Beyond graft function impairment after liver transplantation: the prolonged cold ischemia time impact on recurrence of hepatocellular carcinoma after liver transplantation—a single-center retrospective study

Introduction

Hepatocellular carcinoma (HCC) is an aggressive tumor and the fourth leading cause of cancer-related death globally (Villanueva, 2019). More than 700,000 people die from HCC yearly. Theoretically, liver transplantation (LT) is an optimum treatment for HCC because it can eliminate both the tumor and underlying liver disease that progresses to cirrhosis. LT can eliminate the primary lesions, such as minute lesions that may be difficult to identify. It can also exterminate the possible cancerization of the rest of the liver. However, early HCC recurrence occurs after LT, which might result from micrometastasis when tumor cells are transported through the circulation (Choi et al., 2015). The 10-year recurrence-free survival rate after LT is around 50%–70% (Adam et al., 2018; Pinna et al., 2018). Prolonged waiting time is common among HCC patients diagnosed with LT, and this issue is a cause of tumor progression (Mehta et al., 2017).

The proportion of clinical use of extended criteria donors (ECDs) and the adoption of grafts with prolonged cold ischemia time (CIT) are increasing, considering the shortage of liver grafts, which may consequently lead to serious ischemia-reperfusion injury (IRI). Previous studies have demonstrated that prolonged cold ischemic time can greatly impair graft survival (Chini et al., 2021; Cassuto et al., 2008; Busuttil et al., 2005). Moreover, IRI has a positive correlation with HCC recurrence (Nagai et al., 2015; Grat et al., 2018). The use of ECD-derived grafts in patients with HCC can also accelerate tumor recurrence after LT (Chini et al., 2021; Cassuto et al., 2008; Busuttil et al., 2005). This retrospective study examined LT in patients with HCC to confirm the relationship between CIT and HCC recurrence.

Patients and Methods

Results

Effects of CIT on IRI

We grouped the patients into two: CIT > 568 min and CIT ≤ 568 min according to the optimal cutoff value of the ROC, peak ALT/AST levels, and total bilirubin on POD7. The differences in the rate of EAD was statistically significant (Table 2). Patients with CIT > 568 min showed significantly higher AST, ALT, and TB levels after LT (AST, 712.5 vs. 390 U/L, p < 0.001; ALT, 957 vs. 593 U/L, p < 0.001; TB, 134.05 vs. 99.5 umol/L, p = 0.012; Figs. 1A–1C). However, no statistical differences were found in Crea level and INR level between the two groups (p = 0.373, p = 0.277, respectively, Figs. 1D, 1E). The EAD incidence in the group with CIT > 568 min was 50%, which was significantly higher than in the group with CIT ≤ 568 min (33.0%) (p = 0.006; Fig. 1F). The incidence of biliary complications was 5.6% in the CIT > 568 min group, and 8.6% in CIT ≤ 568 min group. The statistical results also demonstrated that prolonged CIT resulted in more fresh frozen plasma (FFP) infusion during the operation (p = 0.02) and longer postoperative respiratory support time (p = 0.045).

Table 2:

Differences in factors represent IRI in clinic between CIT>568 min and CIT ≤568 min.

	CIT > 568(n = 72)	CIT ≤568(n = 348)	P
Peak 7-day postoperative ALT level (U/L)	957 (552.75–1,355.5)	593 (322.5–957.1)	0.000
Peak 7-day postoperative AST level (U/L)	712.5 (381–1,374.25)	390 (219.5–917.5)	0.000
Peak 7-day postoperative TBIL level (umol/L)	134.05 (72.525–209.35)	99.5 (57.375–167.225)	0.012
Peak 7-day postoperative Crea level (umol/L)	90 (74–125)	89 (71–121.75)	0.373
Peak 7-day postoperative INR level (S)	1.525 (1.31–1.75)	1.46 (1.3125–1.6875)	0.277
Biliary complication	4 (5.6%)	30 (8.6%)	0.385
Acute rejection	3 (4.2%)	15 (4.3%)	0.956
EAD	36 (50%)	115 (33.0%)	0.006
Postoperative respiratory support (hour)	20 (12–45)	16 (11–28)	0.045
Infusion of FFP (Unit)	7 (5–12)	6.5 (4–9)	0.02

DOI: 10.7717/peerj.18126/table-2

Notes:

Abbreviations

AST

serum aspartate aminotransferase

ALT

serum alanine aminotransferase

TBIL

total bilirubin

Crea

creatinine

INR

international normalized ratio

EAD

early allograft dysfunction

FFP

fresh frozen plasma

CIT > 568 min is an independent risk factor for DFS but not for OS

Table 5 lists the parameters related to risk of disease-free survival (DSF). Multivariate Cox regression analysis showed that the recipient gender (male) (OR, 1.774; 95% CI [1.241–2.536]; p = 0.002), AFP ≥ 400 (ng/ml) (OR, 2.032; 95% CI [1.415–2.920]; p = 0.009), exceeding Milan criteria (OR, 1.782; 95% CI [1.156–2.747]; p = 0.009) and poor tumor differentiation (OR, 1.48; 95% CI [1.041–2.116]; p = 0.029) were independent risk factors for DSF (Table 6). DFS rates at 1-and 3-year in CIT > 568 min group were 71.9% and 45.7%, respectively; 79.2% and 66.3%, respectively in CIT ≤568 min group (p = 0.018, Fig. 2A). Table 6 lists the parameters related to risk of overall survival (OS), and this is significant in univariate analysis. In multivariate Cox regression analysis, intraoperative PRBC transfusions (unit) (OR, 1.046; 95% CI [1.018–1.076]; p = 0.001), AFP ≥ 400 (ng/ml) (OR, 2.42; 95% CI [1.648–3.554]; p < 0.001) and poor tumor differentiation (OR, 1.48; 95% CI [1.041–2.116]; p = 0.017) were independent risk factors for overall survival (Table 5). OS rates at 1-and 5-year in CIT > 568 min group were 86% and 53%, respectively; 87.4% and 54.1%, respectively in CIT ≤568 min group (p = 0.533, Fig. 2B).

Table 5:

Prognostic factors for disease-free survival on univariate and multivariate analysis.

	Univariate		Multivariate
Variables	OR (95% CI)	P value*	OR (95% CI)	P value*
Donor characteristics
Age (years)	1.004 (0.992–1.017)	0.485
Body mass index (kg/m2)	0.996 (0.974–1.018)	0.714
Gender (male)	1.247 (0.821–1.893)	0.3
Warm ischemia time (minutes)(80)	0.944 (0.844–1.057)	0.318
Cold ischemia time >568	1.688 (1.146–2.484)	0.008	1.652 (1.111–2.458)	0.013
Sodium level before harvest (mmol/L)	0.999 (0.994–1.004)	0.566
ALT level before harvest (U/L)	1.001 (1.000–1.002)	0.125
AST level before harvest (U/L)	1.0 (0.999–1.001)	0.807
TBIL level before harvest (umol/L)	0.998 (0.989–1.007)	0.651
DCD	0.866 (0.574–1.309)	0.495
Recipient characteristics
Age (years)	0.986 (0.970–1.003)	0.098
Body mass index (kg/m2)	1.046 (0.995–1.101)	0.080
Gender (male)	0.982 (0.588–1.640)	0.982
Complications
Hypertension	1.627 (0.899–2.945)	0.108
Diabetes	1.951 (0.991–3.839)	0.053
Coronary heart disease	1.028 (0.327–3.232)	0.963
MELD	1.022 (1.003–1.041)	0.022	1.013 (0.995–1.032)	0.155
ALT level before LT (U/L)	1.0 (0.999–1.000)	0.544
AST level before LT (U/L)	1 (1–1)	0.868
TBIL level before LT (umol/L)	1.002 (1.001–1.003)	0.052
Operation time (min)	1.000 (0.999–1.002)	0.885
Intraoperative blood loss (ml)	1 (1–1)	0.687
Intraoperative PRBC transfusions (unit)	1.013 (0.987–1.042)	0.391
anhepatic period (min)	1.001 (0.996–1.005)	0.817
Peak 7-day postoperative ALT level (U/L)	1 (1–1)	0.256
Peak 7-day postoperative AST level (U/L)	1 (1–1)	0.205
EAD	1.219 (0.85–1.749)	0.282
Peak 7-day postoperative Creatinine level (umol/L)	1.0 (0.998–1.002)	0.923
Peak 7-day postoperative INR level	0.999 (0.997–1.001)	0.307
	0.872 (0.602–1.259)	0.872
Biliary complication	1.065 (0.574–1.974)	0.842
Acute rejection	1.298 (0.572–2.950)	0.533
Infusion of FFP (Unit)	0.999 (0.993–1.004)	0.611
Exceeding Milan criteriaa	2.085 (1.366–3.184)	0.001	1.742 (1.129–2.687)	0.012
Previous TACE	1.039 (0.729–1.481)	0.832
Previous radiofrequency ablation	0.877 (0.593–1.296)	0.509
Hepatectomy before LT	1.026 (0.637–1.651)	0.917
AFP (ng/ml) ≥400	2.953 (2.093–4.166)	0.000	2.410 (1.679–3.461)	0.000
Tumor numbers >3	1.668 (1.185–2.347)	0.003	NA※	NA※
Tumor size (mm)>50	1.518 (1.078–2.136)	0.017	NA※	NA※
Poor tumor differentiation	1.749 (1.236–2.475)	0.002	1.547 (1.086–2.203)	0.016
Microscopic vascular invasion	3.564 (2.524–5.033)	0.000	NA※	NA※

DOI: 10.7717/peerj.18126/table-5

Notes:

*Cox’s proportional regression analysis.

NA※, Tumor size, number of tumor, and MVI were not included to avoid colinearity, as they are included in Milan criteria.

Abbreviations

DCD

donation after circulatory death

AFP

indicates serum α-fetoprotein

TACE

transarterial chemoembolization

MELD

model for end-stage liver disease

EAD

early allograft dysfunction

LT

liver transplant

HCC

hepatocellular carcinoma

PRBC

packed red blood cells

AST

serum aspartate aminotransferase

ALT

serum alanine aminotransferase

TBIL

total bilirubin

INR

international normalized ratio

FFP

fresh frozen plasma

Table 6:

Prognostic factors for overall survival on univariate and multivariate analysis.

Variables	Univariate		Multivariate
	OR (95% CI)	Pvalue*	OR (95% CI)	Pvalue*
Donor characteristics
Age (years)	1.01(0.997–1.023)	0.137
Body mass index (kg/m2)	1 (0.985–1.015)	0.986
Gender (male)	1.069 (0.701–1.629)	0.757
Warm ischemia time (minutes)(80)	0.913 (0.823–1.012)	0.084
Cold ischemia time >568	1.143 (0.751–1.738)	0.533
Sodium level before harvest (mmol/L)	0.997 (0.989–1.006)	0.559
ALT level before harvest (U/L)	1.001 (1.000–1.002)	0.046	NA#	NA#
AST level before harvest (U/L)	1.0 (1.0–1.001)	0.191
TBIL level before harvest (umol/L)	1.005 (0.998–1.013)	0.143
DCD	0.866 (0.574–1.309)	0.495
Recipient characteristics
Age (years)	1.002 (0.985–1.020)	0.789
Body mass index (kg/m2)	1.021 (0.966–1.078)	0.467
Gender (male)	0.962 (0.585–1.581)	0.877
Complications
Hypertension	1.315 (0.805–2.148)	0.273
Diabetes	1.187 (0.654–2.155)	0.572
Coronary heart disease	1.316 (0.325–5.336)	0.7
MELD	1.028 (1.009–1.047)	0.004	1.016 (0.996–1.036)	0.114
ALT level before LT (U/L)	1.0	0.510
AST level before LT (U/L)	1	0.858
TBIL level before LT (umol/L)	1.002 (1.001–1.003)	0.000
Operation time (min)	1.001 (0.999–1.002)	0.45
Intraoperative blood loss (ml)	1 (1–1)	0.084
Intraoperative PRBC transfusions (unit)	1.036 (1.010–1.063)	0.007	1.048 (1.019–1.078)	0.001
anhepatic period (min)	1.00 (0.996–1.005)	0.863
Peak 7-day postoperative ALT level (U/L)	1 (1–1)	0.000	NA#	NA#
Peak 7-day postoperative AST level (U/L)	1 (1–1)	0.034	NA#	NA#
EAD	1.172 (0.819–1.677)	0.385
Peak 7-day postoperative Bilirubin level (umol/L)	1.001 (1.000–1.003)	0.104
Peak 7-day postoperative Creatinine level (umol/L)	0.999 (0.997–1.002)	0.610
Peak 7-day postoperative INR level	0.975 (0.857–1.110)	0.706
Biliary complication	1.045 (0.547–1.996)	0.895
Acute rejection	1.230 (0.501–3.018)	0.651
Infusion of FFP (Unit)	1 (0.999–1.001)	0.938
Exceeding Milan criteriaa	1.8 (1.174–2.759)	0.007	1.388(0.895–2.154)	0.143
Previous TACE	1.419 (0.962–2.092)	0.078
Previous radiofrequency ablation	1.241 (0.794–1.940)	0.344
Hepatectomy before LT	1.25 (0.727–2.147)	0.42
AFP (ng/ml) ≥400	3.195 (2.239–4.558)	0.000	2.915 (1.982–4.287)	0.000
Tumor numbers >3	1.819 (1.276–2.594)	0.001	NA※	NA※
Tumor size (mm)>50	1.557 (1.093–2.218)	0.014	NA※	NA※
Poor tumor differentiation	1.921 (1.343–2.746)	0.000	1.577 (1.083–2.298)	0.018
Microscopic vascular invasion	3.038 (2.115–4.362)	0.000	NA※	NA※

DOI: 10.7717/peerj.18126/table-6

Notes:

*Cox’s proportional regression analysis.

NA#not applicable.

NA※, Tumor size, number of tumor, and MVI were not included to avoid colinearity, as they are included in Milan criteria.

Abbreviations

DCD

donation after circulatory death

AFP

indicates serum α-fetoprotein

TACE

transarterial chemoembolization

MELD

model for end-stage liver disease

EAD

early allograft dysfunction

LT

liver transplant

HCC

hepatocellular carcinoma

PRBC

packed red blood cells

AST

serum aspartate aminotransferase

ALT

serum alanine aminotransferase

TBIL

total bilirubin

INR

international normalized ratio

FFP

fresh frozen plasma

Discussion

The quality of donor grafts is one of the most important factors affecting ischemia-reperfusion injury of LT. Moreover, ischemic time, cold ischemic time, and warm ischemic time are vital indices for evaluating the quality of the donor’s liver. This study showed that a prolongation of CIT elevated the AST, ALT, and bilirubin levels to 7 days after LT. Unsurprisingly, the EAD incidence was significantly higher in the CIT > 568 min group. This result agrees with the findings of a previous study regarding ischemia-reperfusion and EAD (Ito et al., 2021). In addition, more intraoperative infusion of FFP and longer respiratory support time suggested a worse recovery in CIT > 568 min group. Currently, LT is an optional treatment for HCC based on the tumor characteristics. However, research on donor-related factors and their impact on tumor recurrence is still insufficient. Numerous studies have argued that liver ischemia is a risk factor for liver cancer recurrence (Yoshimoto et al., 2012; Yang et al., 2019; Kornberg et al., 2015b).

This study demonstrated that the CIT of liver grafts is an independent risk factor for the tumor recurrence after LT in patients with HCC. Another study of 103 LT patients reported that prolonged mean CIT and WIT promoted the risk of HCC recurrence (Fisher et al., 2007). The survival rate of patients and grafts was lower in HCC recipients who received DCD allografts which suggested that IRI injury had a greater impact on HCC recipients, despite the lack of HCC recurrence data (Kornberg et al., 2015a). A large retrospective studyused the data of 9,724 patients with HCC who received an LT to conclude that prolonged donor warm ischemia time was a risk factor for HCC recurrence (Croome et al., 2013). A single-center study elaborated that AST ≥ 1,896 U/L was a risk factor for HCC recurrence after LT with DBD grafts (Duvoux et al., 2012). Moreover, donor age > 60 years and donor WIT were the risk factors for increased HCC recurrence (Orci et al., 2018). Thus, these organs are susceptible to intensified IRI, which may lead to higher rates of HCC recurrence after LT.

In our study, 30 of 80 patients, who received a liver graft from DCD donors, experienced HCC recurrence. Compared with the DBD group, the HCC recurrence rate was not statistically significant (p = 0.642; univariate Cox’s regression analysis, the reference to the data file as a Supplementary File), which might be due to the small sample size.

Although numerous studies have reported that IRI could elevate the risk of HCC recurrence (Grat et al., 2018; Orci et al., 2018; Croome et al., 2015), the mechanisms are not yet fully understood owing to the complex network of cellular and molecular interactions. Moreover, the mechanism of injury caused by different forms of IRI varies, and the alteration of the hepatic microenvironment could create a “fertile soil” for tumor cells. For instance, sinusoidal endothelial cells (SECs) are vulnerable to the IRI, particularly during cold ischemia, which may be the risk factor for endothelial cell swelling, unbalanced vasoconstriction, and neutrophil plugging (Tejima et al., 2004). In addition, IRI can break the barrier between hepatocytes and the blood, making molecules and cells spread easily. Microcirculatory disturbances lead to tissue ischemia and hypoxia. It also activates the HIF-1 α pathway, which has been recognized as a vascular endothelial growth factor associated with metastatic disease (Semenza, 2012). IRI-induced inflammatory cytokines, tumor necrosis factor- α, or interleukin-1 may promote the metastasis of different cancer cells through the expression of adhesion molecules (such as E-selectin, ICAM-1, and VCM-1), acting as mediators of tumor growth (Yoshimoto et al., 2012; Reina & Espel, 2017; Kong et al., 2018).

Damage-associated molecular patterns (DAMPs) are other important mechanisms of IRI, promoting tumor progression, which has been released by injured hepatic cells and may perpetuate the non-infectious sterile inflammatory response. The DAMPs released into the blood can cause innate immune activation and robust hepatocellular injury. For instant, high-mobility group protein B1 (HMGB1), which is the most frequently encountered DAMP in cells undergoing stress, could translocate from the nucleus to the cytosol, bind to mitochondrial DNA (mtDNA), and be released by damaged mitochondria during hypoxia (Srikrishna & Freeze, 2009). The complex of HMGB1 and mtDNA subsequently activates the TLR9 signaling pathways to promote tumor cell proliferation (Liu et al., 2015). Tohme et al. (2017) found that upregulated HMGB1 could enhance mitochondrial biogenesis in HCC cancer cells and promote tumor survival and proliferation.

Many marginal donor livers are used clinically owing to the lack of a donation pool. This issue can lead to more serious IRI damage, subsequent complications, and HCC recurrence after transplantation. Thus, evaluating liver function and repairing the liver in vitro is vital to reducing complications after transplantation. Traditional static cold storage (SCS) may be impossible. With the progress of perfusion technology and the upgrading of perfusion machines, mechanical perfusion has been applied significantly for marginal liver donation. The process is effective in refurnishing energy stores during preservation to resist impending ischemia-reperfusion injury after organ implant. In addition, ex situ organ perfusion simulates physiological blood reflow, reflects the quality of organs, and provides an objective functional evaluation for transplant surgeons. Currently, hypothermic (hypothermic oxygenated perfusion, HOPE) and normothermic (normothermic machine perfusion, NMP) and their combination are widely used machine perfusion approaches (Karangwa et al., 2016). Several clinical trials have confirmed that the HOPE technique offers promising advantages compared with SCS in patients whose organs are transplanted from DCD donors (Karangwa et al., 2016; Schlegel et al., 2019). Another group reported an encouraging result, revealing the incidence of symptomatic non-anastomotic biliary strictures in dual hypothermic oxygenated machine perfusion (D-HOPE), which was perfused through the hepatic artery and portal vein, showing that liver graft volumes in 78 patients were significantly lower after LT than in the SCS group (78 patients) after LT from DCD donors (Schlegel et al., 2013). Ischemia-free organ transplantation (IFOT) eliminates the ischemia and hypoxia of the organ by uninterrupted NMP during the procurement and implant periods. A retrospective study also demonstrated that patients who underwent the IFOT had significant higher DFS rates at 1 and 3 years than the SCS group after LT in recipients with HCC (Tang et al., 2021).

The disadvantages of ischemia-reperfusion in treating hepatocellular carcinoma have been documented. Compared with hepatectomy, LT is more vulnerable to IRI because of the sophisticated segments and procedures. However, the assessment of the patients’ oncological risk profile following LT is limited to the clinical characteristics and biological features of carcinoma. CIT is the easiest to evaluate compared with complicated biological and histological tests; although CIT is not the most direct evidence of IRI. Our results show that prolonged CIT can significantly increase the peak level of ALT, AST, TBIL, and the incidence of EAD in patients after LT.

Although several studies had found that prolonged cold ischemia time promotes recurrence after LT for HCC, their research did not cover East Asian people. Our study retrospectively analyzed 420 Chinese HCC patients with LT, demonstrated that CIT > 568 min was an independent risk factor for recurrence after LT in patients with HCC, which move forward the CIT window that warns of HCC recurrence. Therefore, the significance of CIT in patients undergoing LT for HCC should be re-evaluated.

Our study is a single-center retrospective study, which is its limitation. Future studies must conduct multicenter, prospective cohort studies and explore the biological mechanism behind IRI and tumor recurrence, which may guide the selection of grafts for HCC patients undergoing LT, and provide therapeutic targets for reducing ischemia-reperfusion and recurrence after LT.

Conclusion

In conclusion, our study confirms the harmful effect of cold ischemia time on the prognosis of liver transplantation patients with HCC. Cold ischemia time should be reconsidered when evaluating patients with HCC on the waiting list. Shortening the cold ischemia time should be one of the effective strategies to improve the prognosis, especially for those with relatively low tumor load and long expected survival.

Supplemental Information