CHIP-BCIS3 — LV Unloading in High-Risk PCI

1. Publication

• Title: Left Ventricular Unloading in High-Risk Percutaneous Coronary Intervention
• Acronym: CHIP-BCIS3
• Year & Journal: New England Journal of Medicine, epublished March 29, 2026 (print: 2026;394(18):1779-1789)
• Citation: Perera D, Ryan M, Ezad SM, et al; CHIP-BCIS3 Investigators. Left Ventricular Unloading in High-Risk Percutaneous Coronary Intervention. N Engl J Med. 2026;394(18):1779-1789. doi:10.1056/NEJMoa2515704

2. Context & Rationale

Background: Complex, high-risk PCI in patients with severely impaired LV function carries substantial risk of periprocedural cardiogenic shock, which can both harm patients and limit revascularization completeness. Percutaneous LV unloading (microaxial flow pump, e.g., Impella) reduces cardiac work and pulmonary capillary wedge pressure while improving cardiac power output — favorable physiology — but whether this translates into better clinical outcomes had not been robustly tested despite rapidly increasing real-world use (driven partly by reimbursement incentives in the US, Germany, Japan).

Research Question/Hypothesis: In patients with severely impaired LV function undergoing complex PCI, does elective LV unloading with a microaxial flow pump reduce major adverse clinical outcomes compared with standard care?

Why This Matters: More than 75% of real-world microaxial flow pump implants occur in the high-risk (not cardiogenic shock) PCI setting, despite the device's evidence base being strongest in shock; this trial directly tests the much more common, less-validated use case.

3. Design & Methods

• Study Type: Randomized controlled trial
• Setting & Centers: 21 NHS hospitals, United Kingdom
• Population:
• Inclusion: LV ejection fraction ≤35%, extensive CAD (BCIS Jeopardy Score ≥8/12), planned complex PCI (left main lesions, extensive calcium modification, or retrograde CTO)
• Exclusions: Not detailed in available trial text
• Intervention: Elective LV unloading with a microaxial flow pump (Impella CP) prior to high-risk PCI
• Comparator: Standard care (PCI without elective mechanical LV support)
• Blinding: Not detailed in available trial text (device-based intervention); primary outcome events adjudicated by a clinical events committee
• Statistical Power & Follow-Up: Sample size 250 patients planned for >80% power to detect HR 0.62 at minimum 12 months (assuming 150 events across follow-up); 300 patients actually enrolled (148 microaxial pump, 152 standard care). Primary outcome: hierarchical composite of all-cause death, disabling stroke, spontaneous MI, cardiovascular hospitalization, and periprocedural myocardial injury, analyzed by win ratio. Median follow-up 22 months.

4. Key Results

Study population severity: Mean age 73y, 83% male, median LVEF 27% (IQR 20–32), BCIS Jeopardy Score 12 (IQR 10–12), SYNTAX score 38 (IQR 30–47), 76% presenting with acute coronary syndromes.

5. Internal Validity Assessment

• Randomization & Allocation: Randomized across 21 NHS centers; specific allocation-concealment mechanics not detailed in available trial text.
• Protocol Adherence & Separation: Device delivered as an elective procedural intervention; adherence details not fully available, but the hierarchical composite and its components were adjudicated by an independent clinical events committee.
• Blinding & Detection Bias: Device-based intervention is inherently difficult to blind for proceduralists; mortality and MI are relatively objective, reducing (but not eliminating) detection-bias risk.
• Missing Data & Sensitivity Analyses: Time-to-first-event sensitivity analyses (with and without periprocedural MI) were directionally consistent with the primary win-ratio analysis, strengthening confidence in the overall neutral-to-harmful signal.
• Overall Internal Validity Conclusion: Moderate-to-strong — a modestly sized trial (300 patients) but with independent event adjudication, a prespecified hierarchical analysis, and consistent sensitivity analyses; the safety signal (excess CV mortality) is statistically significant despite the trial not being primarily powered for mortality alone.

6. External Validity Assessment

• Population Representativeness: Very high-risk population (mean LVEF 27%, BCIS-JS 12, SYNTAX 38) — representative of complex, elective/urgent high-risk PCI candidates being considered for mechanical support, not a general PCI population.
• Practice Context: Single-country (UK) trial; findings most directly applicable to systems performing elective LV-unloading-supported complex PCI in similar high-risk, non-shock populations.
• Overall External Validity Conclusion: Good for the specific high-risk, non-shock complex PCI population where elective mechanical support is currently used; does not address (and should not be extrapolated to) cardiogenic shock PCI, where the device's evidence base differs.

7. Strengths & Limitations

Strengths:

• Directly tests the much more common (but less validated) elective high-risk PCI use case, as distinct from cardiogenic shock
• Independent clinical events committee adjudication
• Prespecified hierarchical win-ratio primary analysis plus consistent time-to-event sensitivity analyses
• Addresses a genuine evidence gap despite years of increasing real-world device use

Limitations:

• Modest sample size (300 patients) relative to the mortality signal detected
• Single-country (UK) trial
• Open device-based intervention, inherently unblindable for proceduralists
• Trial not primarily powered to definitively resolve the mortality/spontaneous-MI question alone (explicitly acknowledged in the trial's own design rationale)

8. Interpretation & Practice Impact

• Clinical Implications: Does not support routine elective use of microaxial flow pumps for high-risk (non-shock) complex PCI; the excess cardiovascular mortality signal raises a genuine safety concern rather than simple lack of benefit.
• Mechanistic Coherence: The dissociation between favorable device physiology (reduced cardiac work, improved cardiac power output) and worse clinical outcomes suggests procedural/vascular complications and periprocedural risk may outweigh the theoretical hemodynamic benefit in this elective, non-shock population.
• Systems-Level Takeaway: Per accompanying editorial commentary, supports a more selective approach to elective mechanical circulatory support in high-risk PCI, particularly in the absence of clear hemodynamic instability — not routine/reflexive use.

9. Controversies & Subsequent Evidence

• Editorial Commentary: Nallamothu and Wanamaker's accompanying NEJM editorial ("Price of protection") stresses that in the cath lab, "the margin between a good outcome and a catastrophic one is narrow, and the instinct to add protection with mechanical circulatory support can be powerful" — but the trial's data argue for more selective, not routine, use.
• Guideline Integration: This directly challenges prior UK NICE Interventional Procedures Guidance (IPG633, 2018), which permitted use within specialized centers while flagging limited efficacy evidence and known safety concerns — CHIP-BCIS3 now provides the randomized efficacy/safety data NICE's guidance had been awaiting. Contrasts with DanGer-Shock (a related, industry-funded RCT in cardiogenic shock, a different population where the risk-benefit calculus may differ).

10. Summary & Executive Takeaway

Summary: CHIP-BCIS3 randomized 300 patients with severe LV dysfunction (median LVEF 27%) undergoing complex PCI to elective microaxial flow pump support or standard care. The primary hierarchical composite outcome numerically favored standard care (36.6% vs 43.0% wins; win ratio 0.85, 95% CI 0.63–1.15, P=0.30). All-cause mortality trended higher with the pump (32.6% vs 23.4%, HR 1.54, P=0.054) and cardiovascular mortality was significantly higher (26.7% vs 14.5%, HR 1.91, 95% CI 1.11–3.30).

Overall Takeaway: Elective LV unloading during high-risk (non-shock) complex PCI did not improve outcomes and showed a concerning signal of excess cardiovascular mortality — despite widespread and increasing real-world adoption, this trial argues for a substantially more selective approach to mechanical circulatory support in this population, reserving it for clear hemodynamic instability rather than routine elective prophylaxis.

11. Bibliography

• Ryan M, Ezad SM, Webb I, et al; CHIP-BCIS3 Investigators. Percutaneous left ventricular unloading during high-risk coronary intervention: rationale and design of the CHIP-BCIS3 RCT. Circ Cardiovasc Interv. 2024;17(3):e013367.
• Nallamothu BK, Wanamaker BL. Price of protection — microaxial flow pump in high-risk PCI [editorial]. N Engl J Med. 2026.
• Kapur NK, Mangner N, Aghili N, et al. Left Ventricular Unloading in Anterior STEMI without Shock: The STEMI Door to Unload (DTU) RCT. JACC. 2026.