Surgical Placement of Axillary Impella 5.5 with Regional Anesthesia and Monitored Anesthesia Care

Introduction

The Abiomed Impella 5.5^® with SmartAssist^® System is a short term use ventricular support device indicated for the treatment of ongoing cardiogenic shock (CS) unresponsive to optimal medical management. The device serves as a bridge to heart transplant, ventricular assist device (VAD), or cardiovascular recovery. Typically, the Impella 5.5 catheter is inserted via a surgical cut-down through the axillary artery and into the left ventricle through the aortic valve using fluoroscopic and transesophageal guidance. The axillary access is performed with a surgical cutdown over the second or third portion of the axillary artery. An 8 mm Dacron graft is then anastomosed to the artery in an end-to-side fashion to help facilitate the introduction of the Impella catheter. Generally, the right axillary artery is selected based on the location and caliber, however alternative locations include the left axillary artery, through an open chest innominate artery, or directly through a chimney graft in the aorta.^{1, 2} These approaches often necessitate general endotracheal anesthesia to provide an adequate plane of analgesia and stability to place the vascular graft.

For patients with CS the use of general anesthesia presents unique challenges due to the cardiopulmonary effects from the induction medications and as well as the institution of mechanical ventilation. In order to counteract the negative effects of induction and mechanical ventilation, high doses of vasopressors and inotropic medications are often necessary. More often than not, these patients remain intubated and sedated at the conclusion of the case due to the tenuous hemodynamic status. Transporting these patients to the intensive care unit (ICU) requires further titration and maintenance of vasoactive medications, sedatives, and pain medications, as well as delayed ventilator weaning and extubation at the discretion of the ICU.

An alternative approach to general endotracheal anesthesia (GETA) is to use monitored anesthesia care (MAC). MAC anesthesia can be combined with regional anesthetic techniques to create an alternative solution to Impella 5.5 placement. The overwhelming success that transcatheter aortic valve replacement (TAVR) programs have had in shifting the paradigm towards using sedation and local anesthesia for critically ill patients, poses the question: is there an opportunity to change the Impella 5.5 procedure to a similar model?

In this case conference, we present four cases where MAC with regional anesthesia was used to facilitate Impella 5.5 surgical procedures. Informed consent was obtained for publication of these reports and for use of the patient images.

Case Series Presentation

Impella 5.5 MAC with Regional Anesthesia protocol

The surgeons at our institution follow the Abiomed^® recommended protocol for Impella 5.5 placement (Supplement 1 & Figure 1).

Figure 1:

Axillary placement of Impella 5.5^®. Copyright Abiomed, Inc., reproduced with permission.

MAC and Regional Anesthesia Protocol

The patient selection for MAC with regional anesthesia for patients with CS undergoing semielective Impella 5.5 placement is based on the understanding of the patient’s medical condition. A few basic guidelines were used for patient selection for this approach. (Supplement 2). The peripheral nerve blocks used can be seen in Figures 2–4.

Figure 2:

Approach (left) and ultrasound image (right) demonstrating the interscalene block.

Figure 4:

Approach (left) and ultrasound image (right) demonstrating superficial cervical plexus.

Successful MAC with Regional Anesthesia

Case 1

A 55-year-old man with a body mass index (BMI) of 28 kg/m² was transferred to our institution for heart transplant evaluation after presenting to an outside hospital with an ST elevation myocardial infarction (STEMI). He has a past medical history significant for four vessel coronary artery bypass graft surgery (CABG) 11 years prior, heart failure with reduced ejection fraction (HFrEF) with an ejection fraction (EF) of 15–20%, history of ventricular tachycardia (VT) storm with implantable cardioverter-defibrillator (ICD), chronic kidney disease stage 3b, baseline creatinine 1.8mg/dL, and type 2 diabetes (admission A1C of 8.3%).

His right heart catheterization (RHC) demonstrated a right atrial pressure (RAP) of 13 mmHg, pulmonary artery pressure of (PAP) 65/30 mmHg, mean PAP of 42mmHg, pulmonary capillary wedge pressure (PCWP) of 35mmHg, systemic vascular resistance (SVR) of 1472 dynes*s/cm⁵, cardiac output (CO) 3.64 L/min, and cardiac index (CI) 1.82 L/min/m². A prior left heart catheterization (LHC) revealed 100% chronic total occluded venous grafts. He was started on milrinone which worsened his ischemic symptoms, so it was discontinued, and an intra-aortic balloon pump (IABP) was placed.

While undergoing his transplant work up, he was optimized on valsartan, metoprolol, aspirin, clopidogrel, atorvastatin, and ezetimibe, with the IABP augmenting at 1:1. The decision was made to place an Impella 5.5 for refractory cardiogenic shock for an INTERMAC grade II patient.

The patient was consented for monitored anesthesia care (MAC) and regional anesthesia and general anesthesia in case of conversion. In the operating room he was started on intravenous sedation with 0.5mcg/kg/hr of dexmedetomidine and 0.05 mcg/kg/min of remifentanil. 50 mcg/min of phenylephrine was used for hemodynamic support for the duration of the case. After achieving an adequate plane of anesthesia an ultrasound guided right interscalene block was performed with 5mL of 0.5% ropivacaine. Subsequently, a superficial cervical plexus block was performed using 5mL of 0.5% ropivacaine. Finally, a PECS II block was performed using a total of 25mL of 0.5% ropivacaine. 15 mL of 0.5% ropivacaine was placed between the pectoralis major and minor, and 10mL below the pectoralis minor.

The surgeons followed the standard protocol for Impella 5.5 placement and supplemented the regional anesthesia with a field block in the incision using 10mL of 2% lidocaine. The patient tolerated the surgical cutdown and graft placement successfully without requiring converting to general anesthesia. Using fluoroscopy, a 0.035” arterial wire was introduced through the graft and across the aortic valve. The Impella was then loaded on the wire and passed through the aortic valve into proper position. Once the Impella was deployed, a sterile transthoracic echocardiography (TTE) probe used to confirm the Impella position. The Impella was turned on at P4, the IABP was weaned and subsequently discontinued. At the conclusion of the case the sedation was weaned and immediately upon arrival to the ICU he was listed as status II for heart transplantation. Post operatively he was monitored for hemodynamic changes as well as surgical site bleeding. The anesthesiology team evaluated the patient post operatively for the nerve blook. Over the next 24 hours, the patient’s motor and sensory function returned to baseline in his right upper extremity. He was able to participate in physical and occupational therapy on post operative day 1. His pain was controlled with oral oxycodone 5mg q 8 hours. Ultimately, he received an orthotopic heart transplant 20 days following his Impella placement and did not require any mechanical ventilator support.

Case 2

A 58-year-old man with BMI of 28 kg/m² was transferred to our institution in worsening acute cardiogenic shock. His medical history was significant for type 2 diabetes, peripheral vascular disease with previous right common iliac artery angioplasty with stenting, right superficial femoral artery and popliteal artery percutaneous angioplasty, obstructive sleep apnea (OSA), gastroesophageal reflux disease (GERD), and lymphocytic colitis. He had initially presented to an outside hospital with a recent history of dyspnea on exertion and worsening lower extremity edema. He was found to have triple vessel CAD and underwent a four vessel CABG. A right axillary Impella 5.5 was placed for further hemodynamic support. His post-operative course was complicated by significant coagulopathy requiring intraoperative washout, and resuscitation including intravenous vitamin K and a dose of factor VIIa. On post operative day 4, he developed an axillary artery thrombosis, requiring a surgical thrombectomy via right brachial cut down and the removal of the Impella. Given his peripheral vascular disease he was not a candidate for an IABP.

With the loss of hemodynamic support, he was started on epinephrine and milrinone infusions, but his hemodynamics and markers of end organ perfusion continued to worsen with creatinine increasing to 1.9 mg/dL. A TTE at that time showed an EF of 5–10%. Soon thereafter, he went into supraventricular tachycardia and atrial fibrillation, and was started on amiodarone and transferred to our institution.

On admission to our ICU, computed tomography (CT) imaging of the chest demonstrated a large left layering pleural effusion. His mixed venous oxygen saturation (SVO₂) was 32% and he had a rising lactate of 2.3 mmol/L. TTE showed a mildly dilated left ventricle with EF of 10–15%, moderately to severely decreased right ventricular systolic function, moderate mitral regurgitation, and moderate tricuspid regurgitation. His milrinone was increased to 0.5mcg/kg/min and furosemide was initiated.

With limited options, the surgical team decided to place a left axillary Impella 5.5. In the setting of his severe right and left heart failure he agreed to regional anesthesia. He was consented for MAC with regional anesthesia. In the operating room he was started on epinephrine 4mcg/min, his milrinone was decreased to 0.375mcg/kg/min, and the amiodarone remained at 0.5mg/min. Dexmedetomidine was started at 0.5mcg/kg/hr and remifentanil 0.04 mcg/kg/min. He received 0.5mg of midazolam in preparation for his peripheral nerve blocks.

A PECS I block with 20ml of 0.5% ropivacaine was placed under ultrasound guidance. Then, an ultrasound guided superficial cervical plexus block was placed with 10 ml of 0.5% ropivacaine. With his large left pleural effusion and OSA, the interscalene block was avoided to spare his diaphragm.

During case, the dexmedetomidine was increased to 0.6mcg/kg/hr and remifentanil was titrated to 0.03mcg/kg/min. The sedation caused some hypotension, and vasopressin was started at 0.02 units/min. Once surgery commenced, he felt some discomfort during the deep dissection which improved with surgical supplementation with 2% lidocaine, 1mg of midazolam and 50mcg of fentanyl. He tolerated the surgical cutdown, graft placement, and remained hemodynamically stable through the case. The arterial wire for the Impella was guided using Carm fluoroscopy for placement across the aortic valve. Once the Impella was inserted, a sterile TTE probe was placed on the prepped chest and the Impella was interrogated from the parasternal long axis view and apical four chamber view, which confirmed the final position of the Impella. The Impella was turned on at P6. He was taken back to the ICU without requiring mechanical ventilation or general anesthesia. He had limited incisional pain in the axillary site which was controlled with 5 and 10mg doses of oxycodone as needed. As soon as the Impella 5.5 was placed his hemodynamics improved. He participated in physical therapy on post-operative day 1. He was transferred to the cardiovascular medicine floor from the ICU on post-operative day 7. The Impella 5.5 served as a successful bridge therapy to a Heartmate 3 left ventricular assist device for 18 days.

Impella 5.5 Decannulation under MAC and regional anesthesia

Case 3

A 54-year-old man with BMI 25 kg/m² was emergently transferred to our institution due to cardiogenic shock after receiving his first round of fluorouracil for stage IV esophageal cancer. In addition to his esophageal cancer with metastases to the stomach, liver, and lungs, he had a past medical history significant for upper extremity deep vein thrombosis (DVT) for which he was treated with apixaban.

Approximately an hour following his chemotherapy session, he became diaphoretic with severe angina and presented to the emergency room. His electrocardiogram at that time demonstrated anterior wall ST depressions. Serial TTEs were concerning for a worsening EF which began at 35% and downtrended to 10–15% over 24 hours with anterior wall and apical akinesis. A RHC showed a PCWP of 37 mmHg, CI 1.1 L/min/m², and a SVO₂ of 35%. He became increasingly hypotensive requiring escalating doses of norepinephrine, and the decision was made to place a right femoral Impella CP for hemodynamic support. He was also initiated on uridine triacetate for the suspected fluoropyrimidine toxicity.

Upon arrival to our institution, he showed signs of progressive decompensation with rising lactate and persistent hypotension and tachycardia, requiring escalating doses of inotropic and vasopressor support. His Impella CP^® was flowing 2.7L/min. A multidisciplinary discussion including cardiothoracic surgery, oncology, palliative care, anesthesiologists, and the ICU teams discussed the poor prognosis and using mechanical circulatory support as a bridge to an alternative chemotherapeutic regimen. After discussing this with the patient, he was taken to the operating from for conversion from Impella CP to Impella 5.5.

A left axillary Impella 5.5 was placed under general anesthesia. Due to his esophageal cancer, a TEE was not performed, and instead, a sterile TTE was done to provide final Impella positioning. The TTE showed LVEF 8%, no left ventricular thrombus, a right ventricle that was normal in size with mild to moderately decreased systolic function, and normal atria. With the Impella at P6 and milrinone at 0.25mcg/kg/min, his RAP was 3 mmHg, PAP 19/6 mmHg (mean 10mmHg), SVR 1912 dynes*s/cm⁵, SvO2 44.6%, and CI 1.68 L/min/m². He was taken to the ICU and extubated 6 hours later. His serum chemistries demonstrated creatinine decreasing from 1.9 to 0.6 mg/dL.

Over the next 5 days, his heart function improved, and he was weaned off of the Impella. For removal of the Impella and closure of the axillary graft he agreed to a regional anesthetic. After receiving 2mg of midazolam, three peripheral nerve blocks were placed. His regional anesthesia consisted of an ultrasound guided PECS II block with 15ml of 0.25% bupivacaine total, an ultrasound guided interscalene block with 10ml of 0.25% bupivacaine, and an ultrasound guided superficial cervical plexus block using 5ml of 2% lidocaine.

During the case, his anesthetic consisted of 80 mcg/kg/min of propofol and 0.03 mcg/kg/min of remifentanil. Norepinephrine was started at a dose of 2 mcg/min. He received 100 mcg of fentanyl and 1mg of hydromorphone for the entirety of the case and a total of 428mg of propofol, and 0.23mg of remifentanil. The axillary graft site was re-opened, Impella cannula removed, the graft was removed, and the incision was closed in several layers. His closing hemodynamics demonstrated a CI of 2.8 L/min/m² and his TTE showed an EF of 70% with no regional wall motion abnormalities with a normal RV. Post-operatively he had no issues with upper extremity pain and was discharged two days after Impella decannulation.

Failed Impella Placement with MAC and Regional Anesthesia, Stable Conversion to General Anesthesia

Case 4:

A 47-year-old man with a BMI of 24 kg/m² presented to the emergency department with Dressler’s syndrome and “crushing chest pain”. His medical history included a STEMI two months prior to admission with percutaneous coronary intervention (PCI) to the first obtuse marginal and the left anterior descending artery (LAD), polymorphic VT, and type 2 diabetes (A1C 7.7%). Due to his HFrEF with EF 15–20% and LifeVest^®, he was listed as Status 4 for heart transplant.

On admission, his TTE demonstrated global hypokinesis, an EF 15–20%, a moderately dilated and depressed right ventricle, moderate mitral regurgitation, and a moderate circumferential pericardial effusion. Due to the refractory CS, he was started on 0.25 mcg/kg/min of milrinone. His RHC demonstrated RAP of 1 mmHg, PAP of 49/23 mmHg (mean 33 mmHg), PA sat = 65 %, PCWP 16 mmHg, PVR 4.2 Wood Units, and SVR of 1,622 dynes*s/cm⁵, a cardiac output of 4.0 L/min and an index of 2.1 L/min/m². LHC did not demonstrate any lesion requiring intervention. He was medically optimized on aspirin, sacubitril/valsartan, clopidogrel, rosuvastatin and milrinone as he was being worked up for ventricular assist device or heart transplant. He received two separate thoracenteses for symptomatic pleural effusions, his sacubitril/valsartan was held for low SVR, and was being diuresed heavily. His end organ perfusion remained normal throughout his wait for transplant approval. His liver enzymes began to slowly rise and clinically, he appeared weaker and ashen, so the milrinone was increased to 0.5mcg/kg/min. He was quickly listed for Impella 5.5 placement, and he agreed to regional anesthesia for the procedure.

In the operating room, he was started on 0.5 mcg/kg/hr of dexmedetomidine and his milrinone infusion was continued at 0.5mcg/kg/min. For the nerve blocks he was given 1mg of midazolam and 100mcg fentanyl in divided doses. He was prepped and draped in the usual manner and an ultrasound guided right sided interscalene block was placed using 10ml of 0.5% ropivacaine. A superficial cervical plexus block was then performed using 10ml of 0.5% ropivacaine. Lastly, a PECS I block was placed with 10ml of 0.5% ropivacaine.

30 minutes later after the case began, the patient was unable to tolerate the procedure due to discomfort and wanted to be put to sleep instead after being offered more sedation and pain medication. Norepinephrine and epinephrine infusions were started at 6mcg/min, and he was induced and intubated uneventfully with 6mg of etomidate and 70mg of rocuronium with the remifentanil and dexmedetomidine infusions continued.

The remainder of the case was uncomplicated. The patient was extubated successfully with milrinone set at 0.375mcg/kg/min and the Impella at P5. He was approved for transplant that evening, listed as Status 2.

Over the next 24 hours the patient complained of worsening right arm pain, requiring several additional doses of hydromorphone. His pain regimen was increased to scheduled 10mg oxycodone every 6 hours, as needed doses of hydromorphone, and scheduled gabapentin 200mg every 8 hours. By post operative day 2 his arm pain had significantly improved, and he was able to have the narcotics weaned in 48 hours. As the pain regimen taper continued, he developed worsening neuralgia of the operative side. On post procedure day 3 he began to complain of right hand numbness and tingling that evolved into significant right upper extremity weakness limiting his ability to abduct his arm or hold it antigravity. He described intermittent cramping pain and spasms in his forearm and pressure in his right chest near the incision.

He was taken for a CT scan 4 days post procedure for worsening pain at the Impella site and weakness. The CT demonstrated postoperative hemorrhage deep to the insertion site and small hematoma in the right pectoralis minor. The moderate amount of hematoma around the Impella site effaced the course of the distal divisions to branches of the brachial plexus. This was managed conservatively and gradually improved. He was transplanted 11 days later.

Discussion

For patients in CS, the in-hospital mortality is estimated to be around 50 – 80% with a 1-year mortality ranging from 50% to 63%.^3–9 Regardless of management, be it maximal medical therapy or mechanical circulatory support (MCS), progressing a patient through their hospital stay as effectively, efficiently, and quickly as possible is of critical importance to shorten their time to recovery.

Accelerating time to recovery is central to the Enhanced Recovery after Surgery (ERAS^®) care pathway which focuses on improving perioperative management strategies to expedite an early recovery for patients undergoing major surgery. ERAS^® “re-examines traditional practices, replacing them with evidence-based best practices when necessary [and] is comprehensive in its scope, covering all areas of the patient’s journey through the surgical process.”¹⁰ In the same way, employing regional anesthesia for surgical Impella 5.5 placement is a step towards expediting an early recovery for patients by avoiding the additional vasoactive medications, sedatives, and time on the ventilator.

According to a 2022 study by Rock et al., the average time to extubation was 5 (±5) days post-Impella 5.5 implantation, and all patients that required mechanical ventilatory support before Impella placement were able to be extubated post implant. In their population, 70% of the patients were bed-bound, and following Impella 5.5, of these bed-bound patients, 39% were walking or mobile to bedside chair; following explant, 50% were walking or mobile to chair.⁶ Avoiding a general anesthetic for Impella placement circumvents the delay to mobility and activity, the additional sedation and narcotics to facilitate ventilator comfort, and allows patients to progress in their care. Moreover, avoiding GETA and additional sedatives allows the patient a quick recovery to baseline cognitive function after the intraoperative sedation is stopped. This is important as it allows for immediate neurologic monitoring post operatively to evaluate for strokes (a known risk associated with MCS) and allows the patient to participate directly in the transplant work up and discussion, possibly resulting in a faster time to transplant listing.

The Impella 5.5 is an important advancement in CS patients as it not only provides up to 5.5–6 L/min of cardiac output, but its design has a patient’s recovery in mind. MCS devices such as the IABP or veno-arterial extracorporeal membranous oxygenation (VA ECMO) circuits limit a patient from moving and bending, participating actively in physical therapy, and are not considered therapies approved for the standard step-down unit. The Impella 5.5’s axillary cannulation allows patients to get out of bed and participate in physical and occupational therapy¹¹. The ability to do these things prevents delirium and further deconditioning which occurs from consecutive days being confined to a hospital bed.

The paradigm shift from using general anesthesia to regional anesthesia and intravenous sedation in critically ill patients has been successful most notably in transcatheter aortic valve replacement (TAVR) procedures. Conventionally, transfemoral TAVRs were performed under general anesthesia with TEE to confirm valve positioning, function, and transvalvular gradients. As data continued to demonstrate the success of TAVRs using local anesthesia with or without conscious sedation, a new standard emerged in cardiothoracic surgery and interventional cardiology. The 2020 randomized multicenter SOLVE-TAVI trial, demonstrated that patients who received conscious sedation required significantly less inotropes (p<0.001) compared to those who were placed under general anesthesia.^{12, 13} Using less inotropes and vasopressors not only speaks to the hemodynamic stability of the anesthetic techniques but also avoids the possibility of inducing unstable arrythmias and the toxicity associated with high dose vasoactive medications.

In 2020 Butala et al. examined an American multicenter TAVR registry of 120,080 patients from 2016 to 2019. They noted a small but significant decrease in in-hospital mortality (0.2%; p = 0.01) and 30-day mortality (0.5%; p < 0.001) when comparing local anesthetic with conscious sedation to general aensthesia.^{12, 14, 15} Additionally, Hyman et al. showed conscious sedation to be associated with both shorter lengths of stay and lower in-hospital and 30-day when compared to TAVRs with GA, which, added to the mortality benefit, makes a regional based technique an attractive option for patients in CS.¹⁵

Translating TAVR and other major vascular procedures that use sedation and regional anesthesia to Impella 5.5 placement required integrating our institution’s sedation strategy for TAVR with peripheral nerve blocks to cover the chest wall, clavicle, and lower neck to facilitate the deep axillary cut down and retraction needed for Impella 5.5. To adequately provide analgesia for the cut down, retraction, and insertion of the Impella 5.5, the sensory innervation to the anterior and lateral chest wall, shoulder, and clavicle needs to be sufficiently anesthetized. Several types of nerve blocks have been used successfully in ICD placement and Impella 3.5 placement, to include various combinations of PECS I, PECS II, superficial cervical plexus, and transversus thoracis muscle plane blocks.¹⁶

The chest wall and pleura are innervated by branches of the intercostal nerves. The anterior and lateral cutaneous branches of T2-T6 nerves, pectoral nerves, long thoracic nerves, and supraclavicular nerves innervate the anterolateral chest wall specificially.¹⁷ The pectoral nerve block, specifically, the PECS II block, provides comprehensive coverage of the area. The PECS II adds a second pectoral fascial plane block to the PECS I to cover the anterior cutaneous branches of intercostal nerves 3 to 6, the intercostobrachial nerves, and the long thoracic nerve. The PECS I anesthetizes the layer between the pectoralis major and minor, covering the medial and lateral pectoral nerves, which innervate the pectoralis muscles. The second injection is lateral to the PEC I injection point and lies between the pectoralis minor and serratus anterior muscles. The added injection blocks the upper intercostal nerves, the lateral branch of the T2–4 spinal nerves, and sometimes the anterior branch of T2–4 spinal nerves. Additionally, the PECS II block spread can cover the long thoracic nerve (C5–7) which arises from and innervates the serratus anterior muscle, and the thoracodorsal nerve (C6–8), which innervates the latissimus dorsi muscle.¹⁸

To cover the upper arm and clavicle, the interscalene block (which covers C5 and C6) is added. The medial (C8,T1) and lateral (C5-C7) cords of the brachial plexus supply the medial and lateral pectoral nerves which innervate the pectoralis major and minor.¹⁷ Therefore, the interscalene serves as an added layer to the PECS II in addition to covering the upper arm. Lastly, blocking the superficial cervical plexus covers the supraclavicular nerve to assist in pain around the clavicle and neck.

The four cases presented provide four different strategies at anesthetizing patients for Impella 5.5 placement, both in terms of regional blocks and sedation techniques employed. As the protocol for these cases evolves, a few key points stand out from the case series. The first is the added benefit of the PECS II block. As demonstrated in our case series, conversion to general anesthesia is a risk, but the risk is low. Conversion to GA in the SOLVE-TAVI trial occurred in 5.9% of patients and 7.9% in Villablanca et al.’s 2018 meta-analysis.¹⁹ In our case, we suspect that using a PECS II instead of a PECS I block would have prevented to conversion from regional to GA. Secondly, several of the patients required additional boluses of sedation medications during the case. This is likely due to the dynamic nature of the Impella insertions. Most of the case is consistently low stimulation, with incision, electrocautery, and dissection to the axillary artery. During the dissection depending on the depth of the artery, muscle and fat composition of the patient, and ease of graft placement there can be several minutes of startling forceful retraction and pushing. If the sedation is tailored to the less stimulating moments, patients can find the sudden aggressive sensation unnerving. Selecting a deeper plane of sedation while maintaining spontaneous ventilation, and anticipating and warning the patient of periods of intense retraction pressure can prevent additional boluses. Lastly, the incidence of post operative pain occurring within the first 24 hours post procedure is likely due to the patients’ care team not being advised and instructed on how to tailor long-acting pain medications while the nerve block wears off. The phenomenon of rebound pain can occur when the peripheral nerve block wears off, and the untreated acute post-surgical pain sets in. This rebound pain oftentimes requires more frequent and higher doses of narcotics, counteracting the narcotic sparing nature of nerve blocks.^20–22 To prevent the pain of the nerve block wearing off and the potential rebound pain, these patients should continue long acting narcotics in addition to other multimodal analgesic strategies even if they do not complain of pain. Moving forward, care teams will be educated on effective perioperative pain management.

This case conference highlights the potential of using regional anesthesia and sedation to facilitate the Impella 5.5 placement. Using the experience from TAVR, regional cases for vascular surgery, and our limited case series, a more robust and comprehensive protocol can be developed to successfully employ this technique. By avoiding a general anesthetic all together, patients can hopefully afford a shorter time to rehabilitation, listing for transplant, and an improved hospital course.

Expert Commentary – Are Regional Techniques for Impella 5.5 the Answer?

Elana Folbe, MD and Joseph Sanders, MD, FASA

Advancements in cardiovascular medicine have made it increasingly common to perform more complex procedures. However, administering anesthesia to patients with significant cardiovascular disease, as seen in this case conference, presents specific challenges. Many of the medications used in general anesthesia can depress both cardiac and pulmonary function. Additionally, intubation, which is an invasive procedure, carries risks for patients with impaired pulmonary function. As a result, there has been a shift towards incorporating regional anesthesia, either as a primary anesthetic or as an adjunct, into standard practice. In cardiac surgery patients, regional anesthesia has demonstrated several benefits, including reduced opioid requirements, facilitated early extubation, decreased postoperative complications, shorter stays in the intensive care unit and hospital, and lower healthcare costs.²³ Given the high risk of decompensation in patients with severe cardiovascular disease, it is crucial to approach anesthetic selection for high-risk procedures with an individualized, multidisciplinary approach. Although evidence for the efficacy and safety of regional anesthesia as the primary anesthetic for ventricular assist devices is limited, this case conference underscores both the risks and benefits associated with using regional anesthesia as the primary method for patients in cardiogenic shock undergoing axillary Impella 5.5 insertion.

Studies have indicated that patients in cardiogenic shock face a significantly higher risk of morbidity and mortality compared to other patients undergoing general anesthesia.²⁴ The cardiovascular effects and the unpredictable onset of induction agents can have a profound impact. In patients with low cardiac output, it may take longer for intravenous medications to reach the brain and produce sufficient amnesia.²⁵ Induction agents can also induce significant vasodilation and hypotension, exacerbating preload, cardiac output, and end-organ perfusion. Furthermore, many intravenous anesthetic agents can cause dose-dependent respiratory depression, which may be undesirable for patients prone to developing pulmonary edema. Although none of the patients in this case conference required intubation due to respiratory distress or depression, it is not specified whether they had comorbid pulmonary disease or had been optimized before undergoing Impella placement.

There are additional factors to consider when utilizing regional anesthesia for patients in cardiogenic shock. Regional anesthesia carries the risk of adverse events, including inadvertent intravascular injection, local anesthetic toxicity, hematomas, infection, and nerve injury. In patients with cardiogenic shock, local anesthetic toxicity or intravascular injection could lead to life-threatening cardiovascular collapse. It is crucial to discuss these potential complications with patients, particularly those with established peripheral neuropathies from conditions such as diabetes mellitus or peripheral vascular disease.

In this case conference, the authors discuss the utilization of PECS I and PECS II, interscalene, and/or superficial cervical plexus blocks. When employing multiple blocks, it is crucial to be attentive to the maximum doses of local anesthetics to avoid adverse effects. Furthermore, patients in cardiogenic shock frequently have pulmonary sequelae or underlying pulmonary conditions. When administering regional anesthesia with sedation, it is important to consider that the patient will have an unprotected airway, which elevates the risk of aspiration and respiratory depression.

As highlighted in the case conference, PECS I and PECS II blocks are interfascial plane blocks commonly utilized in breast surgery to block specific branches of intercostal nerves, the long thoracic nerve, and the thoracodorsal nerve. The literature extensively describes these blocks in relation to breast surgery. The PECS I block targets the lateral and medial pectoral nerves as they traverse the pectoralis minor, while the PECS II block involves a combination of subpectoral and interpectoral approaches, effectively blocking the second through sixth intercostal nerves, long thoracic nerve, and thoracodorsal nerve.²⁶ It is important to note that performing PECS blocks can be challenging in patients with significant morbid obesity or substantial breast tissue, as these cases may require a larger volume of more dilute local anesthetic for adequate spread. For Impella insertion, a PECS I block might suffice if the proceduralists do not manipulate the ribs, latissimus dorsi muscle, serratus anterior muscle, or distal axilla. By avoiding a PECS II block, a lesser amount of local anesthetic can be utilized for a procedure involving multiple blocks. While most literature reports employ PECS blocks as adjuncts to general anesthesia, there are case reports demonstrating successful outcomes using PECS blocks and sedation for breast surgery.²⁷ However, the available evidence suggests that when performing chest wall surgeries like Impella insertion, PECS blocks should be combined with additional blocks for optimal results.²⁸

In this case conference, interscalene blocks were included as part of the primary anesthetic approach for Impella insertion. Interscalene blocks are typically used for surgeries involving the shoulder, upper arm, and clavicle. During the procedure, the local anesthetic is administered to the space between the anterior and middle scalene muscles, bathing the superior and middle trunks of the brachial plexus. It is important to note that when performing an interscalene nerve block, there is almost always a complete blockade of the phrenic nerve,²⁹ which innervates the diaphragm and plays a crucial role in respiratory function. This becomes a particular concern for patients with pre-existing pulmonary disease, as their respiratory function may be compromised due to the phrenic nerve blockade.

The second patient in the case series, who had a pleural effusion and obstructive sleep apnea, did not undergo an interscalene block due to these specific conditions. Instead, a superficial cervical plexus block was performed. The use of cervical plexus blocks has been extensively studied in patients undergoing carotid endarterectomies (CEAs).³⁰ These patients often have comorbidities such as hypertension, diabetes, and coronary artery disease, which increase their risk of cardiac complications.

Several meta-analyses have examined the outcomes of local anesthesia versus general anesthesia in patients undergoing CEAs. These studies have generally concluded that regional anesthesia, including the use of local anesthesia, is associated with improved outcomes, including reduced incidences of stroke, myocardial infarction, and death.^30–32 However, it is important to note that a more recent meta-analysis conducted by Amer Harky et al. in 2020 showed only a small, significant benefit of using local anesthesia. These findings highlight the ongoing debate and varying conclusions in the literature regarding the benefits of regional anesthesia compared to general anesthesia in specific surgical procedures, such as carotid endarterectomies. The case series discussed in this context sheds light on the emerging debate regarding the choice of anesthesia for axillary Impella procedures, indicating that further research and exploration are needed in this area.

It is important to note that a cervical nerve block can also cause phrenic nerve blockade. A small, randomized control trial by Opperer et al. aimed to quantify the degree of phrenic nerve blockade and hemi-diaphragmatic motion after superficial, intermediate, and deep cervical plexus blocks.³³ The study showed that superficial cervical plexus blocks cause a decrease in diaphragmatic movement in all three blocks (1.00±0.93 cm, 1.60±0.75 cm, and 1.62±0.50 cm, for deep, intermediate, and superficial groups, respectively, p=0.047). Even though the study was underpowered, with 15 patients in each group, the authors commented that deep cervical plexus blocks exhibited the greatest decrease in diaphragmatic movement. Thus, during Impella insertions for patients with compromised pulmonary physiology, it would be advised to avoid both the interscalene and cervical plexus blocks. Instead, one could perform a clavipectoral block,³⁴ which blocks the distal one-third of the clavicle, with a PECS block to provide appropriate anesthesia for Impella insertion.

In addition to the evidence presented in the case conference, the authors also draw upon literature related to TAVR to support the use of regional anesthesia for Impella placement. Studies in TAVR have shown that patients who undergo the procedure under conscious sedation experience reduced use of inotropes, shorter hospital stays, and a slight decrease in in-hospital and 30-day mortality rates compared to those under general anesthesia.^{14, 15, 35}

However, it is important to note that there are significant differences between the patient populations undergoing TAVR and those discussed in this case conference. TAVR is typically an elective procedure, and the patients are generally hemodynamically stable. On the other hand, the patients in the case conference are in cardiogenic shock and already receiving vasoactive medications. Given their critical condition, these patients may potentially benefit more from avoiding general anesthesia. However, to establish the superiority of regional anesthesia over general anesthesia in Impella placement, larger registries incorporating protocolized regional and sedation techniques would be required. This highlights the need for further research and well-designed studies specifically focusing on Impella procedures in patients with cardiogenic shock to better understand the optimal approach to anesthesia in this population.

Lastly, effective communication among the anesthesiologist, surgeon, and patient is crucial when considering the use of regional anesthesia for Impella insertion. Patients requiring Impellas, especially those discussed in this case conference, may perceive this surgery as the most stressful and significant event in their lives. Pre-operative anxiety is highly prevalent, affecting up to 80% of patients, and is often rooted in fear of surgery, anesthesia, complications, and the recurrence of previous complications.³⁶ While the four patients in this conference willingly participated in the new approach involving regional anesthesia for Impella insertion, it is important to acknowledge that patients with predisposed anxiety or fears may not tolerate this procedure with regional anesthesia and sedation. In cases where patients are unable to tolerate the procedure, close communication between the surgeon and anesthesiologist is vital so that anesthesia can be deepened during specific stages of the procedure. Patient movement during a surgery, especially one involving direct insertion into the heart, could have catastrophic consequences. Therefore, obtaining informed consent for this procedure with regional anesthesia necessitates ensuring that patients have a complete understanding of the implications and potential experiences during an awake procedure. As the use of regional anesthesia continues to rise in various surgical specialties, maintaining a patient-centered approach is essential in our practices.

The authors of this case conference have demonstrated that regional anesthesia for Impella 5.5 insertion can be a potentially successful and safe alternative to general anesthesia. Although the case series was small and utilized different combinations of regional techniques, further investigation is warranted. Institutions would require anesthesia providers with expertise in ultrasound and regional anesthesia to ensure safe and effective implementation. It is worth exploring alternative blocks or combinations of blocks that could minimize the risk of pulmonary complications, local anesthetic toxicity,³⁷ and peripheral nerve injury. Given that the patients undergoing these procedures are in cardiogenic shock and likely have multiple comorbidities, the anesthetic approach for these high-risk procedures should be determined on a case-by-case basis, taking into account hemodynamic stability, comorbidities, and the individual patient’s disposition.

Expert Commentary – Adding Regional Anesthesia to the Toolbox for Impella 5.5 Insertion

Rohesh J. Fernando, MD, FASE, FASA

In this manuscript, the authors describe four cases whereby insertion of the Impella 5.5 was attempted via the axillary approach under MAC with regional anesthesia. Three of the four cases were successfully completed without the need for conversion to general anesthesia. The type of regional block varied by each case, and the authors suspect that the PECS II block can play a valuable role in assuring patient comfort. Overall, these cases contribute potential strategies to managing anesthesia for trans-axillary Impella placement.

The Impella 5.5 was approved for use in the United States in 2019. Its use has continued to grow, and it may be particularly helpful for patients as a bridge to transplantation. In a retrospective study by Cevasco et al, 464 patients were identified as having an Impella 5.5 at some point during being waitlisted for heart transplantation.³⁸ Of these patients, 265 patients had an Impella when they were listed. A total of 402 patients ultimately underwent heart transplantation, and 387 were directly bridged from the Impella 5.5 with a median duration of implantation of 16 days. The most common complications after heart transplantation included kidney failure with dialysis (15.7%) and stroke (4.0%). Interestingly, the duration of Impella implantation was not determined to be a risk factor for either of these outcomes. In addition, these patients had good survival at one year (89.6%).³⁸

As the rate of Impella 5.5 implantation increases, the ability to employ techniques such as regional anesthesia could be useful to offer an individualized anesthetic plan to patients. In 2020, Raghunathan et al. described a case report in which an Impella 3.5 was inserted using a regional anesthetic technique.¹⁶ The authors stated that they used a combination of PECS I, PECS II, a transversus thoracic muscle plane block (TTMPB), and a superficial cervical plexus block. A combination of 6mL 0.25% bupivacaine and 2mL lidocaine (concentration not specified) was used for each block. Certainly, one must be judicious with dosing of the local anesthetic to limit toxicity when performing multiple blocks, and although the weight of the patient was not provided, the authors stated they considered this.¹⁶

It is interesting to consider how many blocks are truly required for trans-axillary Impella placement. In the case mentioned previously, Raghunathan et al. described using four blocks. However, the PECS I and PECS II were listed as separate blocks, even though the PECS II block includes the PECS I; although the PECS II requires injection at two different sites, they can usually be performed with a single needle insertion.³⁹ As a result of this confusion, the American Society of Regional Anesthesia and Pain Medicine and European Society of Regional Anesthesia and Pain Therapy have suggested renaming of the PECS I block to “interpectoral plane block” and the deeper injection of PECS II to “pectoserratus plane block.”⁴⁰ Therefore, rather than referring to a PECS II block, one would instead describe the performance of both the interpectoral and pectoserratus plane blocks.⁴⁰

If the PECS II is considered one block, the number of blocks utilized in this case conference ranged between two and three (Table 1). Notably, PECS II was not utilized for the second and fourth patient. The second patient was able to complete the procedure without conversion to general anesthesia, although the patient did have discomfort that required injection of local anesthetic by the surgical team. The fourth patient received similar blocks as the second patient, but also received an interscalene block; despite the additional block, conversion to general anesthesia was required. The authors surmise that general anesthesia may have been avoided if a PECS II block had been performed. It is also interesting to note that the second case was successfully performed without the interscalene block; as mentioned prior, however, it is possible that the field block performed by the surgical team could account for this. Perhaps this also illustrates the point that the ability of the surgeon to supplement with local anesthesia can be a valuable tool. However, close communication between the anesthesia and surgery teams is important to ensure the maximum dose of local anesthetic is not exceeded.

Table 1:

Regional Anesthetic Technique by Case

Block/Case	PECS I (without II)	PECS II	Interscalene	Superficial Cervical Plexus	Additional Local from Surgeon	Conversion to General Anesthesia
Case 1		x	x	x	x	No
Case 2	x			x	x	No
Case 3		x	x	x		No
Case 4	x		x	x		Yes

Perhaps some insight can also be gained from the use of regional anesthesia for TAVR with an axillary/subclavian approach. Alexander et al. reported the successful combination of regional anesthesia and sedation with only a PECS I block.⁴¹ The authors believed it to be important to use at least 20mL local anesthetic to ensure adequate block duration and coverage. The PECS I block was performed with 25mL 0.5% bupivacaine, and the procedure was completed successfully. The authors also mentioned two other cases in which an interscalene block was used in addition to the PECS I for trans-subclavian TAVR; interestingly, they believed that the interscalene block may not be necessary and suggest against its performance to spare the phrenic nerve.⁴¹

Block et al. described 3 cases in which interscalene and PECS II blocks were performed for TAVR via the left subclavian artery.⁴² None of the patients required general anesthesia. One patient experienced neurologic symptoms, but this may be explained by the inadvertent inclusion of the brachial plexus by the vascular clamp during surgery. The authors found the interscalene block to help in cases where a more complex surgical dissection is needed and also to minimize arm movement that could interfere with surgery.⁴²

Toscano et al. described two cases in which the PECS II block was used for trans-subclavian TAVR.⁴³ The pectoserratus plane block was performed with 20mL 0.5% ropivacaine, and the interpectoral plane block was done with 10mL 0.5% ropivacaine. Both patients successfully underwent TAVR with sedation.

Similarly, Kanei et al. enrolled 221 consecutive patients undergoing TAVR, 25 of whom underwent the transaxillary approach.⁴⁴ 24 patients successfully completed transaxillary TAVR with MAC and the PECS I block, with one patient receiving general anesthesia. One patient had temporary numbness of the hand, but this spontaneously improved. Overall, while the studies described above are small, they provide support that trans-axillary and/or trans-subclavian TAVR can be successfully performed with regional anesthesia and sedation.

Whether it be TAVR or Impella placement, use of regional anesthesia rather than general anesthesia will likely have other implications beyond the anesthetic. For example, echocardiography is typically used to access device placement. Prior to the procedure, it should be verified that the patient has adequate windows for transthoracic echocardiography. If none of the clinicians in the operating room are proficient in obtaining these images, then the availability of a sonographer may be an additional consideration. Also, conversion to general anesthesia could be challenging given the proximity of the surgical site to the airway; patients with a challenging airway at baseline could be particularly at risk for difficulty in securing the airway. The authors of this case conference seem to be aware of these issues, as they astutely outlined patient selection criteria in Supplement 2.

Overall, this case conference highlights a useful technique whereby regional anesthesia can be used to facilitate placement of a trans-axillary Impella. Patient selection is important to minimize the risk of unexpected complications. Prior literature on regional anesthesia for trans-axillary TAVR can also help to inform anesthetic management for transaxillary Impella, and vice versa. While it is not completely clear the exact number of blocks or which blocks are absolutely required, these cases provide clinicians with a foundation for additional research.

Figure 3:

PECS II Block - needle approach (left) and ultrasound images (middle: injection between pectoralis major and pectoralis minor, right: injection between pectoralis minor and serratus anterior).