Background: The role of myocardial unloading with Impella prior to elective cardiac operations in patients with diminished left ventricular (LV) function is yet to be established, particularly in the community hospital setting.
Hypothesis: Pre-operative selection of high-risk patients for placement of Impella devices to medically optimise them before surgery and assist with recovery in the immediate postoperative period should allow faster recovery, reduced intensive care and hospitalisation days, improved end-organ function and reduced inotropic needs immediately postoperatively.
Methods: Impella has been used at our hospital since 2014 for high-risk cath lab procedures and treatment of acute chronic heart failure (CHF) exacerbations. For many years, prophylactic intra-aortic balloon pump (IABP) had been used peri-operatively to support higher-risk patients, but starting in May 2017, we implemented a programme of peri-procedural unloading with Impella in select higher-risk patients in order to improve postoperative outcomes and avoid the need for urgent or emergent device implantation postoperatively. Data were collected retrospectively on these patients to identify outcomes in this high-risk cohort and to validate our site-specific algorithm for prophylactic placement of these devices.
Results: Since the initiation of this strategy, we have placed 149 Impella devices, with 36 devices placed in patients ultimately undergoing surgical intervention. The mean age in the surgical cohort was 63 ± 11 years (median 64 years) and 26 were men (72%). Of these, 25 patients were selected for elective Impella placement prior to surgery, with another 11 patients requiring Impella placement in the cath lab for unstable angina or acute CHF decompensation who ultimately underwent surgery following medical optimisation.
Surgical procedures included coronary artery bypass grafting (CABG; n=29), CABG/valve (n=5) and valve (n=2). In the CHF/shock cohort, the mean pre-op ejection fraction was 22 ± 10% (median 20%). There were three deaths in the 36 patients: two of 25 in the elective group and one of 11 in the unstable group. In the surviving patients, the mean total Impella duration was 5 ± 4 days (median 5 days), with a mean postoperative Impella duration of 4 ± 2 days (median 3 days).
Of patients with serious end-organ dysfunction, two of five patients with acute renal insufficiency requiring dialysis and two of four patients with hepatic insufficiency survived to discharge. Total inotrope duration was 2 ± 2 days (median 1 day), with only five patients requiring high-dose inotropes. The mean postoperative intensive care length of stay (LOS) was 7 ± 5 days (median 6 days) and the mean postoperative hospital LOS was 11 ± 5 days (median 10 days). Postoperative mortality was 8% (three of 36, 30 days or in-hospital).
Conclusion: Preoperative identification of high-risk patients who may benefit from LV unloading with Impella is possible in the community setting. More data might permit development of a selection algorithm that would be broadly applicable to optimise outcomes in high-risk cardiac surgery patients.