Skip to main content

Foundational Trials

IABP-SHOCK II

Image: Shutterstock / ChaNaWiT

Publication

  • Title: Intraaortic Balloon Support for Myocardial Infarction with Cardiogenic Shock
  • Acronym: IABP-SHOCK II
  • Year: 2012
  • Journal published in: The New England Journal of Medicine
  • Citation: Thiele H, Zeymer U, Neumann FJ, Ferenc M, Olbrich HG, Hausleiter J, et al. Intraaortic balloon support for myocardial infarction with cardiogenic shock. N Engl J Med. 2012;367(14):1287-1296.

Context & Rationale

  • Background
    Cardiogenic shock complicating acute myocardial infarction (AMI) carries a persistently high early mortality despite contemporary revascularisation and intensive care.
  • Research Question/Hypothesis
    In patients with AMI complicated by cardiogenic shock undergoing early revascularisation, does routine intra-aortic balloon pump (IABP) support reduce 30-day all-cause mortality compared with no planned IABP?
  • Why This Matters
    IABP was widely used as adjunctive support in AMI-related shock, despite uncertain outcome benefit and recognised vascular/bleeding risks; a definitive, contemporary randomised trial was required to inform practice and guideline recommendations.

Design & Methods

  • Research Question: In AMI with cardiogenic shock planned for early revascularisation, does routine IABP reduce 30-day mortality versus no planned IABP?
  • Study Type: Prospective, multicentre, randomised, open-label, parallel-group controlled trial; 1:1 central (internet-based) randomisation stratified by centre; 37 centres in Germany; enrolment June 2009 to March 2012.
  • Population:
    • Adults (≥18 years) with STEMI or non-STEMI complicated by cardiogenic shock and an intention to pursue early revascularisation (predominantly PCI). 1
    • Cardiogenic shock definition: systolic blood pressure <90 mmHg for ≥30 minutes or catecholamine therapy required to maintain systolic blood pressure ≥90 mmHg, plus pulmonary congestion, plus signs of impaired organ perfusion (e.g., altered mental status, cold/clammy skin, oliguria <30 mL/h, or serum lactate >2.0 mmol/L). 1
    • Key exclusions (trial-facing): shock duration >12 hours; prolonged resuscitation (>30 minutes); “no intrinsic heart action”; mechanical causes of shock (e.g., acute severe mitral regurgitation, ventricular septal rupture); severe peripheral artery disease precluding IABP; severe cerebral deficit; age >90 years; comorbidity with life expectancy <6 months; shock of non-cardiac cause. 1
  • Intervention:
    • IABP inserted via femoral access before or immediately after revascularisation, providing counterpulsation support (typically 1:1 ECG triggering) until sustained haemodynamic stabilisation; weaning by stepwise reduction of trigger ratio.
    • Delivered IABP support duration: median 3.0 days (IQR 2.0–4.0; range 1–16 days) among those treated with IABP.
  • Comparison:
    • No planned IABP (standard care including early revascularisation, vasoactive support, ventilation and other organ support as required); IABP permitted as rescue (e.g., for mechanical complications or clinical deterioration).
  • Blinding: Open-label to treating teams and participants; primary endpoint (mortality) objective; clinical endpoint adjudication committee blinded to treatment assignment (as per published design). 1
  • Statistics: Sequential design with interim analyses; powered (80%) to detect a 12 percentage-point absolute increase in 30-day survival (assumed 56% control vs 68% IABP) with overall type I error 0.05 and a final analysis significance threshold of 0.044; planned sample size 600 (300/group); primary analysis by intention-to-treat (with supportive per-protocol and adjusted analyses reported).
  • Follow-Up Period: 30 days for the primary endpoint (with longer-term follow-up reported in later publications).

Key Results

This trial was not stopped early. Recruitment reached the planned sample size (600 randomised); the primary 30-day endpoint was assessed in 598 participants (300 IABP vs 298 control) due to one withdrawal of consent and one patient lost to follow-up.

Outcome IABP Control (no planned IABP) Effect p value / 95% CI Notes
All-cause mortality at 30 days (primary; ITT) 119/300 (39.7%) 123/298 (41.3%) RR 0.96 95% CI 0.79 to 1.17; P=0.69 Primary analysis sample 598 (one consent withdrawal; one lost to follow-up).
All-cause mortality at 30 days (per-protocol) 94/251 (37.5%) 109/263 (41.4%) RR 0.91 95% CI 0.74 to 1.11; P=0.35 Supportive analysis excluding crossovers/protocol deviations (as reported).
Reinfarction in hospital 9/300 (3.0%) 4/298 (1.3%) RR 2.24 95% CI 0.70 to 7.18; P=0.16 Safety endpoint.
Stroke in hospital 2/300 (0.7%) 5/298 (1.7%) RR 0.40 95% CI 0.08 to 2.03; P=0.28 Safety endpoint.
Life-threatening or severe bleeding in hospital (GUSTO) 10/300 (3.3%) 13/298 (4.4%) RR 0.76 95% CI 0.34 to 1.72; P=0.51 Safety endpoint (GUSTO definition).
Peripheral ischaemic complications requiring intervention 13/300 (4.3%) 10/298 (3.4%) RR 1.29 95% CI 0.58 to 2.90; P=0.53 Safety endpoint.
Sepsis in hospital 47/300 (15.7%) 61/298 (20.5%) RR 0.77 95% CI 0.54 to 1.08; P=0.15 Safety/complication endpoint.
Time to haemodynamic stabilisation (days; median [IQR]) 3.0 (1.0–5.0) 3.0 (1.0–6.0) Not reported P=0.50 Process-of-care endpoint.
Duration of catecholamine therapy (days; median [IQR]) 3.0 (2.0–5.0) 3.0 (2.0–5.0) Not reported P=0.81 Process-of-care endpoint.
Duration of intensive care treatment (days; median [IQR]) 6.0 (3.0–12.0) 6.0 (3.0–13.0) Not reported P=0.34 Process-of-care endpoint.
Mechanical ventilation required 254/299 (85.0%) 244/299 (81.6%) Not reported P=0.26 Process-of-care endpoint (denominators reflect available data as reported).
Left ventricular assist device implantation 10/300 (3.3%) 23/299 (7.7%) Not reported P=0.02 Rescue mechanical support use differed between groups.
Renal replacement therapy 62/301 (20.6%) 47/299 (15.7%) Not reported P=0.12 Organ support endpoint.
  • Routine IABP did not reduce 30-day mortality versus control (RR 0.96; 95% CI 0.79 to 1.17; P=0.69), with consistent null findings in per-protocol and adjusted analyses.
  • Despite substantial intervention separation (IABP delivered in 288/301 assigned vs crossover in 30/299 controls) and a median IABP duration of 3.0 days, key process-of-care measures (time to haemodynamic stabilisation, catecholamine duration, ICU length of stay) were similar between groups.
  • Subgroup analyses were broadly consistent; nominal interactions were observed for prior myocardial infarction (RR 1.44; 95% CI 0.93 to 2.21 vs RR 0.86; 95% CI 0.69 to 1.07; interaction P=0.04) and hypertension (RR 1.06; 95% CI 0.84 to 1.34 vs RR 0.67; 95% CI 0.45 to 1.01; interaction P=0.05).

Internal Validity

  • Randomisation and allocation: Central internet-based randomisation (1:1), stratified by centre; baseline characteristics were closely balanced (e.g., age median 70 vs 69 years; left ventricular ejection fraction median 35% vs 35%).
  • Dropout/exclusions: Of 600 randomised, the primary endpoint analysis included 598 participants (300 vs 298) due to one withdrawal of consent and one patient lost to follow-up at 30 days; attrition was minimal.
  • Performance/detection bias: Open-label design creates potential for differential co-interventions; however, the primary endpoint (mortality) is objective, and endpoint adjudication was conducted by a blinded committee (published design). 1
  • Protocol adherence & separation of the variable of interest: IABP was actually received by 288/301 (95.7%) assigned to IABP; 13/301 (4.3%) did not receive IABP; crossover to IABP occurred in 30/299 (10.0%) controls; delivered support duration was median 3.0 days (IQR 2.0–4.0).
  • Co-interventions and rescue therapy: Revascularisation was highly protocol-concordant and contemporary (PCI in 573/598 [95.8%]; CABG in 22/598 [3.7%]); left ventricular assist device implantation was more frequent in the control group (7.7% vs 3.3%; P=0.02), potentially diluting any between-group differences in clinical outcomes.
  • Heterogeneity: Multicentre enrolment across 37 centres increases generalisability and reduces centre-specific confounding; treatment effects were directionally similar across most prespecified/post hoc subgroups.
  • Timing and dose: The intervention was delivered within the practical PCI workflow (before or immediately after revascularisation) and for a clinically meaningful duration; within the IABP arm, mortality was similar whether insertion occurred before versus after revascularisation (as reported).
  • Statistical rigour: Prespecified sequential monitoring with conservative final alpha (0.044) and a largely complete primary endpoint dataset; supportive per-protocol and adjusted analyses did not materially change inference.

Conclusion on Internal Validity: Overall, internal validity is moderate-to-strong—randomisation and follow-up were robust with an objective primary endpoint, but the open-label design, clinically driven crossovers (10% of controls), and differential rescue mechanical support introduce plausible dilution and co-intervention effects.

External Validity

  • Population representativeness: Broad AMI-related cardiogenic shock population (both STEMI and non-STEMI), typical age (median ~70 years), and high pre-randomisation resuscitation rates (~43%); exclusions (shock duration >12 hours, prolonged resuscitation, mechanical complications, severe peripheral vascular disease, age >90) limit applicability to very late presenters, profound anoxic injury, or mechanical causes of shock.
  • Applicability: Highly applicable to high-resource systems with rapid access to PCI and intensive care; extrapolation to settings without timely revascularisation, or to contemporary practice dominated by “active” mechanical circulatory support (e.g., microaxial pumps/VA-ECMO) should be cautious.
  • Intervention scope: Findings apply to conventional IABP counterpulsation as routinely delivered in a modern PCI-era shock pathway; they do not directly address other mechanical support devices or non-AMI shock phenotypes.

Conclusion on External Validity: External validity is good for AMI-related cardiogenic shock managed with early PCI in advanced centres, but is limited for non-AMI shock, mechanical complications, late shock, prolonged arrest, and settings without timely revascularisation or access to escalation therapies.

Strengths & Limitations

  • Strengths: Large, contemporary randomised trial in a high-mortality population; multicentre design; central randomisation; objective primary endpoint with near-complete follow-up; high revascularisation rates reflecting modern care; prespecified sequential monitoring and supportive analyses.
  • Limitations: Open-label design with clinically mandated crossovers and protocol deviations; potential dilution by differential rescue mechanical support (higher LVAD use in controls); powered for a relatively large absolute mortality difference (smaller benefits/harm cannot be excluded); results pertain to IABP (modest haemodynamic augmentation) rather than more potent mechanical support strategies.

Interpretation & Why It Matters

  • Routine IABP is not an outcome-improving default in AMI shock
    In a PCI-era pathway with high revascularisation uptake, adding routine IABP counterpulsation did not improve 30-day survival and did not materially shorten haemodynamic stabilisation, catecholamine exposure, ICU stay, or ventilation duration.
  • Device therapy requires hard-outcome randomised evidence
    IABP-SHOCK II illustrates that plausible physiological mechanisms and entrenched practice patterns can fail to translate into mortality benefit when tested rigorously in modern care systems.
  • Shifts attention to escalation strategies and patient selection
    The neutral result sharpened focus on risk stratification, timing, and selection of more potent mechanical circulatory support options—while reserving IABP for selected indications (e.g., certain mechanical complications or bridging scenarios) rather than routine use.

Controversies & Subsequent Evidence

  • Mechanism–outcome mismatch: IABP provides modest haemodynamic augmentation; in AMI shock treated with early PCI and vasoactive support, this degree of support may be insufficient to alter mortality despite physiological plausibility.
  • Crossover and rescue therapy as diluters (but not explanatory): Control crossover to IABP (10%) and higher LVAD use in controls (7.7% vs 3.3%) could dilute a true treatment effect; however, per-protocol and adjusted analyses remained neutral, supporting a genuinely small/absent mortality benefit.
  • Subgroup signals and multiplicity: Nominal subgroup interactions (e.g., prior MI and hypertension) did not provide a coherent, practice-changing phenotype and are compatible with chance findings given multiple comparisons.
  • Longer-term follow-up: Subsequent publications at 12 months and at 6 years reported no survival advantage with routine IABP, reinforcing the early neutral result. 56
  • Synthesis of evidence: Systematic reviews and meta-analyses incorporating IABP-SHOCK II show no mortality reduction for IABP in AMI-related cardiogenic shock in the contemporary era, aligning with the trial’s conclusions. 234
  • Guideline impact: Major society guidance has moved away from routine IABP use in AMI-related cardiogenic shock, reflecting the evidence base consolidated by IABP-SHOCK II and subsequent syntheses/statements. 78910

Summary

  • IABP-SHOCK II was a large, multicentre, PCI-era randomised trial testing routine IABP in AMI complicated by cardiogenic shock.
  • Routine IABP did not reduce 30-day mortality versus no planned IABP (39.7% vs 41.3%; RR 0.96; 95% CI 0.79 to 1.17; P=0.69).
  • Key process-of-care metrics (stabilisation time, catecholamine duration, ICU stay, ventilation) were similar despite substantial intervention separation.
  • Major safety outcomes (stroke, severe bleeding, limb ischaemia requiring intervention) were not significantly different between groups.
  • Long-term follow-up and evidence syntheses have reinforced the absence of mortality benefit and contributed to guideline de-implementation of routine IABP in AMI shock.

Further Reading

Other Trials

Systematic Review & Meta Analysis

Observational Studies

Guidelines

Notes

  • The trial tested conventional IABP counterpulsation within a modern revascularisation pathway; it does not directly address escalation with more potent mechanical circulatory support devices.
  • Secondary outcomes and process-of-care endpoints were not adjusted for multiple comparisons; the primary endpoint provides the most reliable inference.

Overall Takeaway

IABP-SHOCK II is a landmark device trial because it definitively tested a long-entrenched, guideline-endorsed therapy (IABP) against hard clinical outcomes in a contemporary, PCI-era cardiogenic shock population. Its neutral mortality result—reinforced by longer-term follow-up and meta-analyses—reframed IABP as a selective rather than routine intervention and redirected the field toward more potent, rigorously evaluated mechanical support strategies.

Overall Summary

  • In AMI-related cardiogenic shock treated with early revascularisation, routine IABP did not reduce 30-day mortality.
  • Process-of-care measures and major safety outcomes were broadly similar between groups despite substantial separation in IABP exposure.
  • Subsequent follow-up and evidence syntheses consolidated de-implementation of routine IABP in guideline recommendations.

Bibliography