Temporary mechanical circulatory support (tMCS) has changed substantially over the past decade. The classic picture of a patient in profound shock placed on veno-arterial extracorporeal life support (VA-ECLS) as a last resort now sits alongside microaxial flow pumps, percutaneous right ventricular support devices, combined configurations, and strategies aimed not only at rescue but also at preventing perioperative circulatory failure.
Against this backdrop, the European Association for Cardio-Thoracic Surgery (EACTS), the Society of Thoracic Surgeons (STS), and the American Association for Thoracic Surgery (AATS) have jointly issued a new guideline. Rather than focusing on any single device, the document sets out to organize the entire pathway: patient selection, indication, implantation timing, device choice and combination, ICU management, transport, recovery, weaning, and transition to transplant or durable ventricular assist support.
For the cardiac surgeon, this shift in perspective may be the guideline’s most relevant contribution: temporary support is no longer viewed primarily as a rescue therapy but instead becomes part of a strategy meant to anticipate irreversible deterioration.
A multidisciplinary panel of EACTS, STS, and AATS experts developed the guideline through a systematic review of MEDLINE, EMBASE, and the Cochrane Library, supplemented with relevant subsequent literature. Recommendations were ratified through a formal consensus and voting process among panel members, and expert consensus was used whenever the evidence base was insufficient — a situation that remains common in this field.
The document covers tMCS in cardiogenic shock (CS), advanced heart failure, high-risk cardiac surgery, and post-cardiotomy shock, along with specific scenarios such as mechanical complications of myocardial infarction, heart transplantation, right ventricular (RV) failure after LVAD implantation, and pulmonary embolism.
One of the guideline’s central messages concerns the timing of implantation. In CS, tMCS is recommended before severe organ dysfunction develops (Class I, Level B). An escalating requirement for vasoactive or inotropic agents, together with hypoperfusion and a progressively rising lactate, should be read as signals to consider escalation to mechanical support.
Device choice should be individualized according to the hemodynamic phenotype and the support required. VA-ECLS provides full cardiopulmonary support and remains essential in profound shock, whereas microaxial flow pumps (mAFP) and temporary ventricular assist devices (tVAD) allow for more targeted strategies of ventricular support and unloading.
For high-risk cardiac surgery, the guideline introduces the concept of protected cardiac surgery, proposing preoperative or intraoperative tMCS in selected patients at elevated risk of low cardiac output syndrome (LCOS). When weaning from cardiopulmonary bypass (CPB) proves difficult, the guideline favors early support implantation over prolonging pharmacological strategies that may ultimately fail.
Another point of emphasis is left ventricular (LV) unloading during VA-ECLS. Because retrograde flow raises afterload and can promote ventricular distension and pulmonary congestion, the guideline holds that active LV unloading should be considered, although the optimal technique and which patients benefit most remain open questions.
Finally, the guideline stresses defining the goal of support from an early stage — recovery, transplant, durable device support, or withdrawal of therapy — so that devices designed for temporary stabilization are not indefinitely prolonged.
The authors conclude that tMCS has become an essential tool for managing CS and for selected high-risk surgical or interventional patients. In their view, outcomes depend heavily on appropriate patient and device selection, timely implantation, structured institutional protocols, and experienced multidisciplinary teams. They also highlight the need for registries, standardized criteria, and multicentre studies to strengthen evidence that remains insufficient in many areas.
COMMENTARY:
The guideline contains numerous recommendations, but its most important message can probably be distilled into a single idea: temporary mechanical support should no longer be understood solely as a late rescue therapy but should instead become part of an early, planned strategy.
In CS, the new EACTS/STS/AATS guideline insists on starting tMCS before severe organ dysfunction sets in. A progressive escalation of vasoactive drugs, hypoperfusion, and a rising lactate should be read as warning signs rather than simply as reasons to keep increasing pharmacological treatment — a recommendation that reaches Class I, Level B. For the surgeon, the most immediate translation of this principle plays out in the operating room: when weaning from CPB proves difficult, early implantation of support is recommended. The real challenge remains recognizing the point at which persisting with catecholamines and further weaning attempts stops adding benefit and starts consuming a window of opportunity.
This shift in mindset also accounts for the growing prominence of what the guideline terms protected cardiac surgery. It proposes anticipating the risk of LCOS and considering preoperative or intraoperative support in selected very-high-risk patients, taking into account ventricular function, age, comorbidity, renal impairment, metabolic status, and lactate. This is not about implanting devices prophylactically across the board, but about entering the operating room with a defined plan: which support would be used, through which access, and toward what goal if the heart fails to respond as expected. These decisions should also rest on protocols and shock teams able to act quickly and reproducibly, so that multidisciplinary discussion does not itself become a cause of delayed support.
The second major idea is that the decision can no longer simply be “ECLS or not.” The phenotype of shock should dictate the device. VA-ECLS remains essential in profound cardiopulmonary collapse, while predominantly left-sided, right-sided, or biventricular failure may call for different, and at times combined, strategies. Technological progress has moved the field away from a relatively uniform model of full support toward a more modular approach that tries to deliver just enough support at the lowest possible biological cost.
This is where full-support microaxial flow pumps (mAFP) take on particular importance — one of the guideline’s most relevant developments for the surgeon. In post-cardiotomy shock, the document recommends prioritizing, when feasible, microaxial devices capable of delivering full support over partial-flow devices. With flows of up to 5.5 L/min and the option of implantation through a graft to the supra-aortic vessels or the ascending aorta, these systems combine hemodynamic support, active LV unloading, and a strategy that can be extended for longer while remaining compatible with patient mobilization.
Its role is also particularly appealing in combination with VA-ECLS, since the latter sustains the circulation but increases LV afterload. The guideline holds that active LV unloading should be considered and, in overt biventricular failure, even proposes the VA-ECLS-plus-mAFP combination as a first-line strategy. It is worth noting that the guideline itself assigns the same grade of recommendation (Class I, Level C) to other combinations that are equally valid for biventricular support more generally, such as two microaxial flow pumps or two temporary VADs (tVAD). The message, then, is not to systematically replace one device with another, but to provide adequate support to each ventricle and to choose, from the outset, the configuration that best fits the pathophysiology and the anticipated trajectory of the patient.
Recent Spanish experience fits particularly well with this new approach. In the first Spanish multicentre cohort of Impella 5.5, which we led from A Coruña across centres with a transplant program, nearly three-quarters of patients were INTERMACS 1-2, and 68.8% ultimately reached heart transplantation (HTx), while only 4.3% received a durable LVAD (dLVAD). Prognosis, however, showed a clear gradient according to baseline severity: in-hospital mortality was 60.6% in INTERMACS 1, 37.8% in INTERMACS 2, and 26.1% in INTERMACS 3. These figures do not prove that earlier implantation necessarily improves survival, but they illustrate the guideline’s message well: the more advanced the shock at the time support is started, the less capacity we have to change its trajectory.
The initial trajectory also seems to matter. In a second Spanish multicentre study, also led by our group, involving 114 patients ultimately treated with Impella 5.5, transplantation was achieved in 76.4% of those who received the Impella as primary support, compared with 31.0% of those who reached the device after VA-ECLS; in-hospital mortality was 27.8% and 69.0%, respectively. These figures do not allow a causal comparison between the two strategies, since the ECLS-first group concentrated significantly sicker patients, with a higher proportion of INTERMACS 1, prior cardiac arrest, intubation, and greater metabolic derangement.
In fact, roughly 60% of these patients could be weaned from ECLS after transitioning to Impella 5.5, reinforcing its usefulness as a de-escalation strategy as well. This trajectory is also explicitly supported by the new guideline: when combined VA-ECLS and mAFP support is used, it is reasonable to withdraw the ECLS first and temporarily maintain LV support alone, particularly once the respiratory component and RV function have recovered but LV failure persists. The most interesting reading here is not that one device is superior to another, but that the same device can be part of entirely different trajectories.
The Spanish context adds a particular nuance. The Spanish Heart Transplant Registry shows that support prior to transplantation changed substantially between 2015 and 2024: continuous-flow temporary support now predominates, the intra-aortic balloon pump (IABP) has all but disappeared, and VA-ECLS use has risen in recent years. The registry does not, however, distinguish microaxial pumps from centrifugal devices, so the precise weight of each strategy remains unknown.
Our system is also clearly oriented toward transplantation rather than durable ventricular assist support. Current national urgency criteria require not only advanced circulatory support but also avoidance of multiorgan deterioration that could jeopardize candidacy. In Spain, then, the goal of tMCS should not simply be keeping the patient alive until an organ becomes available, but keeping them transplantable: avoiding prolonged ventilation, renal replacement therapy, and multiorgan failure, and, whenever possible, getting the patient to transplant extubated, mobilized, and rehabilitated.
In this context, configurations that allow for prolonged support and mobility take on particular value — a point the guideline itself elevates to an explicit recommendation, advising mobilization as early as possible in suitable patients and even considering conversion from femoral cannulation when it would facilitate that mobilization. In the first Spanish Impella 5.5 series, more than a third of transplanted patients also received an organ from controlled donation after circulatory death, reflecting another particularity of our system.
All of this connects to another key message of the guideline: temporary support needs a planned exit from the outset. Recovery, transplant, durable device support, a change in configuration, or withdrawal of therapy should all be continually reassessed. The document recommends keeping VA-ECLS in place for as short a time as possible, suggesting roughly 10-14 days as an ideal ceiling, while microaxial flow pumps or extracorporeal VADs allow for longer strategies. Weaning, likewise, should be guided by organ recovery, echocardiography, and hemodynamic stability; a mean arterial pressure (MAP) ≥65 mmHg with minimal vasoactive support (vasoactive-inotropic score [VIS] <10) can serve as a reference point.
Some caution is nonetheless warranted. Many recommendations still rest on Level B or C evidence — observational studies, registries, or expert consensus. Comparisons between devices are particularly susceptible to selection bias, since patients who receive VA-ECLS tend to be sicker from the outset. The guideline itself acknowledges substantial gaps in patient selection, long-term functional and neurological outcomes, and transition strategies.
These new guidelines, then, do not yet resolve every decision in temporary mechanical circulatory support. Their most important contribution is probably conceptual: anticipating support, choosing it according to pathophysiology, unloading when necessary, and considering the patient’s full trajectory from the very first moment. Spanish data, from a system particularly oriented toward transplantation, reinforce that same idea: it is not enough to sustain the patient — it must be done in time, and a genuine exit strategy must be kept open.
For those of us working in cardiac surgery, perhaps the question should no longer be simply “does this patient need support?”, but an earlier, and probably more important, one: “are we waiting too long to provide it?”
REFERENCES:
Potapov EV, Whitman G, John R, Lanmüller P, Tucanova Z, Arora RC, et al. EACTS/STS/AATS Guidelines on temporary mechanical circulatory support in adult cardiac surgery. Eur J Cardiothorac Surg. 2026;68(7):ezaf330. doi:10.1093/ejcts/ezaf330.
