For patients with end-stage liver disease, liver transplantation (LT) represents a definitive treatment. However, the shortage of liver allografts essentially restricts the widespread application of LT and contributes to a prominent waitlist mortality.1 To reduce the gap between supply and demand, allografts from donation after circulatory death (DCD) have been increasingly utilized as one critical approach to expand the donor liver pool. It has been estimated that the use of DCD allografts will maintain if not increase the number of LT procedures performed globally.2 Nonetheless, the benefits afforded by DCD allografts are established at the expense of compromised posttransplant graft performance and consequently increased patient mortality and morbidity, presenting a profound clinical challenge.3
The functional evaluation of liver grafts in an early postoperative phase emerged as a priority for transplant clinicians, as it reflects the technological success of LT and heavily influences recipient outcomes. Early allograft dysfunction (EAD) is designated as an initially marginal function state of the implanted liver, and it profoundly imperils the morbidity and mortality of recipients.4 EAD occurs in 20%–39.5% of patients following LT, with an increased frequency observed among DCD recipients.4,5 The occurrence of EAD has been demonstrated to be closely correlated with kidney function impairment, excess resource utilization, and shortened recipient and liver graft survivals.6-8
With DCD allografts increasingly being utilized for LT, the prevalence and medical burden of EAD are expected to increase. To minimize the risk of EAD and improve posttransplant outcomes, the development of a tool to simply and accurately assess graft function is of vital importance. Over the last decade, serum cytokines and chemokines, lactate clearance, coagulation factors, liver metabolomics, hepatic microcirculation, and microRNAs in preservation fluids have all been introduced as EAD indicators.9-13 Despite the markers available to predict EAD, the modalities to obtain measurements of these parameters are infrequently applied in routine clinical practice. Meanwhile, the majority of these parameters could be determined only after graft revascularization in recipients, which may indicate a result secondary to the pathogenesis of EAD instead of being causal of EAD.
In the experimental assays, several proteins have been identified to constitute key nodes of the molecular network occurring within the liver graft and to be responsible for acute liver parenchymal damage and repair.14-16 For example, vascular endothelial growth factor (VEGF) enhances microvascular permeability, induces local angiogenesis, and promotes hepatocyte regeneration; intercellular adhesion molecule-1 (ICAM-1) is well recognized as a major inflammatory molecule that mediates the rolling and adhesion of neutrophils on the sinusoidal endothelium, exacerbating sinusoidal narrowing and tissue hypoxia; NADPH oxidase (NOX) transfers electrons to molecular oxygen and produces reactive oxygen species, which if excessive would destroy cellular structures and activate inflammatory responses; and cytochrome C, when released from the mitochondria into the cytoplasm, triggers hepatocyte apoptosis. Given this background, we hypothesized that these proteins might essentially determine graft functional recovery following LT and thereby could be used in the clinical setting to predict EAD at early stages.
With this clinical study, we explored the potential relationship between the pretransplant intrahepatic protein profiles and the onset of EAD following LT from DCD. A total of 7 putative proteins were selected, and their expressions were evaluated by immunohistochemistry (IHC) analysis. Relevant clinical data were extracted and analyzed to help construct a predictive model for the estimation of EAD risk.
MATERIALS AND METHODS
This study enrolled a total of 158 patients who received primary LT from DCD at the authors’ center. In all, 121 subjects in the training cohort and 37 in the validation cohort were included. Demographic characteristics and transplant-related data of these 2 populations were collected and are summarized in Table 1.
Information regarding the patients, liver biopsy and allograft weighting, IHC analysis, definitions of EAD and acute kidney injury (AKI), survival data, and statistical analyses are included in the Supplementary Materials and Methods and Figure S1 (SDC, http://links.lww.com/TP/B748).
Written informed consent was obtained from all donors and recipients. Each organ donation or transplant was approved by the Institutional Review Board, First Affiliated Hospital, Zhejiang University, strictly under the guidelines of the Ethics Committee of the hospital, the current regulation of the Chinese Government, and the 1975 Declaration of Helsinki. This study was approved by the Ethics Committee of the First Affiliated Hospital, College of Medicine, Zhejiang University (Reference Number: 2015–383). No organs were sourced from executed prisoners.
Clinical Characteristics of EAD Patients
After transplantation, the incidence of EAD among the 121 recipients was 43.0% (52/121). The clinical characteristics of the donors and recipients, transplant and posttransplant variables, stratified by the development of EAD, are displayed in Table 2. Despite a similar baseline serum creatinine, EAD patients were more likely to develop AKI (46.2% versus 15.9%, P < 0.001) and to require renal replacement therapy (RRT; 21.2% versus 2.9%, P = 0.001) after LT than non-EAD patients. The mean RRT duration was 13.8 days for EAD patients (n = 11) and 12 days for non-EAD patients (n = 2). EAD patients also had longer intensive care unit length of stay when compared with non-EAD patients (7.9 versus 7.1 days, P = 0.04). Moreover, significant differences in the 6-month patient and graft survival probabilities were observed between patients with and without EAD (86.5% versus 98.6%, P = 0.02; 84.6% versus 98.6%, P = 0.01, respectively).
Univariate analysis identified allograft weight as the only clinical parameter that was significantly associated with the onset of EAD (P = 0.02). The median allograft weight was 1.5 and 1.3 kg for the EAD and non-EAD groups, respectively. The EAD group also tended to have older donor age (40 versus 36 y) and a higher proportion of 30% or more macrosteatotic grafts (11.5% versus 2.9%) than the non-EAD group, although these differences did not reach statistical significances (P = 0.09 and P = 0.07, respectively).
Pretransplant Intrahepatic Protein Profiles Correlated With EAD
The expression patterns of 7 studied proteins in pretransplant biopsy specimens were examined by IHC staining. Cytoplasmic expression of NOX-1 and cytochrome C was observed in hepatocytes throughout the hepatic lobules, whereas the distribution of VEGF immunoreactivity was mainly localized to the cytoplasm of hepatocytes surrounding the central veins. Immunostaining showed that ICAM-1, vascular cell adhesion molecule (VCAM-1), NOX-2, and endothelin-1 were all predominantly localized to the hepatic sinusoids. ICAM-1 and VCAM-1 were expressed on the surface of sinusoidal endothelium, whereas NOX-2 and endothelin-1 showed membranous staining. DCD liver grafts with ≥30% macrosteatosis were associated with increased intrahepatic expression of cytochrome C (87.5% versus 38.9%, P = 0.01) and VEGF (75% versus 31%, P = 0.02) compared with grafts with <30% macrosteatosis.
The expression levels of these proteins were compared between EAD and control grafts (Table 3). Compared with the non-EAD grafts, EAD grafts were associated with significantly higher expression of intrahepatic NOX-1 (27.5% versus 46.2%, P = 0.034), ICAM-1 (23.2% versus 42.3%, P = 0.025), cytochrome C (33.3% versus 53.8%, P = 0.024), and VEGF (21.7% versus 50.0%, P = 0.001). Examples of IHC staining in the graft biopsy specimens from the EAD and non-EAD groups are shown in Figure 1.
Allograft Weight and Intrahepatic VEGF Expression Were Independent EAD Predictors
Multivariate logistic regression analysis showed that allograft weight (per gram) (odds ratio [OR] = 1.003; 95% confidence interval [CI] = 1.001-1.004; P = 0.003) and high VEGF expression (OR = 3.74; 95% CI = 1.494-9.365; P = 0.005) were 2 independent predictors of EAD development (Table 4).
The distribution of allograft weight did not vary by the expression of intrahepatic VEGF (Figure 2A) but was found to be mildly correlated with operative time (Spearman r = 0.19; P = 0.04; Figure 2B). An allograft weight of 1.35 kg was selected as the cutoff point in predicting EAD, as evaluated by the receiver operating characteristic (ROC) curve (data not shown). The proportion of patients who developed EAD increased from 29.6% in patients with an allograft weight <1.35 kg to 48.3% in those with allograft weights between 1.35 kg and 1.50 kg; this rate continued to increase to 57.9% in patients with allograft weights >1.5 kg (Figure 2C).
VEGF overexpression was associated with continuously increased peak aspartate transaminase (on postoperative days [PODs] 1, 2, 5, and 7), alanine aminotransferase (on PODs 2 and 5), bilirubin (on PODs 1, 2, 5, and 7), and prolonged international normalized ratio (on PODs 1 and 2) within the first week posttransplantation (Figure 3A–D). Patients who received grafts with high VEGF expression displayed worse survival than those receiving grafts with low VEGF expression (Log-rank: P = 0.004; Figure 3E). The 6-, 12-, and 18-month cumulative survival rates were 96.2%, 93.6%, and 93.6% for recipients of grafts with low VEGF expression and 87.8%, 82.9%, and 75.6% for those of grafts with high VEGF expression, respectively. Graft survival was also significantly diminished when intrahepatic VEGF was overexpressed (Log-rank: P = 0.001; Figure 3F).
To address whether the negative effects of EAD were potentially enhanced in the setting of high VEGF expression, the study cohort was divided into the following groups: group A (n = 54), patients who received grafts with low VEGF expression and did not experience EAD; group B (n = 15), patients who received grafts with high VEGF expression but did not experience EAD; group C (n = 26), EAD patients who received grafts with low VEGF expression; and group D (n = 26), EAD patients who received grafts with high VEGF expression. As illustrated in Figure 4, EAD patients who received grafts with high VEGF expression had the worst survival prognosis. Moreover, these patients showed markedly increased frequencies of experiencing AKI and RRT.
EAD Risk Assessment Model and Model Validation
A risk assessment model predictive of EAD occurrence was established and is summarized as the following equation: EAD risk assessment model score = 2.584 × allograft weight (in kg) + 1.319 × VEGF − 5.959 (VEGF, 0 for low expression, 1 for high expression).
The predictive model discriminated well (area under ROC curve: 0.741, 95% CI = 0.649-0.834; P < 0.001; Figure 5A) and showed a goodness fit (P = 0.616 to reject model). The cutoff value of the score was set as −1.72, which had a sensitivity and specificity of 67.3% and 78.3%, respectively. Seventy-one patients were included in the EAD risk assessment model score of <−1.72 group and 50 in the score of ≥−1.72 group. A significant difference in the incidence of EAD was observed between the 2 groups. In the group with scores <−1.72, only 23.9% (17/71) of patients experienced EAD, whereas EAD occurred in 70% (35/50) of patients in the group with scores ≥−1.72 (P < 0.001).
We then applied the EAD risk assessment model to the validation cohort. After calculating the risk score for each patient, the model achieved an area under the ROC curve of 0.773 (95% CI = 0.610-0.937; P = 0.006; Figure 5B) between EAD and the control, displaying significant potential for predicting EAD. Moreover, the application of −1.72 as a cutoff score in this group also resulted in an excellent discriminative ability. The incidence of EAD was 19.0% (4/21) and 62.5% (10/16) for recipients with scores <−1.72 and ≥−1.72, respectively (P = 0.007).
Cox Survival Analysis
In the univariate Cox analysis, EAD, RRT requirement, recipient Model for End-stage Liver Disease (MELD) score, allograft weight, macrovesicular steatosis ≥30%, high VEGF, and high NOX-1 were associated with increased risk of graft loss (Table 5). These variables were included in multivariate analyses, which showed that RRT requirement, allograft weight, and high VEGF were independent variables. The interaction between recipient MELD score and VEGF was found to be significant (P = 0.014). To eliminate this effect, we excluded 7 cases of donor/recipient mismatch (5.8%), defined as D-MELD ≥1600, a product of donor age and recipient MELD that has been well documented as a simple predictor of post-LT survival and donor/recipient mismatch,17 instead of recipients with high MELD alone, since the cutoff values to define high MELD varied by studies and were arbitrarily set.
In the D-MELD <1600 subgroup (n = 115), recipient MELD and RRT requirement were no longer significant in the univariate analysis, and only allograft weight and high VEGF expression served as independent risk factors of graft loss in the multivariate analysis (allograft weight: OR = 1.002, 95% CI = 1.001-1.004; high VEGF: OR = 9.247, 95% CI = 2.462-34.723).
In addition, Cox analysis of patient survival was performed (Table S1, SDC, http://links.lww.com/TP/B748). The results were similar to those of the graft survival analysis, except that EAD and macrovesicular steatosis ≥30% in the univariate Cox analysis and RRT requirement in multivariate Cox analysis were nonsignificant factors of patient morbidity. In the D-MELD <1600 subgroup, only allograft weight and high VEGF expression were significant in the univariate analysis and remained independent factors in the multivariate analysis (allograft weight: OR = 1.003, 95% CI = 1.001-1.004; high VEGF: OR = 8.369, 95% CI = 2.186-32.041).
EAD represents a common but undesirable event after LT. In a cohort study from 3 US transplant programs, Olthoff et al6 proposed an updated definition of EAD that identified an incidence of 23.2% and was associated with increased 6-month mortality.6 The definition was further confirmed in other institutions and in living donor LT.18,19 Evolving clinical data revealed that DCD allografts have an increased likelihood of experiencing EAD. A series analysis of 205 patients who received DCD allografts at the Mayo clinic reported an EAD incidence as high as 39.5%.5 Likewise, the present study suggested that EAD had a relatively higher incidence following LT from DCD (43.0%) and was linked to inferior 6-month recipient and graft survival. Our data also confirmed previous studies that EAD patients were at increased risk of developing AKI and requiring RRT.7
In this study of adult LT from DCD, intrahepatic VEGF expression and allograft weight constituted 2 independent predictors of EAD and were also found to carry important prognostic implications in well-matched transplant combination (D-MELD <1600). Based on these 2 variables, we further established an EAD risk assessment model that enables transplant physicians to accurately assess the EAD risk of each implanted liver. Interestingly, our analysis showed that the combination of high VEGF expression and the onset of EAD significantly deteriorated renal function and decreased patient and graft survival, indicating an advanced stage of EAD in the setting of VEGF overexpression. This knowledge provides new insights into the molecular mechanism of EAD development and progression and would assist in optimal immediate clinical decision-making for this set of patients, including closer surveillance of post-LT hepatorenal function and consideration of early therapeutic interventions. Based on this evidence, the measurement of intrahepatic VEGF expression may be utilized as a convenient and effective means to risk stratify DCD allografts.
In the current investigation, large allograft weight was found to be independently associated with EAD occurrence. Consistently, a large series study of 1950 consecutive LTs disclosed that large allograft weight was an independent risk factor for EAD; the median liver allograft mass in recipients with and without EAD were 1.7 and 1.5 kg, respectively.18 A large liver graft would receive relatively low portal blood flow, leading to intrahepatic microcirculatory hypoperfusion, sustained warm ischemia injury, and eventually, compromised graft function.20 External compression from the recipient abdominal cavity further contributes to the impaired blood supply to the grafts. In a serious condition, large-for-size syndrome may be complicated.21 Additionally, as indicated by the mildly positive correlation between allograft weight and operative time, transplantation of an oversized whole liver graft represents a technological challenge. Adverse consequences include difficult anastomosis, delayed fascial closure, poor vascular alignment, and prolonged revascularization time, the last of which is a recently identified indicator of unfavorable transplant outcomes.22
Not only large grafts but also small grafts were shown to have an adverse impact on clinical outcome, namely small-for-size syndrome. In animal experiments of rat LT, transplantations of either stepwise increased or reduced graft weight to recipient weight (GWRW) were both found to result in a stepwise decrease in recipient survival; furthermore, recipients of increased GWRW tended to exhibit elevated serum transaminase and impaired bile flow compared with those in recipients of reduced GWRW, as well as a lower survival rate.23 This nonlinear or potentially U-shaped correlation driven by size mismatch may account for the statistically nonsignificant impact of GWRW on the development of EAD.24 Taken together, these data indicate that liver grafts with weights in the appropriate range are generally preferred. Nonetheless, determination of the optimal weight range is beyond the scope of the present study, which merits further investigation with larger cohorts and may consider geographic variation.
The expression of intrahepatic VEGF is constitutively low under basal conditions, whereas it increases dramatically in response to ischemic stress.25 Studies of orthotropic LT rodent models have documented that intrahepatic VEGF expression and serum release increased exponentially in the early stage of graft damage.26,27 In this clinical investigation, we demonstrated that intrahepatic VEGF overexpression was associated with a 2.74-fold increase in the risk of developing EAD and, furthermore, increased posttransplant mortality and morbidity in EAD patients. VEGF plays important roles in liver ischemia injury by inducing the activity of extracellular matrix proteases and enhancing microvascular permeability and endothelial fenestration, while it is also involved in the tissue repair process by promoting local angiogenesis and boosting hepatocyte proliferation.25,28,29 A dual role of VEGF in graft injury was further reinforced by the therapeutic benefits achieved by the endogenous blockage and exogenous administration of VEGF. Endogenous blockage of locally produced VEGF could effectively decrease lipid peroxidation and gelatinase activity along with the biochemical parameters of graft injury.26 Exogenous systematic administration of recombinant VEGF could upregulate the intrahepatic expression of inducible NO synthase and attenuate graft damage.27 The analogous beneficial effect is also exhibited via exogenous VEGF gene delivery, which increased the expression of hepatotrophic mitogen and improved graft function.30
High intrahepatic VEGF expression was also recognized as a negative predictor of recipient and graft survival. Significantly increased serum transaminase, persistent cholestasis, and prolonged coagulopathy in recipients of grafts with high VEGF expression provided evidence of greater liver graft damage in these patients. Furthermore, regarding the presence of micrometastases in hepatocellular carcinoma patients at the time of LT and the interaction between tumor cells and the microenvironment in vivo, regeneration of donor liver in the context of cancer would theoretically foster tumor cell proliferation. Previous clinical studies have demonstrated that recipients with high plasma VEGF levels (>44 pg/mL) before transplantation had significantly increased tumor recurrence rates and inferior survival outcomes.31
Other protein biomarkers, such as NOX-1, cytochrome C, and ICAM-1, were potentially related to the development of EAD but failed to reach statistical significance in the multivariate analysis. Increasing experimental evidence has unraveled the importance of their roles in liver graft injury.32-34 However, most studies focused on the molecular milieu after graft revascularization when the detrimental effects of these proteins were essentially amplified by the reperfusion process. By contrast, as we chose the timing before implantation for assessment of protein expression, a reperfusion event was not involved. The predictive values of these proteins may be underestimated and presumably be better recognized with examination of the postreperfusion milieu.
Although the presence of EAD, as defined by Olthoff et al,6 is a well-known negative predictor of 6-month survival outcomes, its impact on longer-term survival is controversial. By employing the same definition, some studies suggested that EAD is a surrogate marker of poor long-term survival.5 However, the correlation between EAD and overall survival was less pronounced in the current cohort, which was consistent with prior reports.35,36 Instead, allograft weight and high VEGF expression, 2 independent predictors of EAD, were associated with increased patient morbidity and graft loss. Similar situations have also been observed in other reports. For example, post-LT recipient serum V factor, another biomarker of EAD, could continuously predict short- and mid term survival, while EAD failed.10 One possible explanation may be the limited discriminative capability of Olthoff’s EAD definition. Under this binary definition, all patients were categorized as either presenting EAD or not, which therefore was insufficient to grade liver function and failed to capture the disease spectrum of graft dysfunction. For continuous grading of EAD, several innovated definitions have recently followed, including the MEAF score and L-GrAFT risk score.37,38 Unfortunately, the formulas of these definitions for score calculation are extremely complicated and have been scarcely validated in other centers. Further studies are warranted to construct a simple and precise EAD definition for better grading of disease severity.
In the context of LT, pretransplant damage to liver grafts exerts markedly negative impacts on graft quality and postoperative courses.39,40 Theoretically, the severity of graft injury could be assessed by histological parameters of liver biopsy. However, a comprehensive and well-established criterion for describing the histological morphology is not yet available for transplant pathologists to judge whether the graft quality is good or injured.41 Furthermore, liver biopsy before implantation often yielded essentially normal histological morphology.40,42 Although sinusoidal neutrophilic infiltrate was identified to indicate inferior transplant outcomes in our previous study, it was detected only in 7.1% of the biopsy samples and failed to predict EAD.43 Indeed, even after engraftment when reperfusion aggregated liver damage, only severe histological injury portended decreased recipient survival but was observed in a small proportion (<5%) of the biopsy specimen; neither mild nor moderate form was of relevance.44 This disappointing fact may reflect that morphological injury could manifest either immediately or progressively, and progressive injury is not detectable by morphological examinations of perioperative biopsy. In contrast, to delay the morphology changes, changes in the intragraft molecular profiles may be sensitive, reliable, and respond quickly after organ extractions.
As expected, our study demonstrated that pretransplant intrahepatic protein profiles predicted early graft performance and recipient outcome, with IHC performed on liver biopsy. Notably, the clinical applicability of this strategy is promising, given that pretransplant liver biopsies are routinely performed at a large number of LT centers. As a protein detection method, IHC is characterized by its cost-effectiveness and has universally served as an indispensable element of the practice of diagnostic histopathology. Therefore, it may provide novel clinical means to forecast post-LT outcomes based on preoperative intrahepatic protein profiles. There are also concerns regarding the extension of practical implications to pre-LT graft quality assessment or, furthermore, organ allocation scheme. Unfortunately, IHC technology is relatively time consuming, with 2 or 3 days generally required to obtain immunostaining results. The fact that the results of the IHC analysis were not available before transplantation may not facilitate the clinical decisions of accepting or discarding an organ when offered. The primary practicalities of our strategy may be to predict the development and progression of EAD and short- and mid term survival at an early stage following the implantation of a specific liver graft.
There are several limitations to be acknowledged. The study cohort was based on single-center experience, and accordingly, selection bias may be introduced. However, this may avoid multicenter-based confounding factors typical of surgical practice and perioperative management. The final predictive model incorporated only 2 variables of graft characteristics. Therefore, the predictive capabilities of the model may be potentially weakened, but the importance of liver graft itself is accentuated and model complexity is greatly eliminated. Notably, as our study is preliminary, the numbers of investigated proteins are limited (n = 7), and the EAD risk assessment model has not been validated in independent cohorts at other centers. More research is warranted to determine the potential values of other protein markers and for external validation of this predictive model.
In summary, this is the first study demonstrating that intrahepatic protein profiling provides valuable information about graft performance and recipient clinical outcomes. Intrahepatic VEGF was implicated to play an important role in the development and progression of EAD. The mathematical model presented in this study could be utilized to provide accurate estimation of EAD.
We thank Guoqiang Chao and Bingjie Chen for their help in the management of liver biopsy samples and Qi Ling for providing statistical advice.
1. Busuttil RW, Tanaka K. The utility of marginal donors in liver transplantation. Liver Transpl. 2003; 9:651–663
2. Saidi RF, Bradley J, Greer D, et al. Changing pattern of organ donation at a single center: are potential brain dead donors being lost to donation after cardiac death? Am J Transplant. 2010; 10:2536–2540
3. Goldberg DS, Abt PL. Improving outcomes in DCDD liver transplantation: there can only be strength in numbers. Am J Transplant. 2014; 14:1016–1020
4. Deschenes M. Early allograft dysfunction: causes, recognition, and management. Liver Transpl. 2013; 19(Suppl 2):S6–S8
5. Lee DD, Singh A, Burns JM, et al. Early allograft dysfunction in liver transplantation with donation after cardiac death donors results in inferior survival. Liver Transpl. 2014; 20:1447–1453
6. Olthoff KM, Kulik L, Samstein B, et al. Validation of a current definition of early allograft dysfunction in liver transplant recipients and analysis of risk factors. Liver Transpl. 2010; 16:943–949
7. Wadei HM, Lee DD, Croome KP, et al. Early allograft dysfunction after liver transplantation is associated with short- and long-term kidney function impairment. Am J Transplant. 2016; 16:850–859
8. Croome KP, Hernandez-Alejandro R, Chandok N. Early allograft dysfunction is associated with excess resource utilization after liver transplantation. Transplant Proc. 2013; 45:259–264
9. Friedman BH, Wolf JH, Wang L, et al. Serum cytokine profiles associated with early allograft dysfunction in patients undergoing liver transplantation. Liver Transpl. 2012; 18:166–176
10. Gorgen A, Prediger C, Prediger JE, et al. Serum factor V is a continuous biomarker of graft dysfunction and a predictor of graft loss after liver transplantation. Transplantation. 2019; 103:944–951
11. Kim DG, Lee JY, Jung YB, et al. Clinical significance of lactate clearance for the development of early allograft dysfunction and short-term prognosis in deceased donor liver transplantation. Clin Transplant. 2017; 31:e13136
12. Selten JW, Verhoeven CJ, Heedfeld V, et al. The release of microRNA-122 during liver preservation is associated with early allograft dysfunction and graft survival after transplantation. Liver Transpl. 2017; 23:946–956
13. Pulitano C, Joseph D, Sandroussi C, et al. Postreperfusion microcirculatory derangements after liver transplantation: relationship to hemodynamics, serum mediators, and outcome. Liver Transpl. 2017; 23:527–536
14. Abu-Amara M, Yang SY, Tapuria N, et al. Liver ischemia/reperfusion injury: processes in inflammatory networks–a review. Liver Transpl. 2010; 16:1016–1032
15. Peralta C, Jiménez-Castro MB, Gracia-Sancho J. Hepatic ischemia and reperfusion injury: effects on the liver sinusoidal milieu. J Hepatol. 2013; 59:1094–1106
16. Zhai Y, Petrowsky H, Hong JC, et al. Ischaemia-reperfusion injury in liver transplantation–from bench to bedside. Nat Rev Gastroenterol Hepatol. 2013; 10:79–89
17. Halldorson JB, Bakthavatsalam R, Fix O, et al. D-MELD, a simple predictor of post liver transplant mortality for optimization of donor/recipient matching. Am J Transplant. 2009; 9:318–326
18. Lee DD, Croome KP, Shalev JA, et al. Early allograft dysfunction after liver transplantation: an intermediate outcome measure for targeted improvements. Ann Hepatol. 2016; 15:53–60
19. Pomposelli JJ, Goodrich NP, Emond JC, et al. Patterns of early allograft dysfunction in adult live donor liver transplantation: the A2ALL experience. Transplantation. 2016; 100:1490–1499
20. Rangel Moreira Dde A, Aoun Tannuri AC, Belon AR, et al. Large-for-size liver transplantation: a flowmetry study in pigs. J Surg Res. 2014; 189:313–320
21. Allard MA, Lopes F, Frosio F, et al. Extreme large-for-size syndrome after adult liver transplantation: a model for predicting a potentially lethal complication. Liver Transpl. 2017; 23:1294–1304
22. Buchholz BM, Gerlach UA, Chandrabalan VV, et al. Revascularization time in liver transplantation: independent prediction of inferior short- and long-term outcomes by prolonged graft implantation. Transplantation. 2018; 102:2038–2055
23. Xu HS, Pruett TL, Jones RS. Study of donor-recipient liver size match for transplantation. Ann Surg. 1994; 219:46–50
24. Fukazawa K, Nishida S. Size mismatch in liver transplantation. J Hepatobiliary Pancreat Sci. 2016; 23:457–466
25. Fernández M, Semela D, Bruix J, et al. Angiogenesis in liver disease. J Hepatol. 2009; 50:604–620
26. Boros P, Tarcsafalvi A, Wang L, et al. Intrahepatic expression and release of vascular endothelial growth factor following orthotopic liver transplantation in the rat. Transplantation. 2001; 72:805–811
27. Tsurui Y, Sho M, Kuzumoto Y, et al. Dual role of vascular endothelial growth factor in hepatic ischemia-reperfusion injury. Transplantation. 2005; 79:1110–1115
28. LeCouter J, Moritz DR, Li B, et al. Angiogenesis-independent endothelial protection of liver: role of VEGFR-1. Science. 2003; 299:890–893
29. Roberts WG, Palade GE. Increased microvascular permeability and endothelial fenestration induced by vascular endothelial growth factor. J Cell Sci. 1995; 108Pt 62369–2379
30. Cao D, Wang M, Gong J, et al. Exogenous vascular endothelial growth factor delivery prior to endothelial precursor cell transplantation in orthotopic liver transplantation-induced hepatic ischemia/reperfusion injury. Liver Transpl. 2017; 23:804–812
31. Zhang W, Kim R, Quintini C, et al. Prognostic role of plasma vascular endothelial growth factor in patients with hepatocellular carcinoma undergoing liver transplantation. Liver Transpl. 2015; 21:101–111
32. Rentsch M, Post S, Palma P, et al. Anti-ICAM-1 blockade reduces postsinusoidal WBC adherence following cold ischemia and reperfusion, but does not improve early graft function in rat liver transplantation. J Hepatol. 2000; 32:821–828
33. Lambeth JD. NOX enzymes and the biology of reactive oxygen. Nat Rev Immunol. 2004; 4:181–189
34. Weigand K, Brost S, Steinebrunner N, et al. Ischemia/reperfusion injury in liver surgery and transplantation: pathophysiology. HPB Surg. 2012; 2012:176723
35. Hudcova J, Scopa C, Rashid J, et al. Effect of early allograft dysfunction on outcomes following liver transplantation. Clin Transplant. 2017; 31:e12887
36. Croome KP, Wall W, Quan D, et al. Evaluation of the updated definition of early allograft dysfunction in donation after brain death and donation after cardiac death liver allografts. Hepatobiliary Pancreat Dis Int. 2012; 11:372–376
37. Agopian VG, Harlander-Locke MP, Markovic D, et al. Evaluation of early allograft function using the liver graft assessment following transplantation risk score model. JAMA Surg. 2018; 153:436–444
38. Pareja E, Cortes M, Hervás D, et al. A score model for the continuous grading of early allograft dysfunction severity. Liver Transpl. 2015; 21:38–46
39. Briceño J, Marchal T, Padillo J, et al. Influence of marginal donors on liver preservation injury. Transplantation. 2002; 74:522–526
40. Gaffey MJ, Boyd JC, Traweek ST, et al. Predictive value of intraoperative biopsies and liver function tests for preservation injury in orthotopic liver transplantation. Hepatology. 1997; 25:184–189
41. Verhoeven CJ, Farid WR, de Jonge J, et al. Biomarkers to assess graft quality during conventional and machine preservation in liver transplantation. J Hepatol. 2014; 61:672–684
42. Kakizoe S, Yanaga K, Starzl TE, et al. Evaluation of protocol before transplantation and after reperfusion biopsies from human orthotopic liver allografts: considerations of preservation and early immunological injury. Hepatology. 1990; 11:932–941
43. Xia W, Ke Q, Wang Y, et al. Donation after cardiac death liver transplantation: graft quality evaluation based on pretransplant liver biopsy. Liver Transpl. 2015; 21:838–846
44. Ali JM, Davies SE, Brais RJ, et al. Analysis of ischemia/reperfusion injury in time-zero biopsies predicts liver allograft outcomes. Liver Transpl. 2015; 21:487–499