Association of post‐reperfusion syndrome and ischemia‐reperfusion injury with acute kidney injury after liver transplantation

Acute kidney injury (AKI) is frequently observed after orthotopic liver transplantation (OLT) even in patients with previously normal renal function. In this study, we investigated the impact of factors such graft steatosis, post‐reperfusion syndrome (PRS), and hepatic ischemia reperfusion injury (HIRI) on the development of AKI after OLT in adult patients.

factors of AKI. Furthermore, the recent increase in the use of extended criteria donor grafts has led to more frequent development of AKI after OLT. 7 Therefore, the characteristics and mechanisms of AKI in the era of accepting extended criteria donor organs need to be analyzed using universal criteria. In the present study, AKI was diagnosed and classified based on the AKI criteria proposed by the International Kidney Disease Improving Global Outcomes ( KDIGO) workgroup. 8 Several studies have shown that extended criteria donor grafts lead to AKI through hepatic ischemia-reperfusion injury (HIRI) 7,9 and/or post-reperfusion syndrome (PRS), 10,11 which can be related to level of steatosis in the liver graft. 12,13 A large registry study revealed the detrimental effects of macrovesicular steatosis of liver grafts by showing that higher steatosis was independently associated with impaired graft survival. 14 Therefore, the association among graft steatosis, PRS, HIRI, and AKI needs to be investigated to determine the pathophysiological mechanism of AKI after OLT. Furthermore, in order to identify modifiable risk factors and to implement preventive strategies, the impact of intraoperative parameters such as low blood pressure and/or the use of vasopressors on the development of HIRI, PRS, and AKI need to be thoroughly investigated.
To determine the association of graft steatosis, PRS, HIRI, and AKI, we evaluated the effects of graft steatosis on PRS and HIRI together with recipient and donor variables. We also identified independent perioperative predictive factors for AKI development.

| Patient selection
We retrospectively identified 496 consecutive patients who underwent OLT using grafts from donation after brain death at our institution between July 2011 and June 2017. Of these, 110 patients were excluded from this study for the following reasons: recipient age <18 years (n = 58), re-transplantation (n = 33), urgent OLT for acute liver failure (n = 15), combined liver-kidney transplantation (n = 3), and early patient death within 48 hours (n = 1). Finally, 386 patients were included in the study ( Figure 1). Clinical data were collected retrospectively from the registration database as well as from clinical records and stored anonymously. The Model for End-Stage Liver Disease (MELD) score was calculated using the data analyzed immediately before OLT. Cold ischemia time (CIT) was defined as the time from liver cold perfusion in the donor until removal from ice for implantation. Warm ischemia time (WIT) was defined as the time period from removal from ice until portal vein reperfusion. Intraoperative needle biopsy of the liver graft was routinely performed between the graft reperfusion and closing of the abdominal incision. After formalin fixation, the biopsied specimens were stained using the periodic acid-Schiff (PAS) staining technique. The degree of macrovesicular steatosis, which was evaluated for all of the biopsied specimens by two pathologists in our institution, was semiquantified and defined as the percentage of hepatocytes in which fat droplets accumulation resulted in displacement of the nucleus. The data on the percentage of macrovesicular steatosis were retrospectively collected from the histological reports. This study was conducted in accordance with the Helsinki Declaration and was approved by the local institutional internal review board. Informed consent from the patients in this study was waived by the institutional internal review board according to Swedish legislation.

| Definition of PRS, HIRI, and AKI
PRS was defined as at least a 30% decrease in mean arterial pressure occurring during the first 5 minutes after liver graft reperfusion and lasting longer than 1 minute. 11 Peak aspartate aminotransferase (AST) level within 72 hours post-OLT was used as a surrogate marker of HIRI. 2,15 HIRI was defined when AST was ≥2500 U/L, moderate HIRI was defined when AST was 2500-5000 U/L, and severe HIRI was defined when AST was >5000 U/L.

Editorial Comment
Acute kidney injury is a frequent complication after liver transplantation, and this retrospective analysis of a singlecenter cohort study evaluated the risk factors for this complication. Hepatic ischemia-reperfusion injury and cold ischemia time for the donor organ were associated with an increased risk for post-operative acute kidney injury. F I G U R E 1 Flow chart of patient enrolment AKI after OLT was defined based on the AKI criteria proposed by the KDIGO workgroup, and Stage 1 was defined as 1.5-1.9 times the baseline serum creatinine (SCr) level in 7 days or a ≥26 µmol/L increase in 48 hours when compared with the baseline level, Stage 2 was defined as 2.0-2.9 times baseline in 7 days, and Stage 3 was defined as ≥3.0 times baseline or an increase to ≥354 µmol/L or the initiation of renal replacement therapy. 8 The baseline SCr was defined as the SCr level immediately before LT, which was routinely evaluated in our institution.

| Immunosuppressive regimens
All patients received induction immunosuppressive treatment consisting of basiliximab perioperative and on day 4 after LT.
Solumedrol was started at 1 g on the day of the operation with subsequent tapering to 5-10 mg at 2 months and 2.5-5 mg at 1 year after transplantation. The detrimental effect of calcineurin inhibitors, especially tacrolimus, on kidney function has been well documented. In our institution, we have implemented a renal-sparing immunosuppressive regimen that delays perioperative tacrolimus use; tacrolimus was initiated on post-operative day 4 and adjusted to maintain the initial trough level of 6-10 ng/mL.
Mycophenolate mofetil was started on post-operative day 1 with the standard dose of 1 g twice daily.

| Statistical analysis
Data are expressed as means (± standard deviations, SD) or medians (interquartile range [IQR]) as appropriate. Significant differences were determined using one-factor analysis of variance for normally distributed data, Wilcoxon's signed-rank tests for skewed data, and Fisher's exact tests or chi-squared tests for dichotomous data. Univariate and multivariable logistic regression analyses were used to assess the association of several parameters with the incidence of HIRI and AKI. Analyses were performed using JMP Pro 14 (SAS Institute). A P-value less than .05 was considered significant.

| Pre-transplant patient demographics
The median recipient age was 54 years with preoperative median

| Intraoperative Factors
The graft and perioperative factors are listed in Table 2. In the stage 3 AKI group, the graft-to-recipient weight ratio (GRWR) was largest, the rate of macrovesicular steatosis ≥30% was greatest, CIT, WIT, and operative time were longest, and bleeding volume was largest. Regarding factors during the post-reperfusion period, systolic blood pressure was lowest, and noradrenalin dose and lactate level at the end of operation were highest in the stage 3 AKI group. Peak AST levels within 72 hours after OLT, a surrogate marker of HIRI, increased as AKI post-OLT became more severe. Figure 2 demonstrates the correlation of steatosis, HIRI, and AKI.

| Impact of AKI development on 1-year outcomes
We analyzed patient survival rate at 1-year post-LT in AKI stage 2 and 3 patients. At 1-year post-LT, the patient survival rate was lower in AKI stage 2 and 3 patients than those with AKI stage <2 (91.7% vs 94.2%), although statistically insignificant (P = .51).

| Multivariable analysis of risk factors for HIRI development
Univariate analysis for the development of HIRI, which was de-  Table 3). Among these 12 variables, multivariable analysis showed that graft steatosis, CIT, WIT, and lactate level at the end of the operation were independent risk factors for the development of HIRI.

| Multivariable analysis of risk factors for AKI development
Univariate analysis showed that three recipient variables (MELD score, SCr level, and pre-transplant non-insulin-dependent diabetes), two donor variables (male donor and BMI), two graft characteristics (graft steatosis and CIT), four surgical variables (WIT, operative time, bleeding volume, and blood transfusion), and three post-reperfusion characteristics (total noradrenalin dose, lactate level at the end of operation, and HIRI) were risk factors for AKI (Table 4). Among these variables, multivariable analysis identified two independent predictive factors for the development of AKI, namely CIT and HIRI.

| D ISCUSS I ON
The main results of the present study were (a) 37% of OLT recipients transplanted in our center, developed AKI, which was diagnosed using the revised criteria of the KDIGO workgroup, and half of them   Note: Data are given as mean (SD), median (interquartile range), and number (%).
Abbreviations: AKI, acute kidney injury; AST, aspartate aminotransferase; CIT, cold ischemia time; GRWR, graft-to-recipient weight ratio; HR, heart rate; NA, noradrenalin; OLT, orthotopic liver transplantation; SBP, systolic blood pressure; WIT, warm ischemia time. especially in the era when marginal donor grafts are increasingly accepted, the deleterious impact of extended criteria donor graft has been demonstrated in several studies. [1][2][3]7 In the present study, graft steatosis was significantly associated with the development of HIRI, which was then identified as an independent prognostic factor of AKI development. In contrast, multivariable analysis revealed that graft steatosis itself was not an independent predictor of AKI, which suggests that steatotic livers did not directly cause AKI but had indirect effects through the development of HIRI, which is linked with systemic inflammatory response. This seems to be an important finding of this study to elucidate the pathophysiological mechanism of post-OLT AKI.

F I G U R E 2 Correlation (A) between graft steatosis and hepatic ischemia-reperfusion injury (HIRI), (B) between HIRI and acute kidney injury (AKI), and (C) between graft steatosis and AKI
In the present study, even the mildest degree of steatosis (steatosis of 10%-20%) had a significant impact on HIRI development compared with steatosis of <10%. As the degree of steatosis became more severe, the association with HIRI development became stronger.
These results confirmed that even at its mildest degree, graft steatosis had effects on liver graft and patient outcomes, although most of those patients recovered from HIRI. Similar associations between liver steatosis and post-OLT HIRI have been seen in rodent models. 13 16 In our study, longer CIT was an independent risk factor for both HIRI and AKI, suggesting that the use of steatotic liver grafts needs to be cautiously decided when it is predicted that CIT will be long. Several studies have shown that liver grafts with low or moderate macrovesicular steatosis can be safely used in recipients without additional risks. [17][18][19] Considering the mortality rate of patients on the waiting list, accurate estimation of risks for AKI and further strategies for preserving kidney functions seem to be important to avoid wasting liver grafts. In this study, even mild steatosis was identified as a risk factor of HIRI, which is inconsistent with previous studies. 18,19 This inconsistency seems to be caused by the low prevalence of moderate or severe steatosis among our cohort. Regarding severe macrovesicular steatosis (>60%), many studies have demonstrated high primary non-function rate and impaired graft survival. [19][20][21][22] However, McCormack et al reported that 60-day mortality and 3-year patient survival rate were comparable between the severe steatosis group and matched control group in well-controlled cases, although the proportion of patients with longterm ICU and hospital stay was significantly higher for patients who received severely steatotic grafts. 12 Out of 386 cases in this study, Although several studies have reported that pre-OLT SCr level is an important predictor of post-OLT AKI development, 24 the impact of pre-OLT SCr level on AKI development is still controversial. 2,9,25,26  is not specific to the liver, but can be released by other organs such as the heart or muscles. However, at the present, peak AST level post-OLT has been commonly used as a surrogate marker of HIRI and time frame of 72 hours post-OLT was consistent with previous reports. 2,15 In summary, HIRI induced by steatotic liver graft and long CIT was associated with post-OLT AKI development rather than pre-OLT recipient demographics or surgical factors. Although we should not discard mildly or moderately steatotic livers only for that reason, cautious evaluation of expected outcomes and preventive strategy to suppress post-perfusion inflammatory response would be of importance. To prevent AKI development in the era of enhancing donor pools, further strategies of suppressing HIRI, such as machine perfusion for steatotic liver grafts, need to be further examined and implemented.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.