BCIS-3 Finds No Benefit, Signals Harm With LV Unloading in PCI

March 30, 2026—Use of a temporary mechanical pump to achieve left ventricular (LV) unloading during percutaneous coronary intervention (PCI) did not improve outcomes compared with PCI alone in patients with severe coronary artery disease and impaired LV function, according to results of the BCIS-3 trial presented at ACC.26 and published in The New England Journal of Medicine.

The randomized trial addresses a growing practice pattern in complex PCI, where adjunctive hemodynamic support has been adopted despite limited randomized data. The findings suggest no clinical benefit and raise safety concerns, particularly regarding mortality, in this high-risk population.

BCIS-3 enrolled 300 patients across 21 United Kingdom sites (mean age, 73 years; 83% men). Participants had extensive coronary disease and severe LV dysfunction, with a mean ejection fraction of 27%. Approximately 75% presented with acute coronary syndrome. Patients were randomized to PCI with LV unloading using a microaxial flow pump (Impella CP, Johnson & Johnson MedTech) or PCI alone.

The primary endpoint—a composite of all-cause death, disabling stroke, myocardial infarction, cardiovascular hospitalization, or procedural myocardial injury—did not differ significantly between groups. Pairwise comparisons favored standard PCI in 43% of cases versus 36.6% favoring LV unloading, with no difference in 20.4% (win ratio, 0.85; 95% CI, 0.63-1.15; P = 0.30), corresponding to a between-group difference of −6.4 percentage points (95% CI, −18.4 to 5.7).

In sensitivity analyses, results were consistent. The composite endpoint, including periprocedural myocardial injury, occurred in 79.3% versus 73.6% (hazard ratio [HR], 1.24; 95% CI, 0.94–1.62), and excluding procedural injury, in 45.3% versus 45.4% (HR, 1.06; 95% CI, 0.75-1.49).

However, secondary analyses showed higher mortality with LV unloading. All-cause death occurred in 32.6% of patients receiving LV unloading versus 23.4% with standard care (HR, 1.54; 95% CI, 0.99-2.41). Cardiovascular death was also higher (26.7% vs 14.5%; HR, 1.91; 95% CI, 1.11-3.30), representing an absolute increase of 12.2%.

“We found no evidence that use of the temporary pump protected the heart during the angioplasty procedure,” said Divaka Perera, MD, with King’s College London, in an announcement from the American College of Cardiology. “Our findings strongly suggest that we shouldn’t be using this device routinely without more evidence of benefit.”

Patients treated with LV unloading also experienced more myocardial injury, counter to the proposed mechanism of ventricular protection (61.7% vs 50%; risk ratio [RR], 1.23; 95% CI, 0.99-1.54). Rates of bleeding and vascular complications were numerically higher but not significantly different between groups, including major bleeding (10.8% vs 7.3%; RR, 1.48; 95% CI, 0.71-3.09) and minor vascular complications (15.5% vs 9.9%; RR, 1.56; 95% CI, 0.85-2.88). No differences were observed in spontaneous myocardial infarction (HR, 0.64; 95% CI, 0.28-1.47), disabling stroke (HR, 0.53; 95% CI, 0.13-2.11), or hospitalization for cardiovascular causes (HR, 1.20; 95% CI, 0.72-1.98).

Median follow-up was 22 months (IQR, 16-30). Rates of bleeding and other blood vessel injuries were overall low, suggesting these did not account for the observed mortality difference.

The study excluded patients with cardiogenic shock and was conducted entirely in the United Kingdom, which may limit generalizability. Investigators noted ongoing analyses to better understand the mechanisms underlying the observed harm and to assess cost-effectiveness.

The trial was funded by the United Kingdom’s National Institute for Health & Care Research.