Mechanical Circulatory Support Doesn’t Reduce Infarct Size in STEMI

The routine use of mechanical circulatory support with Impella CP starting 30 minutes prior to reperfusion with PCI fails to reduce infarct size compared with PCI alone in patients who have anterior STEMI without cardiogenic shock, according to results from the STEMI Door-to-Unload (DTU) randomized trial.

Use of Impella (Abiomed/Johnson & Johnson MedTech) in this patient population also was associated with a significantly higher risk of major bleeding and vascular complications when compared with PCI alone.

Despite promising preclinical data suggesting that LV unloading might reduce infarct size, Gregg Stone, MD (Icahn School of Medicine at Mount Sinai, New York, NY), who presented the results of the late-breaking clinical trial today during the American College of Cardiology 2026 Scientific Session, said Impella “is certainly not recommended at this time” in patients with anterior STEMI without cardiogenic shock.

There is “a lot of learning from this study that might influence the next trial and that’s what medicine is all about,” Stone told TCTMD. “Even when you have a negative randomized trial, you learn a lot from it, and that helps you design future research or improve patient care.”

Sanjit Jolly, MD (McMaster University/Population Health Research Institute, Hamilton, Canada), another STEMI-DTU investigator, said that while “Impella is great in cardiogenic shock and support, to ask it to reduce infarct size might be a little too much.”

Additionally, the “elephant in the room,” said Jolly, is the higher rate of major bleeding and vascular complications with Impella. “Had we shown a reduction in infarct size, inevitably, with that sort of device, you are going to see more vascular complications,” he said. “As a clinician, I’m going to continue using Impella in cardiogenic shock. In anterior STEMIs, I don’t think we’ll be using this technology.”

Even when you have a negative randomized trial, you learn a lot from it. Gregg Stone

David Kandzari, MD (Piedmont Heart Institute, Atlanta, GA), who wasn’t involved in the trial, said there has been a quest over decades to find a therapy to reduce myocardial infarct size that would translate into improved patient outcomes, including survival.

“There has been truly a graveyard of pharmaceutical and device therapies, all of which were intended to reduce infarct size as an adjunct to revascularization in acute MI,” he told TCTMD. At best, the studies are inconsistent, and none of the approaches have entered into standard of care.

Regarding use of Impella CP in this setting, the risk-benefit ratio is unequivocal, said Kandzari. “There is a sixfold higher incidence of major bleeding or vascular complications consistent with other studies using this technology,” he said. “But in this instance, it’s not countered by net clinical benefit with any efficacy, in this case using infarct size as a surrogate marker.”

While the trial closes the door on Impella in this patient population, Kandzari said, he stressed that STEMI-DTU investigators excluded patients with cardiogenic shock.

“We can’t equate these results with the demonstrable benefit that has been observed in shock patients who do undergo PCI and need the device, perhaps more for hemodynamic support rather than the intention of reducing infarct size,” said Kandzari.  

Moving to Patients Without Shock

The Impella CP is a temporary ventricular support device that is indicated for use during high-risk PCI performed in elective or urgent patients who are hemodynamically stable. It is also approved for the treatment of ongoing cardiogenic shock following acute MI or open-heart surgery or in the setting of cardiomyopathy or myocarditis as a result of isolated LV failure.

In 2024, the DanGer Shock study was the first large randomized, controlled trial to show that the routine use of mechanical support with the Impella CP microaxial flow pump resulted in fewer deaths compared with usual care in patients with acute MI complicated by cardiogenic shock.  

To ask [Impella] to reduce infarct size might be a little too much. Sanjit Jolly

The STEMI-DTU trial, which was published simultaneously in JACC, differed in that investigators randomized patients with anterior MI without shock. Myocardial salvage after timely PCI, the rationale goes, can remain suboptimal, which results in large infarct sizes that subsequently lead to a poor prognosis. A decade ago, Stone published a meta-analysis showing that infarct size, as measured by cardiac magnetic resonance (CMR) imaging or SPECT, was strongly associated with all-cause mortality or heart failure (HF) hospitalization at 1 year.

“With primary PCI, we’ve become very successful and efficient at opening the infarct-related artery,” said Stone. “We’re not as good at salvaging myocardium and we end up with large infarcts in a lot of patients, and that leads to death and heart failure.”

There’s been roughly five decades of experimental data showing that unloading the LV reduces wall stress and myocardial oxygen demand and attenuates myocardial ischemia by increasing microcirculatory support, said Stone. There are also preclinical data suggesting that LV unloading could reduce infarct size compared with reperfusion alone. These experimental studies showed that there is a critical waiting period after LV unloading that needs to occur before reperfusion.

“You have to do it 30 minutes before reperfusion,” said Stone. If reperfusion occurs within 15 minutes of LV loading, or LV unloading is performed after perfusion, there’s no benefit on infarct size, he added.

Unload the Left Ventricle

The STEMI-DTU study, performed at 55 sites, included 527 patients (mean age 61 years; 79.1% male) with anterior STEMI without cardiogenic shock randomized to either a strategy of LV unloading with the transvalvular microaxial flow pump for 30 minutes prior to PCI or PCI alone without mechanical LV unloading. Mean blood pressure was 143/89 mm Hg. The LAD was the location of the culprit lesion in more than 95% of patients.

There was no significant difference in key metrics, such as the times from symptom onset to hospital arrival and from hospital arrival to first device (either PCI or Impella insertion). The time from hospital arrival to PCI was a median of 40 minutes longer, while total ischemic time, defined as symptom onset to PCI, was 47 minutes longer in the Impella group. The median duration of mechanical circulatory support was 10.4 hours.

The primary endpoint—infarct size normalized to LV mass on CMR at 3 to 5 days—was not significantly different between the two treatment arms in the intention-to-treat or per-protocol analyses. In the intention-to-treat analysis, infarct size was roughly 1% smaller in those treated with Impella prior to PCI, and it was 1.9% smaller in the per-protocol analysis. The lack of effect on infarct size was consistent across all subgroups. There was no significant difference in any of the main secondary endpoints, including HF-related events and other clinical outcomes.

There has been truly a graveyard of pharmaceutical and device therapies, all of which were intended to reduce infarct size. David Kandzari

Stone said that clinicians all know that “time is muscle,” with every minute delay in reperfusion increasing infarct size. Still, despite delaying reperfusion by roughly 45 minutes once the microaxial flow pump was inserted, infarct size was not increased; instead, it was 1-2% smaller.

“It suggests that the device is doing something to reduce infarct size, but it may not have been as effective as it could have been,” he said.

Regarding safety, which was assessed at 30 days in the 260 patients in whom Impella insertion was attempted, treatment-related major bleeding (BARC types 3 to 5) or major vascular complications exceeded the predefined performance goal of 26.5%. Treatment-related major bleeding occurred in 30.4% of patients, all of which were related to the access site. Eleven patients (4.2%) had treatment-related major vascular complications, mostly hematomas.

“It did not meet an acceptable level of device-related major bleeding or vascular complications,” said Stone. “There were substantially higher rates of both.”

While bleeding is related to an increased risk of mortality after STEMI, Stone said that non-access-site bleeding carries greater risks and there was no difference in this measure between the two groups. Additionally, he noted that while major bleeding or major vascular events were associated with 12-month mortality in the control group, this was not so in the Impella arm.

What Happens Next?

To TCTMD, Kandzari noted that use of Impella in acute MI without cardiogenic shock has been debated since the pilot study published in 2019. The trial was small, with just 50 patients, but there was no signal that delayed reperfusion after 30 minutes of unloading with Impella CP reduced infarct size. While the study showed that delaying reperfusion was safe, Kandzari said the lack of efficacy had some questioning whether pursuing a larger trial was worthwhile.

“Certainly, the present trial now confirms the absence of benefit and added risk,” he said.

As for why the device had no effect on infarct size, investigators can only speculate, but Stone believes blood pressure might have been the reason. Impella, he said, is most efficient when blood pressure is low. At 140 mm Hg, the device has less impact on increasing cardiac output. In essence, Impella rests the ventricle less when blood pressure is that high. There are plans for a similar type of trial in the same patient population, but with the use of intravenous nitroglycerin and beta-blockade to bring blood pressure down before Impella is inserted, he said.

Jolly, to TCTMD, said that he understands the explanation around blood pressure, but countered that it might also be that these patients simply don’t require mechanical circulatory support. It’s also possible that if investigators focus on patients who have low blood pressures, but who don’t quite meet the definition of cardiogenic shock—borderline patients, perhaps—it might prove effective.

“There are many areas of medicine where you take the sickest patients and provide the intervention and it shows a benefit,” he said. “Then you go after the next tier.”  

Kandzari also noted that Impella is more effective in patients with dysfunctional hemodynamics. From a “30,000-foot” perspective, he added that it’s been more than 10 years since Impella was approved in the US for cardiogenic shock and it’s only in the last few years that randomized, controlled trials have helped inform clinical indications and appropriate patient selection.

“The positive out of the study is that even though it’s a negative trial, it’s showing us at least one indication where this technology is not yielding any benefit,” said Kandzari. He also praised the investigators for completing the trial, noting there are inherent challenges in conducting such research in acute STEMI.