The Device That Scaled Before It Was Tested
Experimental studies show that for a microaxial heart pump to reduce damage during a heart attack, doctors would need to insert it 30 minutes before performing the stent procedure. That timeline is physically impossible, patients arrive at hospitals mid-crisis, not half an hour before symptoms begin. Yet this device, the Impella pump, generated $1.75 billion in sales for Johnson & Johnson in 2025 and has FDA approval for high-risk cardiac procedures, despite never being tested in a randomized trial against standard care until now.
Two major trials presented March 28, 2026 at the American College of Cardiology Scientific Sessions and published simultaneously in the Journal of the American College of Cardiology reveal what happens when that evidence finally arrives: patients treated without the pump either do the same or better than those who receive it, while nearly one in three pump patients suffer major bleeding or vascular complications from the device itself.
A Life-Saving Tool in the Wrong Hands
The Impella pump is genuinely life-saving for patients in cardiogenic shock, a condition where the heart fails so catastrophically it cannot pump enough blood to sustain the body's organs. For these dying patients, the catheter-based device that pulls blood from the heart and pushes it into the aorta provides critical mechanical support. Dr. Gregg W. Stone, Professor of Medicine at the Icahn School of Medicine at Mount Sinai and co-Principal Investigator of the STEMI-DTU trial, confirmed the pump's value in this specific population.
But cardiogenic shock represents a small fraction of heart attack patients. The broader question, whether the pump helps patients undergoing high-risk percutaneous coronary intervention without shock, had never been answered in a randomized trial before CHIP-BCIS3. That study enrolled 300 patients and compared outcomes between those who received mechanical circulatory support during PCI and those who underwent the procedure without it.
The results contradicted the device's widespread adoption. Patients without a pump fared better in 43% of paired comparisons. There was no difference between groups in 20.4% of cases. Most strikingly, 47 patients in the pump group died compared to 33 deaths among those who received PCI alone, 14 additional deaths associated with the device meant to protect them.
Testing What's Already Standard Practice
The STEMI-DTU trial focused on a different but related question: whether the Impella CP device could limit heart muscle damage in patients suffering from STEMI, the most severe type of heart attack, caused by sudden blockage of a major coronary artery, who were not in cardiogenic shock. These patients face serious danger but haven't yet deteriorated to the point where their hearts cannot sustain circulation.
Heart muscle damage occurred in 30.8% of patients who received the pump versus 31.9% in the control group, statistically indistinguishable outcomes. But the device introduced new problems: 30.8% of pump patients experienced major bleeding or blood vessel complications within 30 days. The trial had set a safety threshold of 26.5% for these complications; the pump exceeded it.
The temporal impossibility Stone and colleagues identified in experimental studies explains why. For the pump to reduce infarct size, it must be inserted 30 minutes before the stenting procedure begins. In clinical reality, patients arrive at emergency departments during active heart attacks. The window for intervention is measured in minutes, not the half-hour head start required for the device to condition the heart muscle against damage.
Stone noted that almost all strategies developed to reduce muscle damage during cardiac stenting have failed. The pattern was predictable, yet the device scaled to $1.75 billion in annual revenue before anyone conducted the definitive test.
The $16.6 Billion Bet on Untested Technology
Johnson & Johnson acquired Abiomed, the manufacturer of Impella pumps, for $16.6 billion in 2022. At the time of purchase, the device had FDA approval for use in high-risk PCI procedures and a reputation as cutting-edge technology. What it didn't have was randomized trial data showing it actually improved outcomes in these patients compared to doing nothing.
The regulatory pathway for medical devices differs fundamentally from drug approval. Pharmaceutical companies must complete Phase III randomized controlled trials demonstrating efficacy before the FDA allows widespread marketing. Devices often reach patients through alternative routes, approval for one narrow indication that then expands through clinical practice, or clearance based on similarity to existing devices rather than independent evidence of benefit.
The Impella pump followed this pattern. Its proven value for cardiogenic shock patients, a legitimate, evidence-based use, created momentum for adoption in adjacent cases. Doctors facing high-risk PCI procedures had an FDA-approved device that seemed protective, even life-saving. Without randomized trials showing otherwise, the clinical logic appeared sound: if it helps the sickest patients, why not use it for sick patients who aren't quite as critical?
The answer arrived four years after J&J's acquisition. The device doesn't help these patients. In some comparisons, it appears to harm them. And it introduces a 30.8% risk of serious complications that wouldn't exist without the intervention.
When Financial Momentum Outpaces Scientific Evidence
An editorial accompanying the CHIP-BCIS3 article, authored by Brahmajee Nallamothu and Brett Wanamaker, examined the implications of testing a device after it has become standard practice. The situation reveals a structural problem: the system allowed widespread use, a multi-billion dollar acquisition, and massive revenue generation before anyone checked whether the device worked in these applications.
The Impella story exposes how medical devices can scale on approval creep rather than evidence. A device approved for Condition A, where it demonstrably saves lives, migrates to Conditions B and C through clinical judgment and market forces. Each expansion seems reasonable in isolation, these patients are also critically ill, the device is already approved, why not try it? But without randomized trials, no one knows whether these incremental expansions help or harm patients.
By the time researchers design and complete the studies that should have preceded widespread adoption, the device has become embedded in practice patterns and corporate portfolios. The CHIP-BCIS3 and STEMI-DTU trials represent science catching up to a market reality that developed in an evidence vacuum.
The Billion-Dollar Question
The Impella pump generated $1.75 billion in sales in 2025, four years after its core applications were finally tested in randomized trials. The trials showed no benefit and potential harm for patients outside cardiogenic shock. The device's expansion beyond its evidence-based use case, mechanical support for patients in shock, followed financial and clinical logic that seemed reasonable without contradictory data.
That logic has now been tested. Patients undergoing high-risk PCI without cardiogenic shock don't benefit from mechanical circulatory support, and nearly one in three suffer major complications from the device itself. The pump works for the patients it was designed to help. For everyone else, it's a solution to a problem that doesn't exist, deployed at scale before anyone checked.
The question facing regulators and hospitals is how many other billion-dollar devices are running on the same model, approval for one indication, expansion through practice, revenue before evidence. The Impella trials provide an answer for one device. The system that allowed this trajectory remains unchanged.