Reconsidering FDA Guidelines: A Single-Center Experience of Prolonged Impella 5.5 Support.

OBJECTIVE: Impella 5.5 (Abiomed, Danvers, MA, USA) is approved by the US Food and Drug Administration (FDA) for mechanical circulatory support for ≤14 days. It is unknown whether prolonged support is associated with worse outcomes. We sought to review our single-center experience with Impella 5.5 and compare outcomes based on support duration. METHODS: We retrospectively reviewed adult patients (≥18 years old) supported with Impella 5.5 at our institution (May 2020 to April 2023). Patients on prolonged support (>14 days) were compared with those supported for ≤14 days. RESULTS: There were 31 patients supported with Impella 5.5 including 14 (45.2%) supported >14 days. Median support duration for those on prolonged support was 43.5 (interquartile range [IQR] 25 to 63.5) days versus 8 (IQR 6, 13) days for those who were not (P < 0.001). Overall, the device-related complication rate was 9.7% and did not differ between groups (P = 0.08). Overall, 30-day postimplant survival was 71% and did not differ by support duration (P = 0.2). In-hospital mortality was 32% and did not differ between cohorts (P > 0.99). Among those surviving to explant (n = 22), long-term strategy included bridge to durable ventricular assist device (18%, n = 4), cardiac transplant (55%, n = 12), and cardiac recovery (27%, n = 6). CONCLUSIONS: High-risk patients with cardiogenic shock may be supported with Impella 5.5 beyond the FDA-approved duration without increased risk of complications or mortality.

Utility of Fiberoptic Endoscopic Evaluation of Swallowing After Left Ventricular Assist Device Implantation.

Objective Dysphagia following cardiac surgery is common and associated with adverse outcomes. Among patients receiving left ventricular assist devices (LVAD), we evaluated the impact of fiberoptic endoscopic evaluation of swallowing (FEES) on outcomes. Methods A single-center pilot study was conducted in adults (≥18 years of age) undergoing durable LVAD (February 2019 - January 2020). Six patients were prospectively enrolled, evaluated, and underwent FEES within 72 hours of extubation-they were compared to 12 control patients. Demographic, surgical, and postoperative outcomes were collected. Unpaired two-sided t-tests and Fisher's exact tests were performed. Results Baseline characteristics were similar between groups. Intraoperative criteria including duration of transesophageal echo (314 ± 86 min) and surgery (301 ± 74 min) did not differ. The mean time of intubation was comparable (57.3 vs. 68.7 hours, p=0.77). In the entire cohort, 30-day, one-year, two-year, and three-year mortality were 0%, 5.6%, 5.6%, and 16.7%, respectively. Sixty-seven percent of the patients that underwent FEES had inefficient swallowing function. The FEES group trended to a shorter hospital length of stay (LOS) (29.1 vs. 46.6 days, p=0.098), post-implantation LOS (25.3 vs 30.7 days, p=0.46), and lower incidence of postoperative pneumonia (16.7% vs. 50%, p=0.32) and sepsis (0% vs. 33.3%, p=0.25). Conclusion FEES did not impact 30-day, one-year, two-year, or three-year mortality. Though not statistically significant, patients who underwent FEES trended toward shorter LOS and lower postoperative pneumonia and sepsis rates. Additionally, we report a higher incidence of dysphagia among patients undergoing FEES despite comparable baseline risk factors with controls.

Early versus standard renal replacement therapy after left ventricular assist device implantation.

OBJECTIVES: Renal function may improve after left ventricular assist device (LVAD) implant, however, some patients develop postoperative acute kidney injury (AKI). Randomized trials showed benefit for early renal replacement therapy (RRT) in critically ill patients with AKI, but this practice has not been studied in LVAD patients. METHODS: We performed a single-center, retrospective cohort study of all adults (>18 years) who underwent LVAD placement from 1/2010 to 12/2018. We collected preoperative, hemodynamic, echocardiographic, intraoperative, and postoperative data. AKI was defined according to Kidney Disease: Improving Global Outcomes definition. Early (E) RRT was considered treatment at AKI stage II or below. Standard (S) RRT was considered treatment at AKI stage III. Outcomes and Kaplan-Meier analysis were compared between groups. RESULTS: A total of 184 patients were included (mean age 56.10 years, 81% males, 30.4% African-American race). A total of 71 (38.6%) developed AKI and 17 (9.24%) needed RRT (11 E vs 6 S). A total of 11 remained hemodialysis-dependent at discharge (5 [45.5%] in E vs 6 [100%] in S, P = .043). There was a trend toward shorter intensive care unit stay and ventilation time in E group, and overall hospital stay was significantly less in the E group (48.18 ± 25.95 vs 94.00 ± 53.07 days, P = .028). Thirty-day mortality was similar between groups (E 18% vs S 16%, P = .9), but there was a trend toward improved overall survival in the E group. CONCLUSION: This is the first study to examine early initiation of RRT after LVAD implant. Early RRT was associated with shorter hospital stay, lower need for permanent RRT, and a trend toward improved survival. This practice may provide significant cost savings and should be examined further.

Left Ventricular Assist Device Artifact in EEG.

Left ventricular assist devices are increasingly used as therapy for patients with severe congestive heart failure. These patients typically receive care in the intensive care unit when EEG monitoring is necessary. Identification of artifacts created by these devices is important for accurate EEG diagnosis, thus avoiding unnecessary therapies that may result in complications or require intubation of the patient. This report describes a case of a patient with altered mental status and history of congestive heart failure with a left ventricular assist device who developed seizures following left subdural hematoma and emergent evacuation. While being treated for seizures, the EEG showed a diffuse continuous and regular delta, predominant on the left that mirrored the electrocardiogram (EKG) rhythm strip and appeared to indicate an external artifact. Moving the amplifier farther away from the heart and repositioning the patient resolved the artifact. It is important to recognize potential sources of interference and artifacts when these devices are involved. This case illustrates how clinicians can prevent administration of unneeded therapies to these patients.

EEG Artifact Versus Subclinical Status Epilepticus in a Patient Following Cardiac Arrest.

BACKGROUND: A challenge in ICU EEG interpretation is identifying subclinical status epilepticus versus patterns on the ictal-interictal continuum versus other repetitive patterns. In the electrically noisy intensive care unit, identifying and eliminating interference and artifact allow accurate diagnoses from the EEG, avoiding unnecessary treatment or sedation. METHODS: We present a case during Impella (Abiomed Inc, Danvers, MA) continuous flow left ventricular assist device use where the EEG artifact was initially misinterpreted as seizure by the resident and treated as status epilepticus because of the "focal" sharply contoured repetitive pattern. During percutaneous coronary intervention (PCI), an 88-year-old developed ventricular tachycardia followed by ventricular fibrillation requiring chest compressions for 10 min, multiple defibrillations, and treatment with epinephrine, amiodarone, calcium, bicarbonate, and magnesium. The patient had an Impella placed during PCI with therapeutic hypothermia initiated after the cardiopulmonary arrest. His neurological exam demonstrated preserved pupillary, corneal, gag and cough reflexes and spontaneous respirations. RESULTS: Long-term video EEG monitoring is included in our institution's hypothermia protocol. Initial baseline EEG performed 2 h after PCI showed a persistent rhythmic sharp discharge from the left central temporal region resembling left hemisphere status epilepticus. The sharp waves have an alternating repeating 2:1 relationship with the EKG rhythm strip. This is best seen in the left hemisphere, which we posit is related to the Impella's positioning across the aortic valve pointing toward the patient's left side. A chest x-ray confirmed the device's position immediately before EEG monitoring. Arterial pressure tracings were not available in the chart. CONCLUSIONS: There is a low-amplitude spiky artifact; however, there was no pacing at that time. It is possible that synergistic flow with systole/diastole reinforced the pulsatility with movement of the Impella, resulting in the alternating pattern. The patient's hemodynamic instability precluded extensive troubleshooting with the Impella device, but after EEG repositioning, the artifact was eliminated.