Ischemic Preconditioning in Major Hepatectomy

This is a prospective controlled trial, conducted between January 2005 and December 2007. Three centres specialized in hepatobiliary surgery (Cochin hospital (Paris, France), BEAUJON hospital (Clichy, France) and PONTCHAILLOU Hospital (Rennes, France)) participated to this study, which has been initiated by the department of hepatobiliary surgery and liver transplantation of the Cochin hospital. The protocol was approved by ethics committees of each participating centres and an informed consent was obtained from each patients before they were enrolled. Eighty seven patients were randomized to either receive ischemic preconditioning prior to liver resection under intermittent pedicular clamping or not. Ischemic preconditioning was performed through a sequence of 10 minutes vascular inflow occlusion and 10 minutes of reperfusion prior to intermittent pedicular clamping. Intermittent pedicular clamping was conducted through a sequence of 15 minutes of vascular inflow occlusion and 5 minutes of reperfusion. The randomization process which was centralized was held in the operating room after inclusion criteria had been check and exclusion criteria ruled out. Inclusion criteria were: patients' age (≥ 18 years old), liver resection of 3 segments (as described by Couinaud) or more, posterior lesionectomy (segment VI and VII), liver resection only or associated with a primary digestive or biliary tumor. Exclusion criteria : Patients with cirrhosis, synchronous radiofrequency or cryotherapy ablation, undergoing segmentectomy, left lateral lesionectomy or laparoscopic liver resection were excluded from further analysis as well as pregnant women.

Anaesthetic protocol:

Patients' status was evaluated by the ASA (American Society of Anaesthesiology) scoring system. Surgical procedure was conducted under low central venous pressure (5cm H2O) to avoid excessive bleeding from suprahepatic veins backflow.13 Patients' anaesthesia was performed using a single protocol that was common to all participating centres. Patients in whom this protocol was contraindicated due to medical reasons were excluded. A standardized general anesthesia using thiopental, sufentanil, atracurium and sevoflurane was applied throughout the study period. Mechanical ventilation was carried out using 50 % oxygen in nitrous oxide and was adjusted to keep end-tidal PCO2 between 4.7 and 6.0 kPA. Intra-venous cefazolin (Cephazolin, PANPHARMA laboratory, FOUGERES, France) was administered for antibioprophylaxy. An arterial line was inserted for arterial pressure monitoring and blood sampling. Central venous pressure was not consistently maintained below a predetermined level. Intraoperative Ringer lactate infusion was limited to the minimum, practically below 500 mL, until parenchymal resection was completed. Following liver resection, hydroxyethyl starch 130 (Voluven, Fresenius laboratory, SEVRES, France), 20 mL/kg, was infused in one hour. Thereafter, additional hydroxyethyl starch, fresh frozen plasma, red blood cell packs were administered as indicated by urinary output, hemodynamics, and biological data. Thresholds for blood transfusion were a hemoglobin level of 7g/dL for healthy patients 64 years of age or under, and 9 g/dL for patients 65 years of age or over and/or with preexisting cardiopulmonary disease (1).

Surgical protocol:

All of the patients included in this study were operated in high volume centres by senior surgeons specialized in hepatobiliary surgery. Major resection was defined as resection of 3 or more liver segment as described by Couinaud. Ischemic preconditioning was performed through a sequence of 10 minutes vascular inflow occlusion and 10 minutes of reperfusion prior to intermittent pedicular clamping. Intermittent pedicle clamping was conducted through a sequence of 15 minutes of vascular inflow occlusion and 5 minutes of reperfusion. Intermittent clamping was used during the whole parenchyma transection process. During ischemic preconditioning the liver left in native position to avoid ischemic process due to compression. Surgical liver biopsies were performed before after interruption of the vascular inflow. The samples were all collected in the same centre (Cochin Hospital) for histological and molecular biological analysis. The device used for clamping was left to the surgeon's discretion (Tourniquet or vascular clamp) but had to be the same for both preconditioning and clamping. The technique used for parenchymal liver transection was left to the surgeon's discretion as well as haemostasis and BILIOSTASIS techniques which were performed using bipolar forceps, metallic clips or ligation depending on vessels or bile ducts size. Devices used for parenchymal transection and haemostasis/BILIOSTASIS techniques were recorded in the preoperative data collection form. Postoperative drainage of the abdominal cavity was left to the surgeon's discretion but was collected in the data sheet. All complications occurring during surgery were collected as well as blood loss, blood transfusion and fluid infusion.

Patient's follow-up and data collection:

Patients' follow-up was 3 months and was initiated the day before surgery (Di). Given that primary hypothesis was a 50% reduction of transaminases (AST, ALT) level on postoperative day 1 in the preconditioning group, blood samples were collected on all patients on Di and POD1 for transaminases measurement. All of these samples were collected in a unique biochemistry laboratory (Hospital Cochin) to avoid inter laboratory variations. Besides these two samples, each centres conducted a regular biological follow-up of patients on a previously established schedule which was common to all centres. Biological assessment was performed on blood samples collected on Di and POD 1, 3, 5, 7, 14 and 28. Blood samples were tested for liver biochemistry (AST, ALT, PAL, GGT, BT and BC), prothrombin time, factor V, blood cells count (red blood cells, white blood cells and platelets) and albumine. Clinical examination was performed every day by the senior surgeon in charge of the patients until they were discharge from hospital, as well as on POD 15, POD 30 (5 days) and POD 90 (10 days). All impaired outcomes were noted in the data sheets. Surgical complications collected were: Intra abdominal bleeding, biliary fistula, vascular complication, intra abdominal abscess, wound infection and reintervention. Medical complication collected were: Infection (urinary tract, lung, catheter) liver dysfunction, liver failure, ascitis, pleural fluid effusion necessitating a drainage or not. Postoperative morbidity and mortality were defined as any impaired surgical or medical outcome or death occurring within POD 90, respectively. When hospital readmission was required patients where hospitalised in the surgical department. Finally histological analysis of the liver samples collected at surgery was reviewed by a single pathologist who was unaware of any clinical data.

Statistical analysis:

The analysis was performed in intention-to-treat. The sample size calculation was based on the primary endpoint postoperative (POD1) aspartate aminotransferase (ALT). According to previous data (ref 4-7,9,10 protocol), the sample size calculation was performed with the expectation of a 50% difference in postoperative serum ALT level with a level of statistical significance of 0.05 and a power of 0.80, using a two-tailed t-test. This calculation indicated to include 38 patients in each group. Modification of ALT was also quantified as the difference between Di and POD 1. Demographic data, baseline characteristics, and surgical data are summarized by groups using descriptive statistic. Categorical variables are expressed as numbers (percentages) and comparisons between groups were performed by the chi-square test or Fisher exact test, when needed. Continuous variables are expressed as mean (standard deviation of the mean (sd)) and were compared using two-tailed t-test. Analyses were performed using SAS software, version 9.1, SAS institute inc, Cary, North Carolina.