Highlights
-
•
The use of Impella during high-risk percutaneous coronary intervention (PCI) is increasing, yet current evidence is limited to registry-based data. No robust prospective trials have established its safety or efficacy in patients with severe coronary disease and left ventricular dysfunction who are unsuitable for surgery.
-
•
This is the first meta-analysis to specifically quantify 6-month outcomes in high-risk patients undergoing Impella-supported PCI, reporting a 13.4% mortality rate and acceptable adverse event rates. It offers the most comprehensive synthesis of available real-world data to date.
-
•
Our findings provide a reference for procedural outcomes in this fragile population and support a more confident use of protected PCI in selected nonsurgical candidates, reinforcing the role of mechanical circulatory support in contemporary interventional cardiology.
Surgical revascularization is still considered the gold standard for patients with complex coronary artery disease and left ventricular dysfunction. The advent of Impella has sparked growing interest, yet current evidence on its efficacy remains inconclusive. All studies reporting outcomes beyond 30 days outcomes of pPCI with any Impella device were included. Pooled effect of estimated outcomes was calculated according to a random-effect model with generic inverse variance weighting. Primary endpoint was all-cause mortality. Secondary outcomes were myocardial infarction, repeat revascularization, rehospitalization for heart failure and stroke. Six studies globally encompassing 1,581 patients were included in the quantitative analysis. Median age was 70 years old (IQR 69 to 72) with a median left ventricular ejection fraction (LVEF) of 27 % (SD ± 6) and a SYNTAX SCORE of 31 (IQR 29 to 35). Impella 2.5 was the most common micro axial flow pump used to support high-risk PCI. All-cause of death was observed in 13.4% (95% CI: 10.4 to 16.4) of patients at 6 months median follow-up. Myocardial infarction occurred in 5.8% (95% CI 3.4 to 8.1) of patients, repeat revascularization in 9.1% (95% CI: 4.8 to 13.3) of patients, stroke in 1.6% (IQR 1.2 to 2.1) of patients and, finally, heart failure rehospitalization in 8.4% (95% CI 3.3 to 13.6) of patients. In conclusion, for high-risk patients, PCI with the Impella device represented a viable strategy with an acceptable risk profile when surgical revascularization is not an option, and a poor prognosis is predicted.
Percutaneous microaxial flow pump (Impella, Abiomed, Danvers, MA, US) represents a breakthrough innovation in interventional cardiology, redefining the standards for mechanical circulatory support through its unique design and mechanism of action. Unlike the intra-aortic balloon pump (IABP), which relies on diastolic counterpulsation to increase coronary perfusion and reduce afterload, Impella provides continuous forward flow from the left ventricle into the ascending aorta, actively unloading the left ventricle, reducing myocardial oxygen demand, improving cardiac output and providing systemic perfusion. , This results in more robust circulatory support, particularly in patients with reduced ventricular function. Indeed, current European guidelines recognize the potential benefit of Impella in critical cardiogenic shock (CS) or end-stage heart failure as “bridge therapy” with a Class IIA recommendation.
Although not strongly endorsed by the European Society of Cardiology, Impella may play a critical role in selected high-risk patients, particularly those with complex coronary anatomy and severely reduced ejection fraction, to prevent hemodynamic deterioration during revascularization procedures. However, despite its growing use and promising observational data, widespread adoption is limited by the lack of large-scale randomized controlled trials (RCTs) demonstrating clear mortality or long-term outcome benefits. Ethical concerns, patient selection difficulties and logistical challenges have contributed to this paucity. As a result, current recommendations are largely based on registry data and expert consensus, highlighting the need for further robust clinical evidence. ,,,
This was the basis for conducting this comprehensive meta-analysis to quantify the evidence for high-risk patients undergoing percutaneous coronary revascularization with mechanical circulatory support (MCS) with Impella.
Methods
This analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines and was preregistered in the international prospective register of systematic reviews (PROSPERO CRD420251056404). The data supporting the results of this study are available from the corresponding author on reasonable request. Institutional review board approval for this study was waived due to lack of individual patient information. Written patient consent for publication of the study was not obtained due to lack of individual patient information.
Search study
Randomized controlled trials (RCTs) and observational studies on protected percutaneous coronary interventions were evaluated for inclusion in this metanalysis. Searches were conducted on 20 April 2025 in the following databases: PubMed Central, Ovid EMBASE and Scopus. The search strategy included the following terms: “assisted percutaneous coronary interventions,” “supported percutaneous coronary interventions,” “Impella,” “percutaneous mechanical circulation support device,” “microaxial flow pump,” “clinical outcomes,” “high-risk patients.”
Study selection and data extraction
Database searches, screening and exclusion of duplicate results were performed by 2 physicians (G.D.P., R.I.). Two investigators (G.D.P.; R.I.) screened the searched databases for inclusion and performed data extraction independently. Disagreements were resolved by a third author (M.M.) who also checked the extracted data for accuracy.
The prespecified inclusion criteria were: (1) studies enrolling at least 50 patients, to minimize small-study effects and ensure sufficient statistical stability of pooled estimates; (2) clinical follow-up available beyond 30 days, to focus on outcomes longer than immediate periprocedural events; (3) no overlapping populations, verified by cross-checking enrolment centers and study periods. To avoid duplicating the inclusion of patients, we retained only the report with the largest sample size or the longest follow-up in the event of potential overlap. Exclusion criteria were summarized as follows: (1) patients undergoing Impella-assisted percutaneous coronary intervention with LVEF within the normal range; (2) studies with limited clinical follow-up within 30 days. Animal studies, case reports, conference presentations, editorials, reviews and expert opinions were also excluded. As there is currently no universally accepted definition of high-risk percutaneous coronary intervention (HR-PCI), we applied a pragmatic framework requiring at least 1 clinical and 1 anatomical high-risk feature. The clinical criteria were: age over 75 years; diabetes; left ventricular ejection fraction (LVEF) under 35%; acute coronary syndromes; prior cardiac surgery; peripheral vascular disease; advanced chronic kidney disease (glomerular filtration rate (GFR) under 30 ml/min/1.73 m²); chronic obstructive pulmonary disease; or severe valvular disease. Anatomical features comprised unprotected left main disease, degenerated vein grafts, severely calcified lesions requiring rotational atherectomy, the last remaining conduit and chronic total occlusions in multivessel disease.
The full text of the selected studies was reviewed for a second round of eligibility screening. Reference lists of articles were also searched to identify further relevant studies. The PRISMA flowchart of the study selection process is shown in Figure 1 . For each included study, data on relevant baseline variables were extracted and collected. The quality of the included studies was assessed using the ROBINS-I tool for observational studies and the ROB2 tool for randomized controlled trials. Publication bias was assessed by visual inspection of funnel plots.
PRISMA flow diagram of study selection. PRISMA = preferred reporting items for systematic reviews and metanalysis.
Outcomes measures
The primary endpoint was all-cause mortality at the longest follow-up. Secondary endpoints were: (1) in-hospital outcomes, including acute kidney disease, in-hospital major bleedings and vascular complications requiring surgery; (2) late events, including myocardial infarction, repeat revascularization, readmission for heart failure and stroke.
Statistical analysis
Continuous variables are presented as mean (± SD) or median (interquartile range, IQR). Categorical variables are presented as n (%). Incidence estimates were pooled using a random-effects model with generic inverse variance weighting, and risk estimates with 95% CIs were calculated. The inverse variance method is so called because the weight given to each study is chosen to be the inverse of the variance of the effect estimate (i.e., one above the square of its standard error). Thus, larger studies with smaller standard errors are given more weight than smaller studies with larger standard errors. This choice of weight minimizes the imprecision (uncertainty) of the pooled effect estimate. Hypothesis testing for superiority was performed at the 2-tailed 0.05 level. Cochran’s Q statistic determined heterogeneity, and I 2 values of 25%, 25% to 50%, and 50% to 75% represent low, moderate, and high heterogeneity, respectively. A p-value < 0.05 was considered statistically significant. Statistical analysis was performed using RevMan 5.4 (The Cochrane Collaboration, The Nordic Cochrane Centre, Copenhagen, Denmark, 2014). Additional subgroup analyses were carried out in addition to the main study. The studies were divided into groups depending on the prevalence of unprotected left main disease (>50% of patients), multivessel coronary disease (>50%), and mean left ventricular ejection fraction (<30% vs ≥30%). Between-subgroup heterogeneity was assessed using the chi-squared test and the I² statistic. A leave-one-out analysis was performed to assess the impact of between-study variability on the primary outcome using Python v.3.11.
Results
Seven studies ,,,,,,, with a total of 1,581 patients undergoing Impella-assisted percutaneous coronary intervention were included in the quantitative analysis. The characteristics of the included studies are shown in Table 1 . The baseline clinical characteristics and procedural data of the included patients are summarized in Table 2 . In brief, the median age was 70 (IQR 69 to 72) years, and they were characterized by a high burden of cardiovascular risk factors: 85% had arterial hypertension, 67% had hyperlipidemia and 47% were diabetic. 44% of the included population had a previous myocardial infarction. The pooled median LVEF was 27% (IQR 23 to 31), while the median pre-PCI SYNTAX SCORE was 30 (IQR 29 to 35). Angiography revealed: obstructive left main coronary artery disease in 46% (IQR 39 to 64) of patients, obstructive multivessel coronary artery disease in 66% (IQR 31 to 92) and chronic total occlusions in 14% (IQR 11 to 44) of patients. Most patients (66%, IQR 61 to 81) were treated with the Impella 2.5 during the index procedure.
Table 1
Included studies
| Study | Design | Sample size | Follow-up | Type of Impella |
|---|---|---|---|---|
| IMPELLA PL Registry | Observational | 253 | 12 months | Impella CP: 100% |
| IMP-IT Registry | Observational | 177 | 12 months |
Impella CP: 37,3%
Impella 2.5: 63% |
| German Impella Registry | Observational | 149 | 6 months |
Impella CP: 36,1%
Impella 2.5: 63,9% |
| PROTECT II Study | Randomized Controlled Trial | 225 | 3 months | Impella 2.5: 100% |
| PROTECT III Study | Observational | 504 | 3 months |
Impella CP: 31,9%
Impella 2.5: 68,1% |
| Azzalini et al. | PSM | 187 | 12 months | NA |
| Roma-Verona Registry | Observational | 86 | 3 months |
Impella CP: 14%
Impella 2.5: 86% |
Table 2
Baseline and procedural characteristics
| N = 1,581 | |
|---|---|
| Age, years old (IQR) | 70 (69-72) |
| Female, % (IQR) | 20(12-22) |
| HTA, % (IQR) | 85 (78-89) |
| Dyslipidemia, % (IQR) | 67(61-83) |
| Diabetes, % (IQR) | 46(44-51) |
| Previous CABG, % (IQR) | 14 (10–16) |
| Previous MI, % (IQR) | 44 (39-49) |
| PAD, % (IQR) | 26 (14–30) |
| HF, % (IQR) | 91 (54-98) |
| CKD, % (IQR) | 34 (25-38) |
| EUROSCORE, % (IQR) | 7.2(5.1-9) |
| SYNTAX SCORE, % (IQR) | 31 (29-35) |
| LVEF, % (SD) | 27 (6) |
| Left main, % (IQR) | 46 (39-65) |
| Three vessel disease, % (IQR) | 66 (31-92) |
| Atherectomy, % (IQR) | 27 (14-41) |
| CTO, % (IQR) | 14 (11-44) |
CABG = coronary aortic bypass grafting; CKD = chronic kidney disease; CTO = chronic total occlusion; HF = heart failure; HTA = hypertension; IQR = interquartile range; LVEF = left ventricular ejection fraction; MI = myocardial infarction; PAD = peripheral artery disease.
The overall risk of bias across the included studies was judged to be low to moderate, primarily due to potential confounding factors and patient selection in observational designs ( Figure S1 ). Funnel plot inspection revealed no significant evidence of publication bias ( Figure S2 ) for the primary endpoint.
Outcomes
The overall random-effects rate of all-cause of death at median 6 months follow-up was 13.4% (95% CI: 10.4 to 16.3) with a high degree of heterogeneity (see Figure 2 ).
All-cause mortality. Abbreviations: CI = confidence interval; IV = inverse variance, SE = standard error.
In exploratory subgroup analysis, studies in which >50% of patients presented left main disease had a pooled mortality of 18.2% (95% CI 17.97 to 18.07) compared with 11.6% (95% CI 9.49 to 13.71) in those with ≤50% (p <0.00001, I 2 = 97.2%). Similarly, studies with >50% multivessel coronary artery disease (CAD) showed numerically higher mortality (13.67%, 95% CI 9.62 to 17.71) compared with those with ≤50% (11.67%, 95% CI 8.67 to 14.66), although this difference was not statistically significant (p-value 0.44, I 2 = 0%). Finally, mortality did not significantly differ according to baseline ventricular function (LVEF <30%: 12.67%, 95% CI 7.41 to 17.93 vs ≥30%: 14.0%, 95% CI 10.99 to 17.01, p = 0.67, I 2 = 0%) (see Figure S3 ).
In-hospital secondary outcomes were as follows: 7.6% (95% CI 3.01 to 12.31) for acute kidney disease ( Figure S4 ), 6% (95% CI 2.9 to 9.25) for major bleedings ( Figure S5 ), 2.1% (95% CI 0.8 to 3.40) for vascular complications requiring surgery ( Figure S6 ).
6-months secondary outcomes were reported as incidence rates per 100 person-years and were as follows: 5.8% (95% CI 3.4 to 8.1, Figure 3 ) for myocardial infarction, 9.1% (95% CI 4.8 to 13.3, Figure 4 ) for repeat revascularization, 1.6% (95% CI 1.2 to 2.09, Figure 5 ) for stroke and 8.4% (95% CI 3.2 to 13.6, Figure 6 ) for heart failure rehospitalizations. Refer to Table 3 for a summary of the results.
Myocardial infarction. Abbreviations as in Figure 1 .
