Monitored Anesthesia Care for Axillary Impella 5.5 Placement: A Feasible Option for Patients With Severe Cardiogenic Shock With Relative Contraindication to Intubation

Introduction

The Impella 5.5 (Abiomed, Danvers, Massachusetts) is the latest iteration of percutaneous micro-axial, continuous flow left ventricular assist device (LVAD) commercially available to provide short-term support to the left ventricle (LV) in patients with end-stage systolic failure. ¹ Given the need for a surgical cut down to place a conduit for vascular access and transesophageal echocardiography (TEE) to confirm intraventricular placement, the Impella 5.5 is typically placed under general anesthesia. Unfortunately, the majority of patients in need of an urgent percutaneous LVAD may not tolerate a general anesthetic and its sequelae due to refractory cardiogenic shock or profound pulmonary hypertension. We present a case for which monitored anesthesia care (MAC) was successfully utilized for the placement of an Impella 5.5 in a patient with both end-stage cardiac and pulmonary failure secondary to systemic pressure pulmonary hypertension.

Case Report

A 28-year-old female with a history of congenital heart disease (atrial septal defect, ventricular septal defect, and patent ductus arteriosus, all status post-surgical corrections in early infancy) and pulmonary hypertension (WHO Group 1) treated with triple therapy (sildenafil, macitentan, and treprostinil) presented with signs and symptoms of biventricular heart failure.

Inpatient heart failure treatment consisted of guideline-directed medical therapy and aggressive diuresis. The patient’s ongoing heart/lung transplant workup was accelerated, and she received formal approval from our institution’s transplantation committee on hospital day 4 (INTERMACS score 4, New York Heart Association class IV). Given her history of congenital heart disease, she was classified as status 4. Right heart catheterization revealed a right atrial pressure of 15 mmHg, pulmonary artery pressures of 50/44 mmHg (mean 47 mmHg) on sildenafil, macitentan, and treprostinil, pulmonary capillary wedge pressure of 20 mmHg, pulmonary artery oxygenation saturation of 54%, cardiac output of 2.9 L/min, cardiac index of 1.46 L/min/m², Pulmonary Vascular Resistance of 9.21 Woods Units, and Transpulmonary Pressure Gradient of 27 mmHg. Echocardiographic assessment revealed a left ventricular dominant cardiogenic shock. Laboratory values revealed an increase in BUN/Cr ratio (from 14 up to 22) and a precipitous decrease in serum sodium concentration (from 140 mmol/L down to 133 mmol/L). Temporary mechanical support of the LV was indicated to restore circulation and prevent further end-organ failure as a bridge to transplant. With this development, the patient was emergently listed for heart/lung transplant, UNOS status. ²

On hospital day 5, the patient was taken to the hybrid surgical suite, part of our cardiovascular operating room pavilion for axillary placement of an Impella 5.5 under MAC. Given her fragile cardiopulmonary status and severe pulmonary hypertension, a general anesthetic could result in significant morbidity including the risk of inability to safely extubate following Impella placement. This was a crucial consideration because it was unknown when a suitable match for a heart/lung en bloc transplantation would be available, and leaving the patient intubated on mechanical ventilation for a prolonged period may cause significant deconditioning and worsen her outcome. Also, as the Impella 5.5 only supports the left ventricle, if the right ventricle were to fail from positive pressure ventilation or exacerbation of her existing severe pulmonary hypertension, there would be no option other than veno-arterial extracorporeal membrane oxygenation. Placement of the Impella would hopefully indirectly support the right ventricle by lowering pulmonary pressures with increased forward flow and decreased afterload.

After a discussion between the anesthesia and surgical teams, we felt that proceeding under MAC would be most prudent. A pre-induction arterial line was placed in the left radial artery, and the patient was started on a propofol infusion and dexmedetomidine infusion. Supplemental oxygen was administered via a simple facemask. Propofol was initiated at 50 mcg/kg/min, uptitrated to 75 mcg/kg/min, and then reduced to 25 mcg/kg/min for the remainder of the case; dexmedetomidine was maintained constant at .1 mcg/kg/hr. A phenylephrine infusion was also initiated at 50 mcg/min to normalize the reductions in systemic vascular resistance and the increases in venous capacitance induced by propofol. The patient already had a Swan-Ganz catheter in situ from her earlier right heart catheterization, and we continuously calculated pulmonary artery pulsatility indices to gauge and optimize right ventricular performance during Impella insertion and activation. ²

Just before making the incision, the surgeon generously anesthetized the surgical site with local anesthetic (20 cc bupivacaine .25% with 1:200,000 epinephrine) creating a field block. Small bolus doses of ketamine (5–10 mg) were also administered once the patient had reached a steady state of amnesia under the propofol and dexmedetomidine infusions. An axillary Impella 5.5 was placed under standard fashion, including a right axillary artery cut down and advancement of a guidewire into the left ventricle. ³ Given the lack of full TEE support at this time, the surgeon initially relied on fluoroscopy to advance the Impella guidewire to abut the aortic valve. A pigtail catheter was used to identify the coronary cusps and aortic valve annulus, in a manner similar to transcatheter aortic valve placement. The soft .035” was exchanged for the Impella wire and the device placed 5–6 cm deep to the aortic annulus with fluoroscopic guidance.

In conjunction with fluoroscopic imaging, Impella placement requires final confirmation with echocardiography (either transthoracic or transesophageal). We chose to perform a limited exam under sedation using TEE with the goal of obtaining better acoustic windows. While the patient was still sedated, a pediatric TEE probe was inserted. A mid-esophageal long-axis aortic view was obtained to measure the depth of the Impella into the LV, and color flow Doppler verified optimal unloading of the LV. The Impella inlet was about 5.5 cm from the aortic valve. The TEE probe was then removed, and sedation continued for surgical closure. At the procedure’s end, the patient’s sedation was discontinued, and she emerged from anesthesia without sequelae. Four days later, she progressed to heart/lung transplantation en bloc and tolerated induction of a general anesthetic with the Impella 5.5 in situ, providing circulatory support.

Discussion

By avoiding the hemodynamic swings, positive pressure ventilation, and other complications associated with general anesthesia, MAC may be preferable in cardiovascularly fragile patients. ⁴ Especially for patients awaiting transplantation, avoiding prolonged intubation and subsequent deconditioning will improve long-term outcomes. ⁵ In contrast, regional anesthesia allows for surgical anesthesia, with minimal sedation, avoidance of instrumentation of airways, positive pressure ventilation, and opioid-sparing analgesia. However, regional anesthesia is not without its own difficulties.

Recently, Cohen et al demonstrated that surgical Impellas, such as the Impella 5.5, can be implanted using MAC and regional anesthesia. ¹ . In two of their presented cases, the Impella 5.5 was implanted using continuous infusions of dexmedetomidine and remifentanil in conjunction with strategic regional blocks (superficial cervical plexus, interscalene, and pectoralis II blocks). Many of the newer regional blocks require an expert in ultrasonography and placement to achieve an adequate block. Those trained in regional anesthesia are not always available to provide such blocks at all hours of the day or night. For example, our patient presented urgently over a weekend. The crux of the issue associated with regional anesthesia is that most Impella placements will have to occur at a quaternary medical center.

However, our anesthesiology group has demonstrated that centers can deliver a safe anesthetic with local anesthesia combined with sedation. Our patient tolerated the procedure with generous local anesthetic infiltration of the surgical site and intravenous analgesic adjuncts. Consideration of our approach to analgesia could be relevant in situations where the time or resources needed for regional anesthesia are unavailable. This could improve the number of centers placing the Impella 5.5 and treating those patients with refractory cardiogenic shock. Moreover, the use of MAC with local anesthesia can likely increase the number of patients progressing to bridge to recovery rather than transplant, as those who are awake without post-procedure intubation and sedation will maintain a sympathetic drive for cardiac and/or pulmonary recovery. Prolonged intubation post-procedure can be a less common complication. More importantly, once the procedure is complete, there is not much pain or stimulation from the Impella itself. This should hopefully promote patient mobility and avoid deconditioning as treating postoperative pain with intubation and sedation should not be needed. Pulmonary complications will also be avoided, especially those associated with intubation and positive pressure ventilation. While in our patient, we were able to successfully perform the procedure under MAC, not all patients may be candidates. Body habitus, history of obstructive sleep apnea, and airway concerns also need to be taken into account. Hypercarbia and hypoxia will not be favorable in the presence of right ventricular dysfunction. Continuous end-tidal carbon dioxide was monitored in our patient, and given the expertise of our providers, should airway obstruction have presented, we felt comfortable being able to deliver intermittent positive pressure breaths and reopen the airway as needed.

Also, unlike Cohen et al, we used TEE to confirm Impella 5.5 placement. The surgical Impella 5.5 requires confirmation of its proper placement with echocardiography, and while TTE or TEE is possible, most cardiac anesthesiologists are much more comfortable with TEE. ⁶ An anesthesiologist skilled in TEE can provide the data necessary to ensure proper placement and LV unloading. Cardiology does not need to be involved for a TTE as our anesthetic has shown that TEE can safely be performed under sedation in conjunction with surgical sedation. The cases from Cohen et al also serve as a reminder to consider TTE. Nevertheless, we have shown that it is also feasible and safe to perform TEE under sedation, to confirm correct Impella 5.5 placement. In our patient, our TEE exam under sedation contributed to preventing a risky general anesthetic and prolonged intubation. Our rationale for choosing TEE was the probability of obtaining better visualization and image resolution to confirm Impella placement, combined with the ergonomics of the fluoroscopy setup in our hybrid operating room. A pediatric probe was chosen over the standard probe because we believed its insertion would be less stimulating to the patients and, thus, better tolerated under the current level of sedation. We were mindful to minimize the risk of coughing and bucking, airway obstruction, hemodynamic disturbances, and aspiration.

Summary

MAC anesthesia with or without regional blocks is likely the future for Impella 5.5 placement. As the Impella 5.5 becomes the most adopted mechanical circulatory support, and Impellas combined with ECMO (ECMELLA) come to the forefront, we must adapt our anesthetic and surgical techniques to improve patient outcomes and throughput in this fragile patient population. The fact that general anesthesia can be replaced by simple sedation should also encourage all anesthesiologists at smaller medical centers to be comfortable with placement of the Impella 5.5. Ultimately, close cooperation and communication between anesthesia and surgical teams are paramount, combined with having a patient who is willing to undergo the procedure under sedation will make for successful outcomes. Future studies are needed to confirm the feasibility of this technique for a wider population.