Impella Before or After the Wire: Timing Mechanical Support in STEMI Cardiogenic Shock
A retired postal carrier arrives in cardiogenic shock from an anterior STEMI. The culprit vessel decision makes itself; the real question is whether a mechanical pump goes in before that vessel is opened or waits to see if he needs it.
J.T., a 71-year-old man, spent thirty-two years walking a postal route before retiring six years ago, and still walks four miles most mornings out of habit as much as exercise. He was on that walk when the pain started — crushing, mid-sternal, radiating down his left arm — and called 911 himself, giving dispatchers his own address before it occurred to him to sit down. His only real cardiac history is well-controlled hypertension on amlodipine; he quit smoking fifteen years ago and has no known diabetes. EMS caught ST-elevation across the anterior leads on the first monitor strip and activated the cath lab before he reached the hospital doors. By the time the team was scrubbed in, his systolic pressure had fallen to 78 despite two liters of fluid, his extremities were cool to the elbow, and his lactate had climbed to 4.2 — cardiogenic shock, not just a large infarct. The angiogram confirmed a proximal LAD occlusion as the culprit, with a second, severe, non-culprit lesion in the circumflex that would need its own decision once he was stabilized.
The question in front of the team wasn't whether to open the LAD — that decision made itself — it was whether to place a microaxial flow pump before doing so or to treat the shock first with vasopressors and escalate only if he failed to stabilize. DanGer Shock, the first randomized trial to show a mortality benefit from any mechanical circulatory support device in this exact setting, found routine early pump placement lowered 180-day mortality from 58.5% to 45.8% against standard care, an effect confirmed again on long-term follow-up. It also found a real cost: more bleeding, more vascular complications, more sepsis, and more need for renal replacement therapy in the device arm. Subgroup analysis found the benefit larger in patients with a systolic pressure below the trial's own median of 82 mm Hg, and in those with multivessel disease. J.T. sits inside both: a systolic pressure of 78 and a severe non-culprit circumflex lesion alongside the culprit LAD. That makes his match to the more-benefited slice of the trial a matter of two soft, post-hoc criteria rather than one — which is a reason to weigh the subgroup carefully, not a reason to treat it as settled.
Cath lab, before the wire crosses
I want the pump in before we cross the LAD. DanGer Shock is the first randomized evidence any mechanical device has ever shown a mortality benefit in this setting, and both his multivessel disease and a systolic pressure under the trial's median put him in the subgroups its own analysis found benefited more, not less. The mechanistic argument lines up with that finding — unloading the ventricle before revascularization avoids the extra stress opening the artery itself puts on a heart already this compromised.
I'm not arguing every shock patient needs this, reflexively. I'm arguing this specific patient's own numbers put him inside the group where the trial's benefit was clearest.
The complication numbers from that same trial are real, not a footnote — more bleeding, more vascular complications, more sepsis, more renal replacement therapy. And "mechanical support helps" isn't one finding across devices — ECLS-SHOCK found VA-ECMO made no difference to 30-day mortality in a similar population, with its own real increase in bleeding and vascular complications. I'd start norepinephrine, open the artery, and escalate to the pump only if he doesn't stabilize. Most patients at his pressure do.
I take the subgroup argument seriously — I just don't think two post-hoc subgroups from a single 355-patient trial should become a default for every multivessel shock patient, however well his numbers happen to line up with them.
I'd frame this narrower than either of you. This isn't "place it in every shock patient" versus "place it in none" — his own numbers put him inside the more clearly beneficial slice of DanGer Shock's data on both counts, the vessel burden and a pressure below the trial's median. Those are post-hoc subgroups, not a licence to generalize, but they describe him. That's a case for placing it in him, specifically, today — not a case for a blanket policy in either direction.
Agreed: place the microaxial pump before crossing the LAD, proceed with primary PCI of the culprit vessel, continue norepinephrine as needed, and stage the circumflex once he stabilizes.
Not agreed, and carried forward explicitly rather than smoothed over:
The Critical Care Physician's position is that routine early placement should stay reserved for patients whose numbers match the trial's most-benefited subgroup, not applied as a blanket rule.
The Interventional Cardiologist's position is that the mechanistic case for early unloading will likely hold up broadly, not just in the narrow slice tested so far.
The Second Interventional Cardiologist's framing — that today's decision was about this patient's own numbers, not a policy for the next one — was accepted by both other voices as the actual basis for today's plan, without resolving which of their broader positions is right going forward.