Key Points:
- Acute myocardial infarction is frequently complicated by heart failure, with outcomes worsening alongside any increase in infarct size. It is hypothesized that early left ventricular unloading with Impella may reduce infarct size and subsequently improve clinical outcomes.
- The STEMI-DTU trial was designed to examine the utility of pre-PCI unloading with Impella for patients presenting with acute anterior STEMI.
- There was no difference in the primary endpoint of % infarct size at 3-5 days between Impella unloading and standard of care treatment, but patients treated with Impella had a very high rate of device-related or vascular complication (30.8%0, above the safety endpoint of 26.5%.
Even after timely reperfusion, heart failure (HF) complicates a high proportion of acute myocardial infarctions (MIs), with longitudinal outcomes worsening by 20% with every 5% increase in infarct size. While immediate revascularization continues to be the standard of care for ST-elevation MI (STEMI), preclinical data have suggested that up-front “unloading” of the left ventricle may mitigate myocardial ischemia, largely via increased coronary perfusion and reduced oxygen consumption. As such, a new conceptual framework known as the “door to unload” time has been developed, which postulates that potentially more myocardium could be salvaged if PCI was preceded by a period of ventricular unloading with an Impella device. In a breaking presentation at the 2026 ACC Scientific Sessions today, Dr. Gregg Stone and his team tested this framework, presenting their study: “The STEMI Door to Unload Trial,” or the STEMI-DTU study, which was simultaneously published in JACC.
The STEMI-DTU trial (NCT03947619) was a multi-center trial conducted across 55 sites in the US, Europe, and Canada that randomized 527 adults (age 18-85) presenting with an acute anterior STEMI with indication for PCI to immediate Impella CP placement vs standard of care PCI, with no delay. Prior to randomization, all patients received an iliofemoral angiogram and LV-gram (with LVEDP measurement) in order to rule out any contraindication to Impella and confirm the presence of an anterior wall motion abnormality. Patients were required to be a) presenting with their first MI and to b) randomized within 1-6 hours of symptom onset. Other relevant exclusion criteria were cardiogenic shock, overt mechanical complication of an MI, inferior or RV involvement, or prior heart failure, aortic valve disease, or revascularization with CABG or LAD PCI. The primary outcome was infarct size on day 3-5, as measured by %LV mass on cardiac MRI. The key secondary effectiveness endpoint was a hierarchical composite of CV death at ≥12 months, cardiogenic shock at ≥24 hours from enrollment, LVAD or heart transplant at ≥12 months, heart failure at ≥12 months, ICD or CRT placement at ≥12 months, and infarct size on day 3-5 cardiac MRI. Among those with an attempted or implanted Impella, the key safety endpoint was device- or procedure-related major bleeding or major vascular complications at 30 days.
A total of 265 individuals were randomized to standard caremand 262 to the Impella group. The mean age was 61, with 79% men. On presentation, the average MAP was 107, with a HR of 80, lactate of 2.0, and LVEDP of 25. The culprit lesion was the LAD in 98% of cases. As a result of the mandated 30-minute unloading period, the treatment group had a median of 47 minute longer total ischemic time and a 42 minute longer door-to-balloon time. In the intention-to-treat analysis, there was no significant difference in the primary outcome between the early Impella and standard of care arm (an infarct size of 30.8 ± 16.2 vs 31.9 ± 16.9%, p=0.5). Similarly, there was no difference in the secondary effectiveness endpoint (win ratio 1.04, 95% CI 0.84-1.28; p=0.73). In further secondary analyses, all-cause and CV mortality were similarly not significantly different between the two groups at a year (p>0.05), despite some early visual evidence of separation in the survival curves at 30 days (trended towards significance; HR 0.33; 95% CI 0.09-1.22, p=0.08 for both). In pre-specified subgroup analyses, age>61 was associated with a 4% reduction in infarct size compared to those under 61, with an interaction p-value of 0.052, which also trended towards significance. For the safety endpoint, the treatment arm did not meet the pre-specified performance goal of 26.5%; rather, 30.8% of patients with Impella attempted or implanted experienced a device-related major bleed or major vascular complication.
Important limitations of the study included high rate of femoral artery access for PCI (as driven by protocol), high baseline BP with minimal evidence of shock (baseline lactate ~2), and the potential for suboptimal closure, which may have contributed to higher-than-expected vascular complications. When discussing the clinical implications of the study at the Scientific Sessions, Dr. Stone stated: “The combination of Impella CP plus a 30-minute delay to PCI did not reduce infarct size compared with immediate PCI in patients with anterior STEMI without shock, despite a substantial delay to reperfusion after initiating LV support. Major bleeding or vascular complications occurred more frequently in the treatment group when compared with a prespecified performance goal or the control group…future potential studies will incorporate insights from subgroup analysis and may combine Impella with adjunct pharmacotherapy to optimize loading conditions before PCI.”