An Iterative Design Approach to Development of an Ex Situ Normothermic Multivisceral Perfusion Platform

Abstract

Background/Objectives: Challenges in normothermic machine perfusion (NMP) remain, particularly concerning the duration for which individual organs can be safely preserved. We hypothesize that optimal preservation can be achieved by perfusing organs together in a multivisceral block. Therefore, our aim was to establish a platform for ex situ multivisceral organ perfusion. Methods: Multivisceral grafts containing the liver, kidneys, pancreas, spleen, and intestine were obtained from Yorkshire pigs. Three generation (gen) set-ups were tested during the iterative design process, and minor changes were made throughout. Gen 1 (n = 4) used a custom-designed single perfusion circuit. Gen 2 (n = 3) employed a dual perfusion circuit. Gen 3 (n = 4) featured a single perfusion circuit with an optimized basin and reservoir. Grafts underwent NMP using an autologous blood-based perfusate, while hemostatic parameters and function were assessed. Results: Comparing Gen 1 versus Gen 3, the mean aortic flow improved (1.018 vs. 2.089 L), resistance decreased (0.224 vs. 0.038), urine output increased (51.90 vs. 271.3 mL), oxygen consumption rose (43.56 vs. 49.52 mL O₂/min), perfusate lactate levels dropped (10.44 vs. 3.10 mmol/L), and the pH became more physiological (7.27 vs. 7.30). Cellular injury trended lower in Gen 3. Histological evaluation demonstrated minimal differences in Gens 2 and 3. Conclusions: We demonstrate the feasibility of abdominal multivisceral NMP for up to 8 h. Adequate arterial flow, stable perfusate pH, and high oxygen consumption in setup 3 indicated organ viability. Multivisceral perfusion may serve as a plat-form for long-term NMP.

Keywords: machine perfusion; multivisceral transplantation; organ preservation.

Figure 1

Global overview of design process and each generation of the setups used.

Figure 2

Perfusion parameters of multivisceral grafts in Gen 1. (A) Aortic flow, (B) aortic resistance, and (C) graft temperature during hypothermic machine perfusion (HMP). (D) Aortic flow, (E) aortic resistance, (F) accumulative urine production, (G) oxygen consumption, (H) lactate clearance, and (I) perfusate pH during normothermic machine perfusion (NMP). Data are shown as individual values (dotted lines) and mean (continuous line).

Figure 3

Perfusion parameters of multivisceral grafts in Gen 2. (A) • Aortic flow and • portal flow, (B) • aortic and • portal resistance, (C) accumulative urine production, (D) oxygen consumption, (E) lactate clearance, and (F) perfusate pH during normothermic machine perfusion (NMP). Data shown as individual values (dotted lines) and mean (continuous line).

Figure 4

Perfusion parameters of multivisceral grafts using Gen 3. (A) Aortic flow, (B) aortic resistance, (C) accumulative urine production, (D) oxygen consumption, (E) lactate clearance, and (F) perfusate pH during normothermic machine perfusion (NMP). Data are shown as individual values.

Figure 5

Cellular injury during multivisceral perfusion. (A) alanine aminotransferase (ALT), (B) aspartate aminotransferase (AST), (C) alkaline phosphatase (ALP), (D) lactate dehydrogenase (LDH), (E) creatinine, (F) blood urea nitrogen (BUN), (G) potassium, (H) lipase, and (I) amylase were measured at baseline (−10), 4 h into perfusion (240), and at the end of perfusion (end). Data are expressed as aligned scatter plots and the arithmetic mean + SD. * p < 0.05, ** p < 0.01, **** p < 0.0001.

Figure 6

Histological assessment of the liver during multi-visceral perfusion. (A) Liver parenchyma hematoxylin and eosin-stained sections representing pre- and post-NMP (Gen 1 = pig 3, Gen 2 = pig 8, and Gen 3 = pig 12). (B) biliary mucosal loss, (C) mural stromal necrosis, (D) cholangitis, (E) hepatic arteriosclerosis, (F) hepatic steatosis, and (G) hepatic necrosis were scored. Data are expressed as aligned scatter plots and the arithmetic mean + SD. * p < 0.05, ** p < 0.01. Original magnification: 20×, scale bar is 200 µM.

Figure 7

Histological assessment of the kidney during multi-visceral perfusion using the EGTI scoring system. (A) Renal cortex hematoxylin and eosin-stained sections representing pre- and post-NMP (Gen 1 = pig 3, Gen 2 = pig 8, and Gen 3 = pig 12). (B) Glomerular injury score, (C) renal endothelial injury, and (D) tubular interstitial were scored. Data are expressed as aligned scatter plots and the arithmetic mean + SD. Original magnification: 20×, scale bar is 200 µM. * p < 0.05.

Figure 8

Histological assessment of the pancreas and intestine during multi-visceral perfusion. (A) Pancreatic necrosis score, (B) islet cell integrity, and (C) intestinal musical injury were scored pre- and post-NMP. Data are expressed as aligned scatter plots and the arithmetic mean + SD. *** p < 0.001.