Clinical Question

• In adults presenting with STEMI and cardiogenic shock does the use of a microaxial flow pump (Impella CP) compared to standard care reduce death from any cause at day 180?

Background

• Cardiogenic shock is a frequent complication of STEMI and is associated with significant morbidity and mortality
• Microaxial flow pumps (MAFP) are percutaneously inserted devices that pump blood out from the left ventricle into the ascending aorta
• Prior RCTs have been small and not shown any benefit with a signal towards some harm (bleeding) in registry studies

Design

• International, multi-center RCT
• Open label
• Randomised via web-based system and stratified by localization of STEMI and timing relative to revascularisation
• Randomisation occurred in the catheter laboratory either before or after revascularisation (could occur up to 12 hours after leaving) depending on when cardiogenic shock was recognised
• Sample size calculation
  • 360 patients required to detect a reduction from 60% to 42% in 180 day mortality with 80% power at an alpha of 0.05
• Pre-defined subgroups
• Registered with clinicaltrials.gov 

Setting

• 14 centers in Denmark (4 centers), Germany (9) and UK (1)
• January 2013 to July 2023

Population

• Inclusion:
  •  ≥ 18 years old
  • STEMI
    • STEMI equivalent also allowed
  • Cardiogenic Shock
    • SBP < 100mmHg +/- need for ongoing vasopressor support AND lactate > 2.5 mmol AND LV Ejection Fraction < 45%
• Exclusion:
  • Resuscitated from out-of-hospital cardiac arrest who remained comatose (GCS < 8)
  • Severe right ventricular failure (RV larger than LV, TAPSE < 1cm or septal shift)
  • Shock duration > 24 hours
  • Other causes of shock or shock due to mechanical complication of MI
  • Severe aortic regurgitation or stenosis
  • Severe arterial disease precluding placement
• 1211 screened → 360 enrolled
  • 851 excluded or which only 39 were eligible and not randomised
  • 5 excluded following randomisation as consent unable to be obtained
• Comparing baseline characteristics of MAFP vs. standard care group
  • Well balanced
  • Age: 67 vs 69
  • Male: 79 vs 79%
  • SBP: 84 vs 82 mmHg
  • Lactate: 4.6 vs 4.5 mmol/L
  • LVEF: 25 vs 25%
  • Resuscitation prior to randomisation: 22 vs 19%
  • Anterior MI: 70 vs 73%
  • SCAI
    • C: 56 vs 55%
    • D: 29 vs 28%
    • E: 16 vs 17%
  • Number of disease vessels on angiography:
    • 1: 29 vs 27%
    • 2: 39 vs 36%
    • 3: 31 vs 37%
  • Time from symptom onset to randomisation: 4.8 vs 3.8 hours
  • Randomised prior to revascularisation: 55 vs 58%
• In-hospital management of cardiogenic shock largely similar:
  • PCI: 96 vs 98%
  • Time from admission to balloon inflation: 58 vs 45 mins
  • Placement of Impella CP: 95 vs 2%
  • Median duration of mechanical ventilation: 5 vs 3 days
  • Vasoactive use:
    • Noradrenaline: 87 vs 81%
    • Dopamine: 29 vs 23%
    • Adrenaline: 37 vs 38%
    • Dobutamine: 35 vs 34%
    • Milrinone: 35 vs 33%
    • Levosimendan: 22 vs 22%
  • Median duration of ICU admission: 6 vs 3 days
  • Median duration of hospital admission: 12 vs 7 days

Intervention

• Insertion of MAFP
  • Inserted immediately after randomisation
  • Run at the highest performance level for 48 hours unless complications occurred
  • Criteria and guidance for weaning provided
  • Median duration of MAFP support: 59 hours
  • 8 did not receive a MAFP

Control

• Standard Care – no use of a MAFP
  • 3 crossed over to MAFP

Management common to both groups

• All participants underwent a revascularisation procedure as needed
• In event of haemodynamic instability treatment could be escalated
  • In microaxial group this was Impella 5.0 or Impella RP or ECMO
  • In control group ECMO was recommended but Impella 5.0 allowed
    • If an Impella was used in this manner this was defined as a protocol violation
  • This occurred in 16% in microaxial pump group and 21% in standard care group (largely to VA ECMO)

Outcome

• Primary outcome: Death from any cause at 180 days
  • 45.8 (microaxial flow pump) vs 58.5% (standard care)
    • HR 0.74 (95% CI 0.55 to 0.99)
    • NNT = 8
• Secondary outcomes:
  • Composite cardiac end point (escalation to additional mechanical support, heart transplantation or death)
    • 53 vs 64%; HR 0.72 (95% CI 0.55 to 0.95)
  • Number of days alive and out of hospital at day 180:
    • 82 vs 73; Mean difference 8 (95% CI -8 to 25)
• Safety Outcomes:
  • Composite safety end point (severe bleeding, limb ischaemia, haemolysis, worsening aortic regurgitation, device failure)
    • 24 vs 6%; RR 4.74 (95% CI 2.36 – 9.55)
  • RRT
    • 42 vs 27%; RR 1.98 (95% CI 1.27 – 3.09)
• Subgroups
  • The following had 95% CI in favour of micro-axial flow pump
    • Male: HR 0.66 (95% CI 0.47 – 0.93)
    • Mean Arterial Pressure < 64 mmHg: HR 0.61 (95% CI 0.41 – 0.92)
    • ≥2 disease vessels: HR 0.68 (95% CI 0.49 – 0.94)

Authors’ Conclusions

• Routine use of a microaxial flow pump in the treatment of patients with STEMI related cardiogenic shock led to a lower risk of death at 180 days. The incidence of adverse events was higher

Strengths

• Excellently conducted RCT with high internal validity
• Rapid placement of device following randomisation (14 minutes) with short door to balloon times
• A seemingly sick cohort with a median SBP in the 80s and lactate of 4.5, with well-balanced baseline characteristics
• Similar use of other ICU supports between groups
• Minimal protocol violations
• Minimal loss to follow up
• Clear guidelines on weaning of Impella support
• Excellent data on safety outcomes with clear definitions of what constitutes an adverse event

Weaknesses

• Unblinded
• When as-treated populations analysed confidence intervals lose significance (HR 0.77, 95% CI 0.57 – 1.03) (Figure S4)
• Single centre in UK may limit external validity
  • Figure S2 shows the heterogeneity of treatment effect across enrollment countries (Denmark vs Germany and UK) favours Denmark (not adjusted for multiplicity)
  • It is important to understand how long these devices have been used in each country to understand how and if volume of practice affects outcomes
• Slow enrollment – 10 years to complete recruitment
  • No obvious difference in subgroup analysis
• Data on safety outcomes doesn’t account for competing risk (e.g. if you die you can’t have RRT)
• Does running on the highest support for 48 hours affect the risk of haemolysis?
  • This might increase AKI and need for RRT as hypothesised by the authors
  • It might be that the patients needed less support (lower setting) than the maximal settings to achieve adequate organ support, which may result in lower rates of haemolysis
  • However, within the settings of a clinical trial it is understandable to protocolise these settings
• May be hard to directly compare with other trials looking at mechanical cardiac supports (e.g. ECLS Shock)
  • This trial didn’t include those comatose post cardiac arrest
    • Given this population will have a higher rate of neurological injury and therefore withdrawal of life sustaining therapies mortality rates may be different
    • Low rates of neurological causes of death (6 patients total, Table S3) were demonstrated in this trial
• Guidelines for other ICU support provided in protocol but no data provided on adherence (e.g. anticoagulation)
• Some industry involvement and support (Abiomed)

The Bottom Line

• It is hard to ignore this trial.  Whilst acknowledging the increased rates of adverse events, the mortality benefit is striking
• This trial should promote multi-disciplinary discussions around the potential benefit for MAFP use within local healthcare settings

External Links

• article Microaxial Flow Pump or Standard Care in Infarct-Related Cardiogenic Shock
• editorial Mechanical Circulatory Support in Cardiogenic Shock — Persistence and Progress

Metadata

Summary author: George Walker @hgmwalker89
Summary date: 12th April 2024
Peer-review editor: @davidslessor

Picture by: AI Generation Microsoft Co-Pilot