What should be done when a partial-support microaxial flow pump does not appear sufficient to stabilize a patient with cardiogenic shock before implantation of a durable ventricular assist device? One option is to keep the device in place until surgery; another is to escalate beforehand to a pump capable of providing full support. This European multicentre study compares both strategies and examines whether escalation to a temporary device offering greater support can help prepare the patient before implantation of a durable ventricular assist device.

Microaxial flow pumps are being used with increasing frequency as temporary circulatory support in cardiogenic shock. Partial-support devices, such as the Impella CP®, can be implanted percutaneously and deliver flows of up to 3.5 L/min. Because implantation is straightforward, these pumps are often used early in the acute phase, even at centres without capacity to offer advanced therapies, allowing initial stabilization of the patient and subsequent transfer to a reference centre for implantation of a durable left ventricular assist device (dLVAD). Full-support devices, such as the Impella 5.5®, can by contrast reach flows of up to 5.5 L/min, but require a surgical approach for implantation. This added complexity limits their availability, so many patients reach tertiary centres already carrying a partial-support pump. A relevant clinical decision then arises: should partial support be maintained until dLVAD implantation, or should it be replaced with a full-support pump? This dilemma becomes particularly important when the patient is also supported with venoarterial extracorporeal life support (VA-ECLS). In this situation, the lower flow delivered by the partial-support pump can hinder left ventricular unloading and weaning from extracorporeal support, potentially affecting subsequent outcomes. Although the differences between the two pump types are recognized, and full support is recommended when prolonged temporary support is anticipated, little evidence exists on whether a patient who already carries a partial-support device should be escalated, or on what the optimal timing for such escalation would be. It is precisely this uncertainty that the study sets out to address.

The authors conducted a retrospective multicentre study of 130 patients from 17 European tertiary centres in Germany, Italy, Belgium, the Netherlands, the Czech Republic, Denmark, and Austria. All patients had initially received a partial-support microaxial flow pump (ps-mAFP) before dLVAD implantation. In 92 patients (70.8%), this partial support was maintained until surgery, while 38 (29.2%) were escalated beforehand to a full-support pump (fs-mAFP). The primary endpoint was 30-day survival after dLVAD implantation. As a secondary endpoint, the clinical course of both strategies during temporary support was compared, mainly in terms of mobilization and weaning from VA-ECLS. Complications occurring after implantation were also analysed (right ventricular failure requiring device support, stroke, respiratory, hepatic, or renal failure, and bleeding complications), together with 1-year survival after dLVAD implantation.

Baseline characteristics were similar between the two groups. However, there were differences in the aetiology of cardiogenic shock: acute myocardial infarction was more frequent among escalated patients (76.3% vs 35.9%; p < .001), whereas other causes of shock predominated in the group that remained on partial support. Escalated patients spent longer on temporary support before definitive implantation, with a median of 12 days versus 7 days in the group that remained on partial support (p < .001). Among patients requiring concomitant VA-ECLS, weaning before surgery was achieved more often after escalation to full support (71.4% vs 11.3%; p < .001). No differences were observed in the study’s primary endpoint: 30-day survival was 89.5% in escalated patients and 84.8% in those who remained on partial support (adjusted OR = 1.00; p = .992). During temporary support, escalation was associated with a greater likelihood of weaning from VA-ECLS before implantation (71.4% vs 11.3%; p < .001) and with greater patient mobilization (nearly 30% of patients were mobilized out of bed, compared with only 6% of those on ps-mAFP). At the time of surgery, the escalated group also required less inotropic support and showed a more favourable biochemical profile. No differences were found between groups in the need for right ventricular assist device (RVAD) implantation (p = .273). Although early survival was comparable, mortality during the first year was lower in the escalated group (HR = 0.41; p = .048). Patients who were simultaneously supported with a partial-support pump and VA-ECLS, moreover, had twice the mortality risk of those managed with the other support strategies (HR = 2.06; p = .013).

The authors conclude that escalation to a full-support pump, particularly when VA-ECLS is present, may favour a greater degree of organ recovery and provide better preconditioning before dLVAD implantation, translating into improved medium-term survival.

COMMENTARY:

This study addresses a very specific question with considerable clinical relevance: should we proceed as quickly as possible to implantation of the definitive device, or is it worth investing time in optimizing the patient with more effective temporary support? The findings of this article appear to support the latter strategy in certain patients. Escalation did not improve early survival, but it was associated with an apparently more favourable clinical status before surgery: a greater chance of discontinuing extracorporeal support, a lower requirement for inotropes, greater mobilization, and better markers of organ recovery. In other words, the potential benefit may not simply stem from using a pump with higher flow, but from achieving better conditions to prepare the patient before definitive support. The authors themselves propose this mechanism as a possible explanation for the difference observed in medium-term survival.

Perhaps the finding with the greatest practical impact from this work lies in the subgroup of patients who were simultaneously supported with a partial-support microaxial pump and venoarterial extracorporeal support. In this subgroup, 1-year survival was 55.5%, compared with 79.5% among the remaining patients treated with other strategies, suggesting that continued dependence on extracorporeal support up to the time of definitive implantation may be a marker of greater severity and worse prognosis.

In this scenario, escalating to a full-support device could promote ventricular unloading and allow extracorporeal support to be reduced or withdrawn before dLVAD implantation. Rather than pointing toward systematic escalation, these results particularly identify the patient in whom extracorporeal support cannot be discontinued as a possible candidate for intensified ventricular support. In less compromised patients, an isolated partial-support pump may provide adequate support while avoiding the risks associated with an additional surgical procedure. Escalation should therefore not be applied systematically but individualized according to the patient’s clinical course, support requirements, and prognosis.

These findings should nonetheless be interpreted with caution. The retrospective design and the limited number of patients, particularly in the escalation group, constrain the strength of the conclusions. Furthermore, the decision to escalate support did not follow predefined criteria but instead depended on clinical presentation and on each centre’s own practice. This introduces substantial selection bias and makes it difficult to determine to what extent the observed differences are truly attributable to the change of device.

The clinical message, therefore, should not be to “always escalate,” but rather to avoid proceeding prematurely to definitive implantation in an insufficiently optimized patient. An isolated partial-support pump may be adequate in selected patients, and an unnecessary escalation adds the risks of a surgical procedure. Conversely, in more critical patients (those in whom extracorporeal support cannot be withdrawn, in whom inotropic requirements remain high, in whom ventricular unloading is insufficient, or in whom device-related complications arise), escalation may become a reasonable strategy to gain recovery time and better prepare the patient. This proposal is consistent with the algorithm put forward by the authors themselves, although it should still be regarded as a clinical hypothesis supported by observational evidence rather than a definitively proven strategy.

REFERENCE:

Lewin D, Nersesian G, Rojas SV, Pieri M, Szymanski MK, Koij J, et al. Impact of partial- to full-support escalation with microaxial flow pumps prior to durable left ventricular assist device implantation. Eur J Cardiothorac Surg. 2026;68(5):ezag165. doi:10.1093/ejcts/ezag165.