In LT, the clinical implications of the two types of hepatic steatosis are very different. Although the presence of ≥30% microvesicular steatosis has been associated with an increased risk of postreperfusion syndrome, early allograft dysfunction
[10], rejection, and the need for postoperative renal replacement therapy
[11][12], use of liver grafts with even significant microvesicular steatosis is generally considered to be safe
[9][13]. In contrast, the presence of moderate (30–60%) or severe (≥60%) MaS appears to be more clinically impactful, proportionally to its severity. The utilization of livers with moderate MaS has been associated with an increased rate of early allograft dysfunction, biliary complications, and decreased graft survival, whereas severe MaS has been linked to postoperative poor function, need for renal replacement therapy, and inferior patient and graft survival
[5][6]. Consequently, MaS has been included as a negative prognostic factor in models predicting post-LT patient and graft survival
[14][15]. The high risk associated with the use of severely steatotic livers is reflected by the very low number of patients in the series reporting their use
[7][16][17][18][19], suggesting that, despite some encouraging results that have been reported, these grafts are generally approached with extreme caution and most frequently discarded. Recently, a study based on the Scientific Registry of Transplant Recipients has shown that moderate (≥31%) liver allograft MaS is associated with 87% to 95% lower odds of graft utilization, while utilization of fatty livers increases the risk of graft failure by 53%
[20].
2. Why Are Steatotic Livers More Susceptible to Ischemia-Reperfusion Injury?
Hepatic ischemia-reperfusion injury (IRI) is a sterile inflammatory response commonly encountered during major liver surgery, such as liver resection and liver transplantation, when organ blood supply is restored after a period of ischemia. The pathophysiological bases of IRI have been recently reappraised, identifying mitochondria as the primary targets and initiators of IRI cascade
[28][29]. Under ischemic conditions, cell metabolism is switched to anaerobic glycolysis while, at the mitochondrial level, the lack of oxygen interrupts the electron transport chain and causes the accumulation of reduced electron carrier molecules (succinate), initiates reverse electron transfer, and leads to the detachment of flavin mononucleotide from mitochondrial complex I
[30]. ATP depletion and lactate accumulation result in electrolyte imbalance and cellular acidosis. When oxygen levels are abruptly restored upon reperfusion, the negative potential across mitochondrial matrix generated during ischemia results in the production of high amounts of reactive oxygen species (ROS)
[31]. Hepatocellular ROS initiate the sterile immune response by promoting the release of high mobility group box 1 (HMGB1) and nuclear factor κβ (NF-κβ). HMGB1 and NF-κβ are both central mediators of the reperfusion phase, as their signaling sustains Kupffer cell activation, microcirculation impairment, neutrophil recruitment, and eventually the activation of cell death processes
[32][33].
It is well known that steatotic livers are extremely vulnerable to IRI but the underlying mechanisms are not completely understood. Evidence from experimental models suggests that the inflammatory response to IRI is different in steatotic and non-steatotic livers
[34][35] and that increased mitochondrial oxidative stress and impaired ATP restoration are major determinants of the increased susceptibility of steatotic livers to IRI
[36].
Mitochondrial uncoupling protein-2 (UCP-2) is a mitochondrial protein that regulates proton leakage across the inner membrane. In steatotic livers, UCP-2 expression is increased to reduce oxidative pressure and ROS production, in an attempt to protect the liver from chronic fat accumulation. However, by diminishing ATP synthesis and reducing ATP baseline levels, UCP-2 overexpression compromises hepatocytes capacity to respond to an acute energy demand, similar to how it occurs during IRI, leading to mitochondrial permeability transition (MPT) and membrane potential collapse
[37][38].
In lean livers, cell death due to IRI can occur through different pathways, with apoptosis being the most represented
[32][34][39]. However, since apoptosis is an energy-dependent process, the chronic ATP depletion observed in fatty livers may lead to the failure to induce apoptosis in favor of necrosis or other forms of programmed cell death
[32][34][40]. Indeed, higher levels of RIPK1 and RIPK3, caspase-9, caspase-1, and iron overload have been observed in fatty degenerated hepatocytes exposed to IRI, suggesting an important role of MPT-driven necrosis, necroptosis, pyroptosis, and ferroptosis, respectively
[36]. The considerable overlap and crosstalk between these pathways may have contributed to the confounding and sometimes controversial results reported by the existing studies
[40][41][42].
Fat droplet accumulation in hepatocytes can cause partial or complete obstruction of sinusoids, resulting in a reduction in sinusoidal blood flow
[43][44]. This might be exacerbated, upon graft reperfusion, by the rupture of hepatocyte membrane and the release of fat droplets in the extracellular space, similarly to what happens in lipopeliosis
[45][46]. As a result of chronic hypoxic state, steatotic livers are characterized by an increased expression of endothelin (ET-1) and inducible nitric oxide synthase (iNOS). ET-1 and iNOS imbalance aggravates sinusoidal vasoconstriction, worsening microcirculatory damage upon reperfusion
[47].
The endoplasmic reticulum (ER) serves many roles in the cell including calcium storage, protein synthesis, and lipid metabolism, which are stressed in fatty hepatocytes. Moreover, chaperonin downregulation
[48] contributes to ER stress supporting the unfolded protein response (UPR), a signal transduction cascade that ultimately leads to NF-ĸB, JUN N-terminal kinase, and caspase-12 activation
[36].
The aforementioned mechanisms, although still partially undisclosed, represent the basis to develop strategies to reduce IRI in fatty livers.
3. Impact of Different Preservation Solutions
Since the introduction of Collins solution in 1969
[49], organ preservation by static cold storage (SCS) has been one of the key elements allowing the expansion of organ transplantation worldwide
[50]. The principle of preservation by SCS is slowing down cellular metabolism with hypothermia while preservation solutions prevent or minimize cellular swelling, interstitial edema, intracellular acidosis, and ROS production, and provide energy substrates
[50]. Effective preservation is of utmost importance when dealing with steatotic livers.
First developed in the late 1980s by Belzer and Southard
[51], the University of Wisconsin solution (UW) is still considered the gold standard against which other solutions must be compared. UW is a colloid solution with high potassium, low sodium concentration (intracellular solution), and high viscosity due to the presence of hydroxyethyl starch (HES) as an oncotic agent. Histidine–tryptophan–ketoglutarate solution (HTK), which was originally introduced in the 1970s for cardioplegia
[52], employs mannitol as an impermeant and does not contain colloids, resulting in decreased viscosity as compared to UW. This solution contains histidine and α-ketoglutarate as energy substrates and buffers, and tryptophan as a membrane stabilizer and antioxidant. Similarly to HTK, Celsior solution (CS) does not contain colloids but it is characterized by high-sodium and low-potassium concentrations (extracellular solution) and was specifically designed to limit calcium overload and ROS production
[53]. Institut Georges Lopez-1 solution (IGL-1) is characterized by high-sodium and low-potassium levels (extracellular solution) and by polyethylene glycol (PEG) instead of HES as an oncotic agent, resulting in lower viscosity as compared to UW
[54]. UW, HTK, Celsior, and IGL-1 are nowadays the most widely utilized preservation solutions in liver transplantation
[50]. In the general population, clinical results with either preservation solution have been shown to be roughly equivalent
[55], although large studies based on the European Liver Transplant Registry have shown inferior results with the use of HTK
[56][57][58].
In steatotic livers, most studies comparing the efficacy of preservation by different preservation solutions have been conducted in an experimental setting (Table 1).
PEG is a nontoxic, highly soluble neutral polymer capable of preventing edema and cellular membrane destabilization if administered intravenously in a model of warm ischemia-reperfusion injury
[59]. The benefits of PEG-containing solutions during cold preservation could be associated with reduced shear stress and improved microcirculation due to reduced viscosity. Indeed, replacing hydroxyethyl starch by PEG results in a much lower viscosity of IGL-1 as compared to UW (1.28 versus 5.7 millipascal-second). Cellular protection is also associated with the reduction of mitochondrial damage by the increased activation of protective cell mechanisms such as adenosine monophosphate-activated protein kinase (AMPK) and endothelial NO synthase (eNOS)
[60], as also recently demonstrated in human hepatocytes exposed to IRI in vitro
[61].
In 2006, Ben Mosbah et al. first reported the superiority of IGL-1 in preserving steatotic rat livers
[62]. Compared to UW, livers preserved with IGL-1 showed less transaminases release, increased bile production, lower malondialdehyde (MDA, a marker of lipid peroxidation and oxidative injury) levels, lower glutamate dehydrogenase (GLDH, a marker of mitochondrial injury) activity, and reduced vascular resistance. The authors postulated that nitric oxide (NO) was involved in the IGL-1 protection against IRI, as suggested by the overexpression of eNOS in the IGL-1 group and by the suppression of IGL-1 protective effects when a NO-inhibitor was added to the preservation solution. The same group then investigated the mechanistic aspects of IGL-1’s apparent superior preservation of steatotic livers in a series of subsequent experiments. IGL-1 enriched with either insulin-like growth factor-1 or epidermal growth factor further increased eNOS activation and improved protection against IRI
[63][64]. High levels of hypoxia-inducible factor 1-alpha (HIF-1α) were found in livers preserved with IGL-1, and the overexpression of heme-oxygenase 1 (HO-1), one of the HIF-1α downstream genes, supported the cytoprotective role of this signaling pathway
[65]. Trimetazidine, an anti-ischemic drug, enhanced HIF-1α and sirtuin 1 induction and reduced HMGB1 levels, thus promoting autophagy to mitigate IRI
[66].
In a study comparing IGL-1 and Celsior, Tabka et al. obtained similar results
[67]. Rat livers preserved by IGL-1 showed increased eNOS levels and reduced activation of the pro-apoptotic mitogen-activated protein kinase (MAPK) pathway. Arterial relaxation was found to be highly dependent on NO levels during preservation with IGL-1, corroborating the hypothesis that IGL-1 solution may prevent endothelial dysfunction through eNOS activation. Supplementing IGL-1 with bortezomib, a proteasome inhibitor, resulted in AMPK activation and the downstream expression of eNOS and GSK3β, leading to reduced hepatocellular injury, oxidative stress, and apoptosis
[68]. Similarly, carbonic anhydrase II, an enzyme involved in many IRI-related processes, enhanced IGL-1 capacity to induce AMPK and consequently reduce UPR- and MAPK-related events, resulting in superior liver function and histology
[69]. Altogether, these results confirmed that IGL-1 benefits in fatty liver preservation are related to AMPK and eNOS activation
[70][71].
Subsequent studies
[54][72][73], besides confirming the advantages of IGL-1 over UW and HTK, showed that the IGL-1 benefits in the preservation of steatotic livers were linked to proteasome inhibition
[73], aldehyde dehydrogenase 2 (ALDH2) upregulation
[74], and autophagy induction
[75].
Based on IGL-1 studies, similar improvements were obtained through activation of the AMPK pathway and eNOS induction during SCS using UW. Supplementing UW with trimetazidine, aminoimidazole-4-carboxamide ribonucleoside, carvedilol, or bortezomib during the preservation of steatotic rat livers resulted in lower perfusate transaminases, increased bile production, reduced vascular resistance, and lower MDA and GLDH activity during normothermic reperfusion
[76][77][78][79].
Eipel et al.
[80] investigated the effects of supplementing HTK with erythropoietin for the preservation of steatotic mice livers. Increased oxygen consumption, better preservation of the endothelium, and a slight reduction in AST levels were observed in the treated group after 2 h or normothermic reperfusion. However, UCP-2 expression was not influenced by erythropoietin supplementation and the signaling pathways explaining better preservation in the treatment group were not completely clarified.
To further reduce ROS production and cell damage, a new IGL solution (IGL-2) was developed with higher PEG (5 versus 1 gr/L) and glutathione concentrations, and the addition of histidine and mannitol instead of raffinose as an impermeant
[81]. Importantly, IGL-2 was designed for use during both SCS and machine perfusion, possibly avoiding the need for repeated graft flushing between different phases of organ preservation
[82]. In fatty livers, preservation by IGL-2 resulted in reduced mitochondrial injury and oxidative stress, as reflected by increased levels of HO-1, glutathione, ALDH2, and mitochondrial complex I and II, all key actors in the response to IRI
[81][82][83]. Interestingly, livers stored with IGL-2 retained the lowest amount of water during preservation, suggesting that PEG could decrease the interstitial formation.
In conclusion, experimental evidence suggests that PEG-containing solutions provide advantages in terms of mitochondrial integrity and protection against oxidative stress and that IGL-1 and IGL-2 seem the most appropriate preservation solutions for SCS of fatty livers. Theoretically, use of IGL-2 would also be associated with the logistical advantages of using the same solution for SCS and machine perfusion. However, these findings must be interpreted with caution due to the lack of experimental models involving transplantation. Furthermore, IGL-2 is still awaiting approval for clinical use and the advantages of PEG-containing solutions should be confirmed in clinical studies.
Table 1. Experimental studies evaluating the impact of different preservation solutions in the preservation of steatotic livers.
Author, Year |
Intervention |
Experimental Model |
Findings |
Ben Mosbah et al., 2006 [62] |
IGL-1 (vs. UW) |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Lower perfusate transaminase, MDA, and GLDH levels; improved bile production; lower vascular resistance. Inhibition of NO production suppressed IGL-1 effects. |
Ben Mosbah et al., 2007 [76] |
UW (+trimetazidine +aminoimidazole-4-carboxamide ribonucleoside) |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Lower perfusate transaminase, MDA, and GLDH levels; improved bile production; lower vascular resistance. Increased AMPK activation. Inhibition of AMPK suppressed the protective effects. |
Ben Mosbah et al., 2010 [77] |
UW (+carvedilol) |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Lower perfusate transaminase, MDA, and GLDH levels; improved bile production; lower vascular resistance; increased ATP. Increased AMPK activation. |
Zaouali et al., 2010 [65] |
IGL-1 (+trimetazidine) |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Lower perfusate transaminase, MDA, and GLDH levels; improved bile production; lower vascular resistance. Increased levels of HIF-1α and downstream genes. Better results and HIF-1α induction after addition of trimetazidine. Inhibition of NO production suppressed the protective effects. |
Zaouali et al., 2010 [64] |
IGL-1 (+IGF-1) |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Compared to IGL-1 alone: increased NO production, lower perfusate transaminase, MDA, and GLDH levels; improved bile production; lower vascular resistance, reduced oxidative stress. |
Zaouali et al., 2010 [63] |
IGL-1 (+EGF) |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Compared to IGL-1 alone: increased NO production, lower perfusate transaminase, MDA, and GLDH levels; improved bile production; lower vascular resistance, reduced oxidative stress; increased ATP. |
Eipel et al., 2012 [80] |
HTK (+erythropoietin) |
24 h SCS followed by 2 h normothermic reperfusion in ob/ob mice livers |
Compared to HTK alone: lower perfusate AST; improved endothelial integrity; higher oxygen consumption. |
Bejaoui et al., 2014 [68] |
IGL-1 (+bortezomib) |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Compared to IGL-1 alone: activation of AMPK signaling, lower perfusate transaminase; improved bile production; lower vascular resistance, apoptosis inhibition. Inhibition of AMPK expression reduced IGL-1 protective effects. |
Zaouali et al., 2013 [78] |
UW (+bortezomib) |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Lower perfusate transaminase, MDA, and GLDH levels; improved bile production; lower vascular resistance. Increased AMPK activation. |
Bejaoui et al., 2015 [69] |
IGL-1 (+carbonic anhydrase II) |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Compared to IGL-1 alone: activation of AMPK signaling, lower perfusate transaminase; improved bile production; increased ATP; downregulation of MAPK and UPR pathway; apoptosis inhibition. |
Bejaoui et al., 2015 [60] |
PEG preconditioning |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Lower perfusate transaminase, and GLDH levels; lower vascular resistance. Increased AMPK activation. |
Tabka et al., 2015 [67] |
IGL-1 (vs. Celsior) |
24 h SCS followed by 2 h normothermic reperfusion in Sprague-Dawley rats rat livers |
Increased NO production, lower perfusate transaminase, MDA, and GLDH levels; improved bile production; lower vascular resistance, reduced oxidative stress, downregulation of MAPK pathway. |
Zaouali et al., 2017 [66] |
IGL-1 (+trimetazidine) |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Compared to IGL-1 alone: lower perfusate transaminase and GLDH levels; increased levels of sirtuin 1 and reduced levels of HMGB1 and TNFα. |
Zaouali et al., 2017 [73] |
IGL-1 (vs. UW) |
24 h SCS followed by 2 h normothermic reperfusion in Zucker rat livers |
Lower perfusate transaminase and GLDH levels; increased ATP; reduced levels of HMGB1 and TNFα. Proteasome inhibition. |
Panisello-Roselló et al., 2017 [54] |
IGL-1 (vs. HTK) |
24 h SCS of Zucker rat livers |
Lower perfusate transaminase and GLDH levels; increased ATP; reduced levels of HMGB1 and TNFα. Proteasome inhibition. Increased AMPK activation. |
Panisello-Roselló et al., 2018 [74] |
IGL-1 (vs. HTK vs. UW) |
24 h SCS of Zucker rat livers |
Lower perfusate transaminase levels; increased ATP; reduced apoptosis. ALDH2 upregulation. |
Panisello-Roselló et al., 2018 [72] |
IGL-1 (vs. HTK) |
24 h SCS of Zucker rat livers |
Lower perfusate transaminase and GLDH levels; reduced membrane mitochondrial depolarization; reduced apoptosis; reduced levels of HMGB1; increased autophagy. |
Lopez et al., 2018 [84] |
IGL-1 (vs. HTK vs. IGL-0 *) |
24 h SCS of Zucker rat livers |
Lower perfusate transaminase levels, preserved glycocalyx integrity. |
Bardallo et al., 2021 [81] |
IGL-2 (vs. IGL-1, vs. IGL-0 *) |
24 h SCS of Zucker rat livers |
Lower perfusate transaminase and GLDH levels; increased ATP; increased autophagy; ALDH2 upregulation. |
Bardallo et al., 2022 [83] |
IGL-2 (vs. IGL-1, vs. IGL-0 *) |
24 h SCS of Zucker rat livers |
Increased ATP; reduced succinate accumulation; increased complex I and complex II levels: increased HO-1; increase glutathione levels; reduced oxidative stress. |
Asong-Fontem et al., 2022 [82] |
IGL-2 (vs. UW) |
24 h SCS +/− 2 h HOPE followed by 2 h normothermic reperfusion in Zucker rat livers |
Lower perfusate AST; preserved glycocalyx integrity; reduced levels of HMGB1; increased weight loss (surrogate of edema formation). |