Prophylactic Impella CP versus VA-ECMO in Patients Undergoing Complex High-Risk Indicated PCI

Objectives To compare two different forms of mechanical circulatory support (MCS) in patients with complex high-risk indicated PCI (CHIP): the Impella CP system and veno-arterial extracorporeal membrane oxygenation (VA-ECMO). Background To prevent hemodynamic instability in CHIP, various MCS systems are available. However, comparable data on different forms of MCS are not at hand. Methods In this multicenter observational study, we retrospectively evaluated all CHIP procedures with the support of an Impella CP or VA-ECMO, who were declined surgery by the heart team. Major adverse cardiac events (MACE), mortality at discharge, and 30-day mortality were evaluated. Results A total of 41 patients were included, of which 27 patients were supported with Impella CP and 14 patients with VA-ECMO. Baseline characteristics were well-balanced in both groups. No significant difference in periprocedural hemodynamic instability was observed between both groups (3.7% vs. 14.3%; p = 0.22). The composite outcome of MACE showed no significant difference (30.7% vs. 21.4%; p = 0.59). Bleeding complications were higher in the Impella CP group, but showed no significant difference (22.2% vs. 7.1%; p = 0.22) and occurred more at the non-Impella access site. In-hospital mortality was 7.4% in the Impella CP group versus 14.3% in the VA-ECMO group and showed no significant difference (p = 0.48). 30-Day mortality showed no significant difference (7.4% vs. 21.4%; p = 0.09). Conclusions In patients with CHIP, there were no significant differences in hemodynamic instability and overall MACE between VA-ECMO or Impella CP device as mechanical circulatory support. Based on this study, the choice of either VA-ECMO or Impella CP does not alter the outcome.


Introduction
In patients with complex coronary artery disease (CAD) or one or more chronic total occlusions (CTO) combined with the presence of comorbidities, determining the optimal revascularization strategy (percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG) remains a challenge. Tese high-risk patients pose an extreme surgical risk and are frequently assigned to medical therapy instead of CABG. However, with the development of new interventional techniques and materials, PCI is a good alternative to CABG and is referred to as complex high-risk indicated PCI (CHIP).
During CHIP, hemodynamics can deteriorate because of temporary complete coronary occlusion or profound myocardial ischemia. Tis could result in loss of cardiac output and fattening of the arterial curve once the ventricle "uncouples" from the aortic pressure with a risk for cardiogenic shock [1]. Mechanical support during CHIP facilitates native cardiac function by achieving a stable hemodynamic state to withstand repetitive derangements such as ischemia caused by prolonged and repeated balloon infations, and resume original cardiac function immediately postprocedure or shortly thereafter [2].
Tere are several mechanical circulatory support (MCS) systems available, i.e., intra-aortic balloon counterpulsation (IABP), Impella, TandemHeart, and veno-arterial extracorporeal membrane oxygenation (VA-ECMO). Tese MCS have been widely studied in patients with acute myocardial infarction (MI) complicated by cardiogenic shock and showed conficting results [3][4][5][6][7][8]. Te most important purpose of MCS is to provide an adequate level of hemodynamic support by augmenting mean arterial pressure and cardiac output and thereby avoiding the use of supplementary vasopressor or inotropic therapy.
However, studies regarding the use of MCS in the setting of CHIP are much less abundant and to our knowledge, no study has compared Impella CP with VA-ECMO in CHIP patients [9][10][11]. Tis retrospective observational study aims to evaluate all patients who were declined for surgery by the heart team and underwent high-risk PCI with the support of either Impella CP or VA-ECMO.

Materials and Methods
A multicenter comparative observational study was performed including all patients undergoing CHIP with the mechanical circulatory support of either Impella CP or VA-ECMO.
In patients who underwent Impella CP support, an Impella catheter was inserted through the femoral artery by using a modifed Seldinger technique and forwarded into the left ventricle under fuoroscopic guidance. Closure of the Impella femoral access site was performed with a minimum of two Perclose ProGlide suture systems (Abbott Vascular, Redwood City, California). Cannulation of the VA-ECMO was performed using open cutdown and decannulation was accomplished through surgical vascular repair of the femoral artery.
All patients were discussed in the heart team prior to intervention and complex high-risk PCI was defned as patients with an unprotected left main artery, last patent vessel, or complex 3-vessel coronary artery disease, with a left ventricular ejection fraction (LVEF) of ≤30%-40% and severe comorbidities (e.g., severe valve disease(s), renal, pulmonary, or cerebrovascular) that were declined for CABG.
Te coronary anatomy of both groups was assessed in regard to the target vessel for revascularization, multivessel disease, and the presence of a CTO lesion. For all individual patients, the SYNTAX score and SYNTAX 2 score were calculated using the online Syntax-score calculator (https:// syntaxscore.org). Te euroSCORE I and II were also calculated for all patients using the online calculator (https://www. euroscore.org/). If there was a CTO lesion present the Japanese-CTO score was calculated using the available online J-CTO calculator (https://www.progresscto.org/cto-scores).
Te outcome was analyzed in regard to the successful revascularization of all planned target vessels.
Hemodynamic instability during MCS support was defned as a systolic blood pressure of < 90 mmHg for ≥ 30 minutes or ventricular arrhythmia. If hemodynamic instability occurred, it was resolved with either fuid resuscitation and/or intravenous inotropic or vasopressor agents.
Observed mortality was analyzed in regard to periprocedural mortality, mortality at discharge, and mortality at 30 days. Major adverse cardiovascular events (MACE) were analyzed for both groups.
MACE at 30 days was defned as a composite of all-cause death, myocardial infarction, stroke or transient ischemic attack (TIA), major bleeding events according to the Bleeding Academic Research Consortium (BARC) criteria, repeat revascularization, need for cardiac surgery, limb ischemia, cardiopulmonary resuscitation or acute renal insufciency. Limb ischemia was assessed by palpation of either the posterior tibial or dorsalis pedis artery of the ipsilateral cannulated femoral artery during the procedure and the occurrence of limb ischemia postprocedure.
Major bleeding events postprocedure were defned according to the BARC criteria and events with a BARC 3 or higher were considered signifcant and included in the analysis [12].
Deterioration of renal function postprocedurally was the increase of one or more KDIGO stages above baseline. Blood hemoglobin levels in mmol/l were assessed prior to and after the procedure as well as thrombocyte count in 10 6 units per ml. Transfer to the Intensive Care Unit (ICU) and the mean stay in the ICU as well as on the ward were also analyzed.
Statistical analysis was performed using a Kolmogorov-Smirnov test to assess the normal distribution of continuous data. If there was a normal distribution of the data a Student's t-test was performed otherwise a Mann-Whitney U test was performed. Categorial variables were analyzed using a Chi-Square test. A two-tailed p value of < 0.05 was considered statistically signifcant.
Te data used in this article are derived from previously published data [11,13]. Both registries were approved by the Medical Ethical Committee at each participating center and written informed consent was obtained from all patients.

Results
Between 2017 and 2020, a total of 41 patients underwent a complex high-risk PCI facilitated by mechanical circulatory support: 27 patients received support with the Impella CP device (between March 2018 and October 2020) and 14 patients received VA-ECMO support (between January 2017 and July 2018). Te decision for the use of mechanical circulatory support was made by the heart team.
Te mean age (p � 0.44) and the number of men (p � 0.15) did not difer signifcantly between the groups. Other baseline characteristics were well-balanced in both groups (Table 1). Most patients had an LVEF < 35% (70.4% vs. 71.4%). In both groups, 6 patients were known with peripheral artery disease, but either already treated or not signifcant.
A left anterior descending(LAD) lesion was signifcantly more apparent in the Impella group (88.9% vs. 57%; p � 0.02 ), but other lesions showed no diference. A majority of the patients in both groups had a CTO (74.1% vs. 71.4%). EuroSCORE, SYNTAX, and J-CTO scores between the two groups did not difer (Table 2), however, a higher J-CTO score occurred in the Impella group (2 vs. 1).
Hemodynamic instability during the procedure occurred in one patient in the Impella group (3.7%) and in two patients in the VA-ECMO group (14.3%). Tis diference was not statistically signifcant (p � 0.22). All hemodynamic unstable patients developed hypotension with a systolic blood pressure < 90 mmHg for more than 30 minutes and this hemodynamic instability was treated successfully with the administration of intravenous vasopressor agents in two patients and the other patient with fuid resuscitation.
Te composite outcome of MACE was higher in the Impella-facilitated CHIP group but showed no signifcant diference (30.7% vs. 21.4%; p � 0.59) ( Table 3). Bleeding complications (BARC >3 or higher) occurred more often in the Impella group, but showed no statistically signifcant diference (22.2% vs. 7.1%; p � 0.22). Te VA-ECMO patient developed a cardiac tamponade which was successfully evacuated by pericardial drainage. Te 6 bleeding complications in the Impella groups had femoral access-site-related bleeding of which 2 patients (7.4%) had signifcant bleeding from the Impella access site and needed acute vascular surgery to restore the defect. Te remaining 4 patients (14.8%) had a grade 3A hematoma at the non-Impella access site but absence of active bleeding. [13].
Limb ischemia did not occur in either group periprocedurally.
Revascularization was successful in 25 patients in the Impella group (92.6%) and in 14 patients in the VA-ECMO group (100%) and showed no signifcant diference (p � 0.47).
Kidney function deteriorated in 4 patients in the Impella group (14.8%) and 3 patients in the VA-ECMO group (21%).
Tere was no statistically signifcant diference in regard to hemoglobin levels post PCI (6.7 vs 6.4; p � 0.35) between both groups. Te mean length of hospital stay showed no signifcant diference between the groups but was overall longer in the VA-ECMO group (3 vs. 7 days; p � 0.23). Tere was a trend towards more transfer to ICU after Impella in comparison with VA-ECMO (18.5% vs 0%; p � 0.06).

Discussion
Our study results indicate that in patients with CHIP who underwent either Impella CP or VA-ECMO assisted PCI, no statistically signifcant diference in the occurrence of periprocedural hemodynamic instability was observed (3.7% vs 14.3%; p � 0.22). In the 3 patients, where hemodynamic instability during the procedure did occur, the signifcant drop in blood pressure was resolved with temporary inotropic agents. In comparison with data from the PROTECT II trial, hypotension during support with an Impella 2.5 occurred in 10.2% of the patients whereas with the Impella CP it occurred in only 3.7% [10]. Data on hemodynamic support with VA-ECMO are scarce. One single-center prospective study on 12 patients stated that the procedures were well tolerated and this could be explained by the stable hemodynamic status of patients in absence of cardiogenic shock or cardiac arrest [14].
VA-ECMO is a modifcation of the cardiopulmonary bypass circuit that provides a continuous, nonpulsatile cardiac output [15]. It is the only MCS that also oxygenates the blood by removing carbon dioxide from and adding oxygen to venous blood via an artifcial membrane [16]. VA-ECMO can provide signifcant hemodynamic support but has the propensity to increase LV afterload and wall stress, which in turn can increase myocardial oxygen consumption and therefore limit any cardioprotective beneft [17,18]. To prevent these hemodynamic unfavorable efects, VA-ECMO is in long-term support and generally used in combination with other devices for MCS such as Impella or IABP, but not for short-term support as in CHIP [19,20]. During VA-ECMO support, vasodilators might reduce afterload and LV end-diastolic pressure, while inotropes can increase contractility [21].
Te Impella CP (Abiomed Inc., Danvers, Massachusetts) is a micro-axial pump positioned across the aortic valve which aspirates blood from the left ventricle into the ascending aorta [20]. Te efect of LV unloading reduces end-diastolic wall stress, improves diastolic compliance, increases aortic and intracoronary pressure and coronary fow velocity reserve, and stimulates a decrease in coronary microvascular resistance [22].
In our study, the choice of MCS was left to the discretion of the operators, local expertise, and available facilities. Not every patient is suitable for all MCS and the choice depends on patient characteristics such as anatomy and procedural characteristics.

Journal of Interventional Cardiology
In-hospital and 30-day mortality were not diferent between both groups, but a trend toward higher 30-day mortality was seen with VA-ECMO-assisted PCI. In both groups, 2 patients died during hospitalization. Careful consideration of these patients by several physicians, including the attending intensivist, cardiologist, interventional cardiologist, and anesthesiologist, concluded that all in-hospital deaths were not related to either the PCI itself or the MCS support. All patients were known with end-stage heart failure and several comorbidities and died as a result of multiorgan failure due to   persistent cardiogenic shock. In the VA-ECMO group, mortality is up to 30 days extended to 1 patient. Te patient who died after discharge died as a result of persistent heart failure. No diference in MACE or bleeding complications between both groups was observed in our study population. Tere were nonetheless numerically more bleeding complications in the Impella group, mainly driven by femoral access site bleeding at the non-Impella site. In only 2 patients there was signifcant bleeding at the access site. A possible solution to reduce the number of access site bleedings of the non-Impella site is to use the single-access technique for Impella [23]. Using the Impella introducer sheath for PCI access decreases the number of access sites and potentially reduces complications related to multiple access sites.
Te reason for no access site bleeding complications in the VA-ECMO group might be explained by a controlled insertion and removal of the cannulas. Te cannulas of VA-ECMO are inserted and extracted by surgical cutdown and closure of the femoral artery. In comparison, the femoral access site of the Impella in these patients was closed with the ProGlide closure device. Nonetheless, in only 2 patients in the Impella group the bleeding complications were related to bleeding from the Impella access site, so most bleeding complications arose from the non-Impella access site.
Limb ischemia did not occur in either group. Tis illustrates the safety of short-term support of MCS during CHIP in regard to limb ischemia. However, in this semielective setting of CHIP, patients with known signifcant peripheral artery disease which could be compromised during or after the procedure, would not be selected for Impella or VA-ECMO. One patient developed venous thromboembolism several weeks after VA-ECMO cannulation at the venous cannulation site. Tis was resolved without further consequences by treating this patient with oral anticoagulation.
Tis is to our knowledge the frst comparative study between Impella and VA-ECMO in the setting of CHIP. A comparison study between Impella and VA-ECMO in patients with cardiogenic shock showed lower in-hospital mortality, fewer complications, decreased hospital costs, and decreased length of stay in patients supported with Impella in comparison to VA-ECMO. Moreover, there was a higher rate of ischemic stroke and vascular complications in the VA-ECMO cohort [24]. Our study does not display these differences in the setting of CHIP and therefore it is justifable to use either device in CHIP. Keeping in mind that each MCS has its contra-indications for use, such as LV thrombus for Impella CP and severe aortic regurgitation for both MCS.
Tis is an observational study and therefore no direct comparison between both groups in a randomized fashion was performed. Data was collected retrospectively with all shortcomings of such. Te sample size is relatively small and therefore hypothesis-generating, but may pave the way for future randomized studies.

Conclusions
In this observational study, there was no diference in hemodynamic instability and MACE in CHIP with either VA-ECMO or Impella CP device for mechanical circulatory support. Although statistically not signifcant, bleeding complications were higher in patients supported with the Impella CP, but these were mainly driven by bleeding complications from the non-Impella access site. Based on this study the choice of either VA-ECMO or Impella CP does not alter the outcome. Future research, preferably in a randomized fashion, is needed to establish the most efective, safe, and fnancially favorable form of MCS in patients undergoing high-risk PCI.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author (F.S. van den Brink) upon request.

Disclosure
Tere is no fnancial support or relationship that may pose conficts of interest.