedural complication leading to termination of the procedure, the procedure time was calculated as the duration between groin puncture and groin closure.

Clinical outcomes

The modified Rankin Scale (mRS) at 90 days after stroke was used as the primary outcome of the procedure. An mRS score of 0–2 was considered to be a “good outcome”, while an mRS score >2 was considered a “poor outcome”. The mRS scores at 90 days were collected during routinely scheduled follow-up visits at a neurology clinic 90 (±15) days) after stroke onset. Mortality rates and post-procedural hemorrhages following the procedure were used as a secondary outcome in this study. Post-procedural hemorrhage was classified based on European Cooperative Acute Stroke Study (ECASS) criteria,14 and symptomatic intracranial hemorrhage (sICH) was defined as post-procedural hemorrhage with associated decline in NIHSS.

Statistical analysis

Statistical analysis was done using SPSS 24 (IBM Corporation) and GraphPad Prism 8 (GraphPad). Univariate analyses were performed using student’s t-test for Gaussian distributed parametric variables, Mann-Whitney test for non-parametric variables, and χ2 or Fisher exact test for categorical variables. Multivariate models were performed to assess the impact of the pump on technical and clinical outcomes while adjusting for covariates based on p<0.1 on univariate analyses or clinical significance. Logistic regression models were used for prediction of 90 day mRS scores dichotomized into good and poor outcome and incidence of sICH/PH2 type hemorrhage. Linear regression models were used for prediction of procedure time and number of attempts required to achieve recanalization. An α<0.05 was considered for statistical significance.


Patient population

A total of 194 patients (48% females, 52% white race) were included in this study, of whom 118 underwent ADAPT thrombectomy for AIS using the MAX pump (28.5 inHg), while 76 underwent the same procedure using the ENGINE pump (29.2 inHg). The mean age at presentation was 69±15 years, 91% of the patients had a pre-stroke mRS between 0–2, and the average NIHSS at admission was 16±8 (table 1). The majority of patients (92%) were treated for anterior circulation stroke with a median ASPECT score of 8, and M1 strokes were the most common location treated (47% of patients (table 1)). Among patients presenting with posterior circulation occlusion (8%), basilar occlusions were the most common (5% of full cohort). The average onset to groin time for this cohort was 459±702 min.

Table 1

Patient variables and outcomes between patients treated using different perfusion pumps

Patients treated using either the MAX pump or the ENGINE pump had similar baseline demographic characteristics, similar incidence of comorbidities, and similar admission deficit scores (p>0.1 for all variables). The mean NIHSS were similar between patients undergoing thrombectomy using MAX or ENGINE pumps (16±7 vs 16±7, p>0.1). Additionally, the mean onset to groin time and rate of IV-tPA use were comparable between the two groups (p>0.1, table 1).

Technical variables

The distribution of reperfusion catheters by size (large bore (ACE 68/Jet 7 vs smaller bore ACE 60/3MAX/4MAX) was similar in patients undergoing MT using MAX or ENGINE pumps (p>0.1). More than half of the patients in either group underwent MT using larger bore catheters (table 1). However, the procedure time was significantly shorter in patients with the ENGINE pump compared with the MAX pump (20±17 mins vs 27±21 mins, p=0.017) (table 1); similarly, the average number of aspiration attempts required to achieve recanalization was lower in the ENGINE versus MAX groups (average of 2.8 vs 2.2, p=0.047), and a lower percentage of patients required more than three attempts in the ENGINE versus MAX groups (p=0.043). Although procedure time was lower in the ENGINE group, the percentage of MT procedures completed within 60 min was still comparable between the two groups (p>0.1).

On multivariate regression analysis, and while controlling for covariates and confounding variables, the use of the ENGINE pump compared with the MAX pump was an independent predictor of shorter procedure time (coefficient −2.23, p=0.027) (table 2). Using an independent linear regression model for the number of aspiration attempts, there was a trend toward significantly lower aspiration attempts with the use of the ENGINE pump compared with the MAX pump (coefficient −1.04, p=0.09) (table 2).

Table 2

Multivariate models for impact of reperfusion pump on mechanical thrombectomy outcome

There was no significant difference in the distribution of mTICI scores (p>0.1) achieved using either pump, nor in the rate of first pass success (p>0.1) (table 1). Yet, a higher percentage of patients required rescue therapy with a stent retriever in the MAX group compared with the ENGINE group (33% vs 14%, p=0.004) (table 1).

Clinical outcomes

The median mRS scores at 90 days was 3 for the full cohort (IQR=3), and it was comparable in patients treated using the ENGINE or MAX pumps (table 1). Good outcomes were defined as mRS 0–2 and occurred in 46% of patients treated with MAX and 43% of patients treated with ENGINE without significant difference between the two groups (table 1). The overall mortality rate was 17% and was comparable between the two groups (table 1). On multivariate logistic regression analysis and while controlling for significant confounding variables (table 2), the use of the ENGINE pump vs the MAX pump was not an independent predictor of favorable outcome at 90 days (OR 1.06, 95% CI 0.46 to 2.45; p=0.89).

Procedural complications

The overall rate of procedural complications was 4%, and the rate of distal embolization was 6% with no significant difference between patients treated with either pumps (table 1). Patients treated with the ENGINE pump had lower rates of PH2/sICH hemorrhage on follow-up CT scans; however, the difference was not statistically significant (5% vs 9%, p=0.2) (table 1).

Using multivariate logistic regression analysis, there was a trend toward significantly lower odds of sICH and PH2-type hemorrhage in patients treated using the ENGINE pump compared with the MAX pump (OR 0.227, 95% CI 0.04 to 1.16; p=0.07) (table 2).

Combined effect of reperfusion catheter and pump use

We also assessed the impact of the ENGINE pump versus the MAX pump on procedure time and rates of hemorrhage between patients treated with the same aspiration catheters (figure 1). The reduction in procedure time with the use of the ENGINE pump compared with the MAX pump was apparent even when the cohort was split by reperfusion catheter (3MAX, 4MAX, ACE68), and was statistically significant in the 4MAX and ACE 68 groups (figure 1A). Notably, even in patients treated with the larger-bore catheter ACE 68 (0.068 inch), the use of the ENGINE pump was associated with a significant reduction in procedure time by 7 min (p<0.05, figure 1A). The rates of sICH and PH2-type hemorrhage were quantitatively lower for each catheter group with the use of the ENGINE pump compared with the MAX pump (figure 1B). No significant difference was observed in rates of hemorrhage between large bore catheters (ACE 68 and Jet 7).

Figure 1

Impact of reperfusion pump on technical outcomes split by reperfusion catheter. (A) Comparison of procedure time (in min) between the ENGINE pump and MAX pump for each of 3MAX, 4MAX, and ACE 68 catheters. No comparisons were made for ACE 60 or Jet 7 since they were used only with MAX and ENGINE pumps, respectively. Student’s t-test was used. *P<0.05. (B) Comparison of the rates of post-procedural PH2/sICH hemorrhage between the engine pump and the max pump for each of 3MAX, 4MAX, and ACE 68 catheters. sICH, symptomatic intracranial hemorrhage.


Following the success of the major randomized controlled trials on MT for AIS using stent retrievers as the frontline device,1–8 frontline aspiration thrombectomy using ADAPT has been shown to be a faster and non-inferior alternative to stent retrievers.6 9 Faster thrombectomy procedures were associated with better clinical outcomes following MT in datasets derived from real-world reports and clinical trials.11 13 15 Based on studies comparing different thrombectomy aspiration catheters, higher aspiration forces are likely to increase the success rate of ADAPT thrombectomy leading to the current question of whether higher powered aspiration pumps can augment the efficacy of ADAPT thrombectomy without compromising safety. Our findings demonstrate that the use of the ENGINE pump was associated with faster procedure times and a lower number of aspiration attempts compared with the MAX pump with lower aspiration power without an increase in complications, distal embolization and hemorrhage rates.

With the advent of the ADAPT technique, the physics of the thrombectomy procedure became more dependent on the use of high force applied to the occlusive clot to allow its removal in one piece. Given the differences in thrombus consistency, the force applied, referred to as thrombus removal force (TRF), may be sufficient to alter the shape of the clot, allowing complete ingestion into the catheter or, in more organized clots, it allows for holding the corked clot in the catheter during retrieval. The TRF is directly proportional to the area of clot engagement and the vacuum differential between the pump pressure and the arterial pressure. Therefore, optimizing the TRF can be achieved by either increasing the surface area of engagement (using larger catheters) or by using pumps with higher aspiration powers.

The impact of aspiration force on technical and clinical outcomes after ADAPT thrombectomy was previously investigated by comparing reperfusion catheters with different calibers.10 12 16 Increasing catheter caliber from 0.054 inch (5 MAX) to 0.068 inch (ACE 68) is associated with a 60% increase in the aspiration force and subsequent reduction in procedure time, higher success rate of aspiration and better clinical outcomes.10 However, comparison of different large bore catheters did not show an improvement in technical or clinical outcomes compared with the ACE 68 catheter.12 An alternative approach to improve aspiration force and success involves the use of stronger aspiration pumps. Aspiration through reperfusion catheters can be either performed using manual suction via a syringe or using reperfusion pumps; the latter allow a more systematic comparison of aspiration force due to the uniform force produced by different pumps.

In this work, we compared two pumps with different aspiration forces; the ENGINE pump with 29.2 inHg power and the MAX pump with 28.5 inHb power. During ADAPT, this additional power with ENGINE is directly translated to the catheter’s engagement, holding and ingestion of the clot. We show that the use of the ENGINE pump was an independent predictor of faster recanalization rates, and was associated with a lower number of aspiration attempts and lower likelihood of the need for rescue therapy with stent retrievers. Interestingly, the effect of the pump in our cohort was preserved even when patients treated with a larger bore (ACE 68) catheter were strictly included. Due to the low number of patients in the Jet 7 group, comparison of increased aspiration force with Jet 7 compared with ACE 68 could not be performed.

On average, the use of the ENGINE pump saves an additional 7 min during ADAPT which corresponds to a 26% reduction in thrombectomy procedure time. Although significant from a technical perspective, this reduction in procedure time was not associated with improvement in clinical outcomes using 90 day mRS scores. The likely rationale is that the average procedure duration was low in our cohort where 80% of patients were recanalized within 60 min of groin puncture; therefore, the reduction in procedure time within this first 60 min does not have a significant impact on clinical outcome as shown previously in larger cohort studies.11 Alawieh et al previously reported, using a large cohort of patients, that among patients recanalized within 30 or 60 min of procedure time, changes in procedure time did not independently predict outcome.11 In our cohort, the average procedure time was 24 min which falls within the 30 min window, thus explaining why the technical improvement did not translate into a clinical improvement when assessed by 90 day mRS scores.

Additionally, our findings demonstrate that the use of a stronger pump was associated with a lower number of recanalization attempts required to achieve successful treatment, and, at the same time, a lower likelihood of a need for alternative rescue therapy using a stent retriever. This provides an additional economic advantage and decreases the overall number of reperfusion devices required per procedure. Additionally, the number of attempts have previously been associated with an increased risk of post-procedural PH2 or sICH. This is also reflected in our data where the use of the ENGINE pump showed a trend toward a significant reduction in the risk of PH2/sICH hemorrhage (p=0.07). The reduction of risk of hemorrhage was also observed for each individual reperfusion catheter (figure 1B). At the same time, there were no significant changes in risk of complications or distal embolization with use of the ENGINE pump compared with the MAX pump.

Notably, a prior report using in vitro analysis suggested that higher maximal aspiration forces may be achieved with manual aspiration compared with the MAX pump.17 Comparison of technical and clinical outcomes between manual aspiration and pump aspiration has not been performed.17 However, mean procedure times in the reported literature using manual aspiration was 35 min17 compared with 27 min with the MAX pump and 20 min with the ENGINE pump in this study. One feature of using aspiration pumps that may contribute to a lower number of recanalization attempts and faster procedure is the presence of “flip-the-switch” technology that allows the application of maximum aspiration force immediately at the time of engagement with the clot. Using manual aspiration, the aspiration force ramps up to the maximum force gradually while engaging the clot. Notably, the ENGINE pump allows for faster accumulation of aspiration force in the pump compared with the MAX pump and a more user-friendly visual indicator of readiness to engage the clot.

Although the operator change may impact technical variables using the different catheters or pumps including procedure time, we limited the impact of operator experience by selecting only patients treated by two operators and matched patients by operators, and included patients treated within 6 months for each of the pumps to avoid dilution or augmentation of pump effect by operator experience.


This study is a retrospective review of a limited number of patients (n=194) from a single institution, and thus selection bias cannot be avoided. Due to the lower number of patients in the ENGINE pump group (n=76), the study may have been underpowered for smaller effect sizes, especially when assessing the risk of hemorrhage where a statistical trend (p=0.07) was observed. We have used a 1:1.5 matching (n=76 in ENGINE group vs n=118 in MAX group) to improve power. We did not include additional patients to limit our study to a specific interventionalist and reduce the impact of procedure timing (later treatment vs earlier treatment). Therefore, we have used a limited study period that included the transition period between the use of the two pumps at our center. In any case, our data had sufficient power to detect a 7 min reduction in procedure time observed with the use of a more powerful pump.


The use of higher powered aspiration pumps with capability for immediate application of maximal force can improve technical outcomes after ADAPT thrombectomy for ischemic stroke without compromising the safety of the procedure.



  • Twitter @ChalhoubReda

  • Contributors Each author listed should receive authorship credit based on the material contribution to this article, their revision of this article and their final approval of this article for submission.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests JL: consultant–Penumbra. AS: consultant–Penumbra, Stryker, Cerenovus, research support–Penumbra, Stryker, Cerenovus.

  • Patient consent for publication Not required.

  • Ethics approval The study was approved by the Institutional Review Board at the Medical University of South Carolina under expedited review with informed consent waived given the retrospective nature of the study.

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

  • Data availability statement Data are available upon reasonable request.

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.