In patients with severely depressed left ventricular (LV) function, percutaneous coronary intervention (PCI) may result in hemodynamic instability, leading to suboptimal revascularization and increased complications. Although there is a clear evidence of mechanical circulatory support (MCS)-supported PCI in selected patients with cardiogenic shock, continued efforts are being made to understand the potential benefits of MCS-supported PCI in populations with severely depressed LV function not in shock. Despite a lack of clear supporting evidence, the use of MCS for high-risk PCI has grown exponentially, particularly in the United States (US) from microaxial flow pumps (Impella)– from 60/10,000 PCIs in 2012 to 443/10,000 in 2021– sevenfold increase. , Given the increased costs and complications associated with the use of Impella during PCI, a better understanding of which populations may derive benefit is needed.
In a recent issue of The American Journal of Cardiology , Panoulas and colleagues provide a statistically contemporary meta-analysis of analyzed data from the PROTECT-II (P-II) and the PROTECT-III (P-III) studies. P-II was a randomized controlled trial of high-risk PCI comparing Impella versus intra-aortic balloon pump (IABP) support that enrolled subjects between November 2007 and December 2010. P-III was an Impella-only high-risk PCI cohort that enrolled subjects between March 2017 and March 2020. Novelly, Panoulas et al employed a hierarchical 5-component (death, stroke, spontaneous myocardial infarction [MI], rehospitalization, and periprocedural MI) Win Ratio (WR) to compare Impella-supported, from P-II and “P-II like” subjects from P-III versus IABP-supported patients from P-II. They also looked at 3 subgroups ([1] patients from P-II prespecified to atherectomy, unprotected left main [ULM], chronic total occlusion [CTO, 2] all patients excluding the first patients from the P-II cohort [to exclude learning curve of Impella use], and [3] separated P-II and P-III cohorts).
A total of 211 IABP and 719 Impella-supported patients (503 from P-III and 216 from P-II) were included in the analysis. The only statistically significant demographic difference was age, with an older cohort for the Impella group (68.8 vs 66.9, p = 0.034). Ultimately, Panoulas and colleagues found a WR in favor of Impella in the primary group and the subgroups. However, per the nonhierarchical analysis of the data, this appears to be driven entirely by an increased rate of rehospitalization (39.3% vs 24.6%) in the IABP group. This suggests that Impella-supported PCI may be associated with a decreased rate of rehospitalization compared to IABP-supported PCI. Unfortunately, likely due to constraints of the available data from the original cohorts, this study does not specify the drivers of rehospitalizations in addition to other clinical outcomes such as vascular complications, length of stay, and procedural costs. An additional concern with combining data from P-II and P-III is the temporal differences of the cohorts. PCI techniques, including the increased use of intravascular imaging and physiological testing, evolved between the 2 trials, as did the armamentarium of Guideline-Directed Medical Therapy (GDMT). Also of note, the REVIVED-BCIS2 study randomized patients with ischemic LV dysfunction to PCI versus GDMT alone. At a median follow-up of 41 months, there were nearly identical rates of a combined primary endpoint of all-cause death and hospitalization for heart failure. However, this study did not test the efficacy of MCS for PCI.
Panoulas and colleagues work is timely considering the recently published CHIP-BCIS3 Trial. This was a randomized controlled trial, which also utilized WR in comparing 300 patients with severe LV dysfunction and extensive coronary disease to unloading with Impella CP (n = 148) versus standard care (no upfront MCS, including IABP) (n = 152) during complex PCI. The trial used a hierarchical composite outcome (death, disabling stroke, spontaneous, periprocedural MI, and cardiovascular hospitalization) that was analyzed as a WR. At a median of 22 months, there was a statistically insignificant trend in the WR favoring standard care over Impella and a trend toward increased death with Impella. The ongoing PROTECT IV (P-IV) trial comparing Impella-supported high-risk PCI versus standard care will hopefully shed light on these findings.
The increasing use of WR in cardiovascular studies has been driven by a need to create a hierarchical composite of event outcomes and quantitative measures that more accurately weigh clinical priorities. Despite the limitations of the current study, the authors have demonstrated a novel way to use WR analysis to merge trials that were conducted separately. However, the results using such a technique can only be considered hypothesis generating.
CRediT authorship contribution statement
Ruchi Patel: Conceptualization, Data curation, Formal analysis, Validation, Writing– original draft, Writing– review & editing. Daniel R. Schimmel: Conceptualization, Data curation, Formal analysis, Supervision, Validation, Writing– review & editing. James D. Flaherty: Conceptualization, Data curation, Formal analysis, Supervision, Validation, Writing– review & editing.
Funding: None.
References
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
