CHIP-BCIS3
Percutaneous Left Ventricular Unloading During High-Risk Coronary Intervention in Patients With Severe LV Dysfunction
Presented by Divaka Perera — King's College London
Subspecialty: Interventional
Published in NEJM
Key Result
LV unloading with Impella CP during complex PCI did NOT reduce adverse events in severe LV dysfunction (win ratio 0.85, p=0.30) and was associated with HIGHER all-cause mortality (32.6% vs 23.4%, HR 1.54) and CV death (26.7% vs 14.5%).
What did this trial find?
The CHIP-BCIS3 trial randomized 300 patients with severe LV dysfunction (LVEF ≤35%) undergoing complex PCI to elective LV unloading with Impella CP or standard care. LV unloading did not reduce the primary composite endpoint (win ratio 0.85, p=0.30) and was associated with higher all-cause mortality (32.6% vs 23.4%) and cardiovascular death (26.7% vs 14.5%). The results have generated significant expert reaction, with the trial's lead investigator and invited discussant both calling for a halt to routine use of Impella in this setting, and editorialists urging a more selective approach to mechanical circulatory support in high-risk PCI.
Why does this trial matter?
Real controversy with meaningful debate. This trial directly challenges the rapidly growing 'protected PCI' paradigm with Impella, showing not just no benefit but a strong signal of harm (nearly double the CV death rate). The lead investigator, invited discussant, and NEJM editorialists all converge on the same conclusion — routine use cannot be recommended — but the finding that the device may actively cause harm rather than just being futile adds genuine clinical urgency. The contrast with the positive DanGer Shock trial in cardiogenic shock creates an important nuance in the discussion.
Study Design
Prospective, open-label, randomized trial across 21 UK sites
Clinical Implications
Routine use of LV unloading with Impella CP during high-risk PCI in patients with severe LV dysfunction cannot be recommended. The findings encourage a more selective approach to mechanical circulatory support, particularly in the absence of clear hemodynamic instability.
Abstract
The CHIP-BCIS3 trial was a prospective, open-label trial that randomized 300 patients with severe left ventricular dysfunction (LVEF <=35%) and extensive coronary disease across 21 UK sites to either elective LV unloading with a microaxial flow pump (Impella CP, n=148) or standard care (n=152) during complex PCI. Mean age was 73 years, 17% were women, median LVEF was 27%, and median SYNTAX score was 38. At median 22-month follow-up, the primary hierarchical composite (all-cause death, disabling stroke, spontaneous MI, CV hospitalization, periprocedural myocardial injury) occurred in 79.3% of the pump group vs 73.6% of standard care (win ratio 0.85, p=0.30). All-cause mortality was 32.6% vs 23.4% (HR 1.54); cardiovascular death was 26.7% vs 14.5%. Periprocedural myocardial injury occurred in 61.7% vs 50.0%. No between-group differences in bleeding or vascular complications.