Ivor Lewis esophagectomy in a patient with left ventricular assist device

Robotic-assisted esophagectomy in an LVAD-supported patient.

Central Message.

Esophagectomy can be safely performed in selected patients with durable LVAD support using a physiology-driven approach addressing preload dependence, OLV tolerance, and conduit perfusion.

With advances in device technology and resultant survival, patients with durable left ventricular assist devices (LVADs) are increasingly presenting with indications for major noncardiac surgery, including malignancies.¹ In these patients, cardiac output is preload-dependent, and right ventricular (RV) function critically influences systemic perfusion. Esophagectomy has historically been considered prohibitively high-risk in this population because of concerns regarding tolerance of one-lung ventilation (OLV), gastric conduit perfusion, and anastomotic healing under LVAD physiology. We report the first successful esophagectomy in a patient supported by a durable LVAD and outline a physiology-driven approach that enabled safe resection and reconstruction. Institutional review board approval was waived, and patient provided informed consent for the publication of study data.

Case Presentation

A 74-year-old man with a HeartMate 3 LVAD (Abbott Laboratories) as destination therapy for ischemic cardiomyopathy and an implantable cardioverter-defibrillator presented with gastrointestinal bleeding and symptomatic anemia requiring transfusion. Anticoagulation was held, and he underwent endoscopy, which revealed an esophageal adenocarcinoma 35 cm from the incisors with an adherent clot (Figure 1). Positron emission tomography-computed tomography findings demonstrated a fluorodeoxyglucose-avid distal esophageal lesion without nodal or distant metastatic disease.

Figure 1.

Endoscopic image of esophageal cancer.

During multidisciplinary cancer conference including heart failure cardiology, cardiovascular surgery, and anesthesiology, definitive chemoradiation was considered too high-risk because of limited curative potential and possible damage to cardiac devices, particularly the implantable cardioverter-defibrillator, which is more susceptible to radiation exposure than the LVAD, for which radiation tolerance has been reported. Given high functional status (New York Heart Association class II), need for local control of bleeding, and the patient’s general reluctance to receive chemotherapy, the patient elected to proceed with upfront surgical resection and underwent robot-assisted minimally invasive Ivor Lewis esophagectomy with curative intent 6 weeks after presentation.

At operation, invasive arterial, central venous, and pulmonary artery catheterizations were performed. A milrinone infusion was started in anticipation of an increase in pulmonary vascular resistance accompanying abdominal insufflation and OLV with capnothorax. A robotic-assisted approach was selected for the abdominal phase, allowing port placement inferior to the epigastric LVAD driveline in the usual locations 18 cm inferior to the xiphoid. The driveline caused an insignificant bulge into the peritoneum (Figure 2). Gastric conduit creation and lymphadenectomy were completed without issue. Indocyanine green fluorescence angiography confirmed adequate conduit perfusion despite the absence of arterial pulsatility (Video 1).

Figure 2.

Intraoperative view of the abdomen demonstrating insignificant extraperitoneal bulge by the left ventricular assist devices driveline.

The thoracic phase was also performed robotically. After the initiation of OLV and capnothorax, pulmonary artery pressures and partial pressure of carbon dioxide, arterial, increased with concomitant decreases in minute ventilation, cardiac output, and LVAD flows. These changes were managed with escalation of vasoconstrictors to support RV perfusion, ephedrine boluses to enhance RV inotropy, and increased minute ventilation. A tension-free intrathoracic esophagogastric anastomosis was created using a 25-mm circular stapler. Estimated blood loss was 200 mL. The patient was extubated in the operating room and brought to the intensive care unit (Figure 3). A cardiovascular surgeon and LVAD perfusionist were present throughout the operation.²

Figure 3.

Comparison of preoperative (right) and postoperative (left) radiographs of the chest.

Postoperative care involved close coordination between thoracic surgery and LVAD teams. Oral-contrast computed tomography on postoperative day (POD) 5 demonstrated no evidence of an anastomotic leak, allowing removal of the nasogastric tube and initiation of a fluid diet. Upper endoscopy on POD 8 demonstrated a healed, patent anastomosis, and the pylorus was dilated to 18 mm. Thereafter, anticoagulation with warfarin was resumed. The patient was discharged home on POD 15 and continues to do well 2 months after surgery. Final pathology revealed a moderately differentiated adenocarcinoma, pT3N1, with 2 of 24 lymph nodes positive and negative margins. The patient elected to defer adjuvant chemotherapy, and there is limited evidence supporting its use after upfront surgery.

Discussion

The evidence guiding management of LVAD-supported patients with malignancy undergoing thoracic surgery is limited to case reports, necessitating individualized strategies that balance oncologic benefit against perioperative risk. To date, a single report has demonstrated the feasibility of definitive chemoradiation in an LVAD-supported patient with metastatic esophageal adenocarcinoma, despite the well-documented risks of cardiotoxicity, fluid shifts, infection, and potential damage to cardiac devices.³ In contrast, our patient had potentially curable locoregional disease with active gastrointestinal bleeding in the setting of mandatory long-term anticoagulation, making esophagectomy an attractive definitive intervention.

Beyond the usual benefits of minimally invasive surgery, robotic assistance in the abdominal phase allowed for port placement below the epigastrium to avoid the LVAD driveline and enabled effective exposure with lower insufflation pressures to reduce effect on preload. In the chest phase, tolerance of OLV represents a critical physiologic prerequisite for a minimally-invasive approach. In LVAD-supported patients, OLV-related increases in pulmonary vascular resistance from hypoxic pulmonary vasoconstriction, capnothorax, and lateral decubitus positioning with mediastinal manipulation can compromise right ventricular function and LVAD inflow. Continuous pulse oximetry was unreliable because of nonpulsatile flow; therefore, cerebral oximetry and frequent arterial blood gas analysis were used to assess oxygenation. Right heart function was evaluated with invasive pressure monitoring, serial cardiac output assessments, and frequent review of LVAD settings. Transesophageal echocardiography is typically valuable for intraoperative monitoring of patients with LVADs but was not possible, given the nature of the surgery. Importantly, reductions in LVAD flows represent a late clinical sign of deterioration and should not be relied upon as a primary monitoring tool.⁴ Although we were prepared to convert to open thoracotomy with intermittent 2-lung ventilation if required, this would incur increased postoperative pain, bleeding, respiratory dysfunction, and overall morbidity.

Fluid management posed additional challenges, requiring careful balance between maintaining preload and minimizing anastomotic risk. Steep reverse Trendelenburg positioning during the abdominal phase leads to venous pooling in the lower extremities, reducing right ventricular preload, an effect further exacerbated by laparoscopic insufflation. Abrupt loss of preload may precipitate acute right heart dysfunction and risk a catastrophic suction event. Accordingly, although conventional guidance favors fluid resuscitation to maintain LVAD flow and systemic perfusion, a liberal fluid strategy increases the risk of conduit edema and anastomotic leak.⁵ In this case, a careful, goal-directed fluid strategy and close coordination with anesthesia ensured that positioning, insufflation, and mechanical disruptions of preload were minimized and applied gradually.

Continuous-flow LVAD physiology introduces additional theoretical considerations for anastomotic healing following esophagectomy. The absence of pulsatility in LVAD patients may further alter microvascular perfusion, shear stress, and tissue oxygen delivery, factors critical to wound and anastomotic healing. Meticulous conduit assessment using indocyanine green fluorescence angiography and tension-free anastomotic construction resulted in uncomplicated healing.

Conclusions

This case demonstrates the safety and feasibility of providing anesthesia for a minimally invasive esophagectomy in a patient with a durable LVAD. Although limited to a single case, our experience suggests that minimally invasive foregut resection with reliable anastomotic healing is feasible in LVAD-supported patients.

Conflict of Interest Statement

The authors reported no conflicts of interest.

The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.

Supplementary Data

Video 1

Indocyanine green fluorescence angiography demonstrating perfusion of the gastric conduit and absence of arterial pulsatility in a patient supported by a continuous-flow LVAD. Video available at: https://www.jtcvs.org/article/S2666-2507(26)00086-6/fulltext.