Successful perioperative management of ex vivo hepatic resection and auto-transplantation

Dear Editor, Ex vivo resections have been developed to overcome anatomical and technical difficulties with certain liver tumour resections, which include complete hepatectomy followed by auto-transplantation after ex vivo hepatic resection. Good anaesthetic management of these patients with adequate pre-emptive strategies for haemodynamic instability, massive bleeding, and ischaemia-reperfusion injury (IRI) is essential for good outcomes. With less than 50 such cases being reported worldwide with significant poor outcomes,[1-3] we report our experience of a similar patient who underwent successful perioperative management. A 48-year-old male patient presented with vague upper abdominal discomfort and was found to have a 7 × 6 cm intrahepatic cholangiocarcinoma involving the right lobe and encasing the inferior vena cava (IVC) 180 degrees. A decision to do ex vivo resection was made due to the lesion being close to all three hepatic veins with complete blockage of right and middle hepatic veins and IVC involvement precluding a safe parenchymal transection. He underwent right portal vein embolisation and neoadjuvant chemotherapy. He did not have any additional comorbidities and was American Society of Anesthesiologists physical status class 1. Pre-anaesthetic assessment, including dobutamine stress echocardiography, was unremarkable. The veno venous bypass (VVB) team was kept on standby. Non-invasive cardiac output monitoring (NICOM) (Starling system, Baxter, United States) was used for assessing fluid management intra-operatively. An increase in stroke volume by >10% as a response to fluid bolus was considered as a fluid responsive state. N-acetylcysteine (NAC) infusion at 200 mg/kg/24 hours was started as prophylaxis for IRI. A four-lumen central venous catheter and an 11Fr sheath were placed in the left internal jugular vein (IJV), and the left radial artery was cannulated. The right IJV and right femoral artery were strategically left available for VVB. Initiation of VVB was planned when a target mean arterial pressure of 65 mm Hg could not be achieved with adequate intra-vascular volume and noradrenaline dose up to 0.3 μg/kg/minute. Hourly arterial blood gas and second hourly rotational thromboelastometry (ROTEM) assessments were performed to guide transfusion of blood and blood products. Severe haemodynamic instability at trial caval clamping was managed with additional fluid resuscitation. Noradrenaline was the vasopressor of choice. Glycaemic control, acidosis, and calcium levels were meticulously managed. Complete caval clamping lasted for 20 minutes. The cava was reconstructed with a Dacron patch, and a temporary portocaval shunt was performed for haemodynamic stability and to avoid splanchnic congestion. The ex vivo resection was simultaneously performed on the bench with University of Wisconsin cold perfusate solution. After a cold ischaemia time of 40 minutes, the reduced liver (left lateral segment), weighing 464 grams, was implanted within warm ischaemia time of 30 minutes and reperfused. The patient’s graft-recipient body weight ratio was 0.66. Transient hypotension was encountered. The total anhepatic phase lasted for 70 minutes. Intravenous methylprednisolone 350 mg (5 mg/kg) was given to prevent IRI. The duration of surgery was 600 minutes, with blood loss of around 1800 ml. Haemoglobin dropped to 10.4 g/dl from 14.2 g/dl. Transfusion triggers were a haemoglobin level below 8 g/dl and abnormalities in ROTEM. Coagulation status was normal, and no transfusions were required. His urine output remained more than 0.75 ml/kg/hour. The peak lactate level was 9.1 mmol/L, which dropped gradually after re-implantation [Figure 1]. The patient was shifted to the intensive care unit and extubated after 7 hours with normal sensorium. Vasopressors were tapered and stopped within 6 hours, and lactates reached levels <2 mmol/L in the next 4 hours.Figure 1: Trend of intraoperative haemodynamic parameters. Intra-operative Timepoints. T0: Baseline; T1: At caval clamp; T2: At explant; T3: 5 minutes before reperfusion; T4: At reperfusion; TS: 5 minutes after reperfusion; T6: End of surgery; HR = Heart rate in beats per minute; MAP = Mean arterial pressure in mmHg; SV = Stroke volume in mlPatient-controlled analgesia (PCA) with morphine and paracetamol was used for analgesia. Although diminished hepatic clearance of morphine has been well described, a protocolised use of morphine PCA has been found to offer better analgesia with a safe adverse effect profile in liver resections when compared to fentanyl.[4] NAC was continued for 24 hours. Enoxaparin was started on post-operative day (POD) 1, and aspirin was started on POD 6. The rest of the post-operative course remained stable. The patient was discharged home on POD 8. Laboratory values are shown in Table 1.Table 1: Perioperative laboratory parametersMaintenance of haemodynamic stability is the most challenging, especially during caval clamping/reconstruction, anhepatic phase, and reperfusion. VVB is preferred in many centres to avoid such intense haemodynamic instability but is associated with multiple complications like bleeding, haemolysis, and thrombocytopenia.[5] Therefore, we did not start VVB pre-emptively. The intraoperative haemodynamic parameters were measured at various time points in our case and their trend is shown in Figure 1. The intra-operative haemodynamic monitoring strategy, nevertheless, should be chosen after discussion with the surgical team. Strategies for the prevention of IRI have not been well described. We used methylprednisolone and NAC extrapolating our experience with orthotopic liver transplantation and a few positive outcomes reported in animal studies and humans undergoing laparoscopic liver resection.[6] NAC infusions have demonstrated attenuated transaminitis in the post-operative period. In a series of 43 cases, bleeding, biliary leakage, liver failure, renal failure, and pneumonia were commonly observed post-operative complications.[1] Liver failure, renal failure, and sepsis contributed to nine deaths. In another series, one patient succumbed to acute respiratory distress syndrome which was presumed to be due to IRI.[3] In two cases previously reported by our team, one developed post-operative respiratory failure, while the other demonstrated an uneventful course.[2] Nevertheless, ex vivo resection is an invaluable technique for patients with no conventional surgical options. A multi-disciplinary approach with proper planning is crucial in the anticipation of intra-operative events and to ensure smooth post-operative recovery. Author contributions SP and AR: Conceptualisation and writing the original draft. RR and MR: Review and editing. Disclosure of use of artificial intelligence (AI)-assistive or generative tools No AI assistance was used in preparing this project. All content was generated independently by the author. Declaration of use of permitted tools No permitted tools were used. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.