*2.5. Surgical Procedure and In-Situ Perfusion*

All procedures were carried out by an expert surgeon trained in microsurgery and ex-situ perfusion models. Time periods required for bile duct cannulation (from laparotomy to cannula placement in the bile duct), portal cannulation (from bile duct cannulation to portal cannula placement), in-situ cold-flush, liver procurement (from flushing to procurement), warm ischemia time (from the end of in-situ flush to the start of the cold storage), and liver connection to NMP (from end of SCS to connection to the circuit) were examined as indexes of technical success and reproducibility of the surgical procedure. The rats that were randomized to native group (*n* = 5) were not included in this analysis because they did not undergo any surgical procedure. There were no statistically significant differences among the "surgical" groups in terms of surgical preparation 37 ± 1 min for DMEM, 35 ± 2 min for BLOOD, and 35 ± 2 min for SCS. Cold flush lasted 60 ± 4 sec for DMEM, 71 ± 18 sec for BLOOD, and 61 ± 3 sec for SCS, while warm ischemia time 16 ± 1 min for DMEM, 16 ± 3 min for BLOOD, and 16 ± 2 min for SCS.

#### *2.6. Ex-Situ Liver Perfusion*

#### 2.6.1. Perfusion Fluids Preparation

Perfusate composition is shown in Supplementary Materials (Table S5). DMEM perfusate was obtained by adding to DMEM, 4% human albumin, streptomycin and penicillin, glutamine, and insulin for a total volume of perfusate of 100 mL. Perfusion fluid was obtained according to Op den Dries and colleagues [13]. We added to DMEM 4% human albumin, streptomycin and penicillin, glutamine, insulin, heparin, and 15% hematocrit human RBC for a global volume of perfusate of 100 mL.

Human red cells were provided by the Blood Bank of our Institution (Department of Transfusion Medicine and Hematology). A packed red blood cell concentrate was prepared by centrifugation from one unit of a whole blood donation, which was unsuitable for human use because of an insufficient volume collected. The same unit (blood group: AB Rh+) was used throughout the experiment.

#### 2.6.2. Normothermic Machine Perfusion Setup

The perfusion system (Figure 2) consisted of a customized circuit derived from the isolated lung perfusion system (Hugo Sachs Elektronik, Harvard Apparatus, March-Hugstetten, Germany) implemented and described by our laboratory [21–24]. It consisted of a reusable heated glass reservoir, a heated bubble trap, a circulating tube derived by an infusion system, an octagonal peristaltic pump, a polystyrene lid, and a single-use artificial lung (Supplementary Tables S3 and S4). The liver was placed onto the glass chamber modified to let the liver laid on the diaphragmatic surface on a modeled ad hoc, perforated parafilm. The liver was connected to the circuit through the portal vein cannula and the bile duct cannula to a 2 mL tube to collect bile. The chamber was closed to maintain humidity. Temperature inside the chamber was recorded with a 1.3 mm probe. The heat-exchanger was set at 40 ◦C to obtain a graft temperature of 37 ◦C. We carefully avoided air embolism during priming and liver connection. The artificial lung was ventilated with a 200 mL/min flow, with 95% FiO2 and 5% FiCO2.

**Figure 2.** (**A**) Schematic representation of normothermic machine perfusion circuit. 1: pre-liver sampling stopcock; 2: post-liver sampling from IVC; 3: bile collection cuvette; A: heated glass reservoir; P: peristaltic pump; ML: membrane lung; H: heated exchanger and bubble trap. (**B**) Detail of the portal vein and bile duct cannula during ex-situ perfusion. Liver was laid on the diaphragmatic surface. Arrow: inferior vena cava; § right lateral lobe; \* median lobe; ◦ left lateral lobe; # caudate lobe.

Circuit hemodynamic parameters were monitored using the Colligo system (Elekton, Milan, Italy). Portal vein and pre-lung pressures were recorded every 5 min during rewarming, and every 30 min during normothermic machine perfusion. Vascular resistances were calculated as mean portal vein pressure divided by blood flow in the portal vein (mL/min). Temperature was monitored by a thermo probe located between the lobes of the liver.
