Objectives: Acute kidney injury is a frequent complication of liver transplant. Here, we assessed the rate and contributing factors of acute kidney injury and need for renal replacement therapy in patients undergoing liver transplant at a transplant center in Tehran, Iran.
Material and Methods: We identified all patients who underwent liver transplant at the Imam Khomeini Hospital Complex from March 2018 to March 2019 and who were followed for 3 months after transplant. Acute kidney injury was defined based on the
Acute Kidney Injury Network criteria. We collected demographic and pretransplant, intraoperative, and posttransplant data. Univariable and multivariable models were applied to explore independent risk factors for acute kidney injury incidence and need for renal replacement therapy.
Results: Our study included 173 deceased donor liver transplant recipients. Rates of incidence of acute kidney injury and need for renal replacement therapy were 68.2% and 14.5%, respectively. The 3-month mortality rate among those with severe and mild or moderate acute kidney injury was 44.0% (14/25) and 9.7% (9/ 93), respectively (P < .001). Multivariable analyses indicated that serum albumin (relative risk of 0.55; 95% confidence interval, 0.34-0.87; P = .021), baseline serum creatinine (relative risk of 2.11; 95% confidence interval, 1.56-2.90; P = .037), and intraoperative mean arterial pressure (relative risk of 0.76; 95% confidence interval, 0.63-0.82; P = .008) were independent factors for predicting posttransplant acute kidney injury. Independent risk factors for requiring renal replacement therapy were pretransplant serum creatinine (relative risk of 1.99; 95% confidence interval, 1.89-4.47; P = .044) and intraoperative vasopressor infusion (relative risk of 1.41; 95% confidence interval, 1.38-2.00; P = .021).
Conclusions: We found a high incidence of acute kidney injury among liver transplant recipients in our center. There was a significant association between severity of acute kidney injury and 3-month and in-hospital mortality.
Key words : Acute Kidney Injury Network, Posttransplant outcomes, Liver transplantation, Renal replacement therapy
Liver transplant (LTx) is the most effective treatment for patients with advanced hepatic failure,1 allowing patients significantly increased overall survival and improved quality of life.2 So far, LTx has been established successfully in Iran, resulting in 4485 transplant procedures over a period of 23 years at 6 centers with an overall 10-year survival rate of 71% for adults.3
Acute kidney injury (AKI) is a frequent complication of LTx with a range of incidence that widely varies from 5% to 94%. About 11% to 17% of these patients present with severe AKI requiring renal replacement therapy (RRT) early after transplant.4 This variation in incidence rates might be due to the use of different diagnostic criteria for AKI, which include the Risk, Injury, Failure, Loss of Kidney Function, and End-Stage Kidney Disease (RIFLE), the Acute Kidney Injury Network (AKIN), and the Kidney Disease: Improving Global Outcomes (KDIGO) classifications.5,6 Many factors may contri-bute to AKI development after LTx, including renal ischemia from surgery-related events (ie, blood loss, hypotension), use of nephrotoxic medications, hepatorenal syndrome, infections, and sepsis.7,8
Acute kidney injury should not be considered as a transient presentation. Rather, it is a complication with long-lasting adverse effects, such as the presence of worsening graft function and increasing risk of long-term renal dysfunction and its associated morbidity and mortality, especially among those with severe AKI.9,10 Moreover, some studies have mentioned that patients with AKI might have longer admission stays in the intensive care unit (ICU) and hospital with a considerable impact on total costs.11,12
In addition to intermittent hemodialysis (IHD), continuous renal replacement therapy (CRRT) has been recently established for the treatment of AKI in our center. The CRRT as a slow and continuous technique can be tolerated in patients with hemodynamic instability better than IHD. In addition to facilitating fluid and drug management and providing better control for electrolyte and acid-base disorder, it has been postulated that CRRT could remove inflammatory cytokines and humoral mediators during the course of sepsis.13
Although different studies have assessed AKI after LTx, the incidence rate and pretransplant and intraoperative risk factors are not yet clear, especially among Iranian patients. In this retrospective observational study, we assessed the incidence rate of AKI in the early posttransplant period, its contributing factors, and the need for RRT among patients undergoing LTx in one of the referral centers for transplant in Iran (the Imam Khomeini Hospital Complex [IKHC], Tehran, Iran). We also evaluated the impact of AKI on hospital course and 3-month survival of patients. These findings could help us improve methods for early prevention and intervention to ameliorate kidney injury and could potentially have implications for patient outcomes, preserving financial resources and allowing better clinical management.
Materials and Methods
This study was approved by the Ethics Committee of Tehran University of Medical Science (approval number: IR.TUMS.IKHC.REC.1398.120).
Patients and follow?up
From March 2018 to March 2019, we identified all patients who underwent LTx for treatment of advanced hepatic failure at the IKHC and retrospectively reviewed their medical records. We excluded patients if they met the following criteria: (1) age ?18 or ?75 years, (2) mortality within 24 hours after LTx, (3) combined kidney and liver transplant, (4) incomplete clinical and follow-up data, and (5) RRT before LTx.
During the study period, 4 patients had undergone LTx with organs from living related donors. Because all 4 of these patients were younger than 18 years, they were excluded from the study. Thus, all remaining 173 patients included in the analyses were LTx recipients of deceased donors.
Follow-up was considered as the time after transplant to death or the time of hospital discharge, which included visits during the first 3 months posttransplant at our LTx clinic. We collected demographic information, comorbidities, drug history, Model for End-Stage Liver Disease (MELD) score and Child-Pugh grade, cause of end-stage liver disease, and pretransplant laboratory values. Intraoperative factors were also analyzed, including hemodynamic parameters, use of diuretics and vasopressor infusion, blood product transfusion, and anesthesia duration. Posttransplant data included length of stay in the ICU and any specialized ward, mortality rates, laboratory values, urine output volume, and use of antibiotics, diuretics, and vasopressor agents (as a surrogate of possible ischemic injury. Need for retransplant and data related to RRT among those with severe AKI were also recorded for analyses.
In our center, LTx is performed according to the guideline of the ethics committee of our hospital and the Declaration of Helsinki. All patients underwent routine surgical techniques and were treated with the immunosuppression protocol of our center, which has been previously described.14 We provided standardized posttransplant care for all patients, with hemodynamic status monitored in a specialized ICU for hepatic surgery.
Definition of outcome
Acute kidney injury was defined based on the AKIN criteria,15 which have a high sensitivity for its diagnosis in LTx patients.16 We assessed AKI within the first 7 days after LTx. This time frame was selected to evaluate pretransplant and intraoperative factors more accurately and to exclude factors that could affect kidney function chronically, such as long-term use of immunosuppressive agents.
We categorized AKI into 3 stages, based on its severity: stage I patients had increase in serum creatinine (SCr) of ?0.3 mg/dL or ?1.5 to 1.9 times the baseline or urine output of <0.5 mL/kg/h for 6 to 12 hours; stage II patients had SCr increase of >2.0 to 2.9 times the baseline or urine output of <0.5 mL/kg/h for ?12 hours; and stage III patients had SCr increase of ?3.0 times the baseline or increase in SCr to ?4.0 mg/dL or urine output of <0.3 mL/kg/h for ?24 hours or anuria for ?12 hours or need for initiation of RRT (defined as severe AKI). The last laboratory test of SCr before LTx was considered as the baseline level.
Baseline characteristics of participants were presented based on having AKI and need for RRT among those with AKI. We expressed continuous variables as means ± SD or medians with interquartile ranges (IQR). We used t test analysis to show differences in continuous variables if normally distributed and Mann-Whitney rank sum test for variables with skewed distribution. Categorical data are shown as numbers and percentages. Pearson chi-square analyses or Fisher exact tests were used to compare differences in categorical variables, where appropriate. Kaplan-Meier curves were used to compare the in-hospital and 3-month survival rates between patients who developed AKI versus those who did not, as well as patients who received RRT versus those who did not
We used a modified Poisson regression model17,18 with backward selection (alpha = 0.2) to explore possible independent risk factors of AKI and reported adjusted relative risks (RR) with 95% confidence intervals (CI). Based on Akaike information criterion, an alpha of 0.157 was appropriate. We applied 3 clinical predictive risk models. Model 1 was adjusted for age and body mass index (BMI, calculated as weight in kilograms divided by height in meters squared). Model 2 was further adjusted for pretransplant variables (albumin and SCr). Finally, model 3 was further adjusted for intraoperative variables (mean arterial pressure [MAP] and vasopressor infusion). Similarly, the univariable and multivariable linear regression analyses were applied to explore independent risk factors of need for RRT among LTx recipients who were diagnosed with AKI for the variables with unadjusted P < .2. The multivariable model was adjusted for pretransplant albumin SCr levels, intraoperative MAP and vasopressor infusion, and retransplant status. We used SPSS version 21 for Windows for all analyses. P < .05 was considered statistically significant.
Our study included 173 patients. All were LTx recipients of organs from deceased donors. Table 1 shows the baseline characteristics and comparisons of quantitative and categorical variables in patients with and without AKI.
Incidence of acute kidney injury
According to the AKIN criteria, the incidence of AKI in the posttransplant period was 68.2% (118 of 173 patients). Most patients (53.2%) developed AKI on the first day after LTx, 22.2% on the second day, 21.4% on the third day, and 3.2% developed AKI more than 72 hours after LTx. Moreover, the lowest urine output and highest mean levels of SCr among our patients were on the first and third days after LTx, respectively (Figure 1). Stage I, stage II, and stage III of AKI episodes were observed in 58 (49.2%), 35 (29.7%), and 25 recipients (21.2%), respectively.
Comparison between liver transplant recipients with and without acute kidney injury
Compared with LTx recipients with no AKI events, recipients who developed AKI were about 8.4 years older (P = .045), had about 4.8 higher BMI (P = .033), had 0.9 mg/dL lower serum level of albumin (P < .001), had 0.6 mg/dL higher SCr concentration (P = .001), and had 10.7 mm Hg lower MAP (P < .001). Moreover, intraoperative vasopressor infusion was more prevalent among the AKI group (P < .001). Other characteristics were similar in patients with and without AKI (Table 1).
Comparison between recipients with severe versus mild or moderate acute kidney injury
Among 118 recipients with AKI, 21.2% (25 of 118) experienced severe AKI and required RRT. Of these patients, 44.0% (11 of 25) underwent CRRT, and IHD was prescribed for the remaining 14 cases (56%). Baseline characteristics of patients with AKI according to need for RRT are shown in Table 1. Compared with patients with AKI and no need for RRT (mild and moderate AKI), participants who needed RRT (severe AKI) had about 0.7 mg/dL lower serum level of preoperative albumin (P = .016) and about 0.9 mg/dL higher preoperative SCr concentration (P = .025). Furthermore, compared with those with mild or moderate AKI, the intraoperative vasopressor infusion was 20% more prevalent among those with severe AKI (group who underwent RRT).
Three-month and in-hospital mortality and length of stay in intensive care unit
In our study, the overall mortality rate in the first 3 months after LTx was 19.7% (34 of 173). The 3-month mortality was not significantly different between those with and without AKI (19.5% vs 20.0% respectively; P = .319). However, the 3-month mortality rate was 44.0% (14 of 25) among patients with severe AKI and 9.7% (9 of 93) among those with mild or moderate AKI, with a significant association between AKI severity and mortality (RR of 4.53; 95% CI, 3.60-7.13; P < .001) (Table 2). In more detail, 47.3% of patients who underwent IHD and 66.6% of patients who underwent CRRT died within 3 months after LTx. Similar to this pattern, recipients with and without AKI had no significant difference with regard to in-hospital mortality (P = .207), whereas recipients who required RRT (severe AKI) had a significantly higher in-hospital mortality compared with those with mild or moderate AKI (36.0% vs. 7.5%; RR of 4.87; 95% CI, 3.24-7.56; P = .002) (Table 2 and Figure 2). Furthermore, patients with incidence of AKI had significantly longer ICU stay compared with patients without AKI (6.0 ± 1.9 vs 3.0 ± 2.0 days; P <0.001). However, there was no significant differences among patients with AKI who underwent RRT and those who did not (6.6 ± 2.9 vs 5.4 ± 1.6 days; P = .506) (Table 2).
Potential risk factors
Univariable and multivariable RRs and 95% CIs of potential risk factors for incidence of AKI among LTx recipients are shown in Table 3. Adjustment for age, BMI, pretransplant values of SCr and albumin, and intraoperative variables (MAP and vasopressor infusion) in multivariable model 3 (full model) indicated that the baseline serum albumin (RR of 0.55; 95% CI, 0.34-0.87; P = .021), baseline SCr (RR of 2.11; 95% CI, 1.56-2.90; P = .037), and intraoperative MAP (RR of 0.76; 95% CI, 0.63-0.82; P = .008) were independent factors to predict incidence of AKI after LTx. Table 4 shows the univariable and multivariable RR and 95% CI results of potential risk factors for need for RRT among LTx recipients with AKI. Independent risk factors in the multivariable analysis for requiring RRT included pretransplant SCr (RR of 1.99; 95% CI, 1.89-4.47; P = .044) and intraoperative vasopressor infusion (RR of 1.41; 95% CI, 1.38-2.00; P = .021).
In this retrospective cohort study, rates of incidence of AKI and RRT requirement within 7 days after LTx were 68.2% and 14.5%, respectively. The 3-month survival rates among patients with mild or moderate AKI and severe AKI were 90% and 56%, respectively. In addition, incidence of AKI could significantly increase the length of ICU stay. Multivariable regression analyses revealed that serum albumin, pretransplant SCr, and MAP during the operation were notable risk factors for AKI incidence. Moreover, pretransplant renal function and intraoperative vasopressor infusion could contribute to requirement for RRT among recipients with AKI.
Several studies have reported AKI as a common complication of LTx. In 2019, a meta-analysis estimated an incidence rate of 40.7% for AKI after LTx and a rate of 7.7% for severe AKI needing RRT.4 More recent studies conducted in Sweden,19 Belgium,20 China,21 Germany,22 United States,23 and Turkey24 revealed that 37% (141/386), 57.1% (117/205), 50% (66/132), 9.3% (14/149), 19.6% (100/511), and 30.5% (35/177) of LTx recipients developed AKI, respectively. Therefore, compared with most previous studies, our incidence rate was significantly higher. This difference could be explained by the variation in AKI definitions,25 baseline characteristics, surgical technique, and disparities in pre- and posttransplant management.
For example, based on KDIGO criteria, the Belgian study assessed AKI among 205 participants and excluded those with a preoperative SCr value?>1.5 mg/dL. Moreover, in the Chinese study, 132 participants with less severity of liver disease (MELD score=11.9) were analyzed, and those who died within 48 hours after LTx and those with documented glomerular filtration rate <60 mL/min/1.73 m2 for 3 months (with or without dialysis requirement) were excluded. Furthermore, the Swedish study consisted of 386 participants who were older than patients in our study (59.3 vs 44.6 years) and who had greater prevalence of hepatitis C at baseline (26.7% vs <15.8%). The Swedish study also excluded patients with retransplant and acute liver failure and those who died within 48 hours after LTx, which was inconsistent with our study.
Our study indicated that the level of SCr plays the most important role in the incidence of mild/moderate AKI and severe AKI, which is in line with existing literature.7,26,27 Haan and colleagues stated that impairment in preoperative renal function, either as chronic kidney disease or AKI, could be a risk factor for post-LTx AKI.26 Interestingly it has been suggested that those with serious preoperative renal dysfunction who require RRT have an increased risk of severe complications.28 On other hand, other studies29-31 did not show this effect, probably because of limited sample size and exclusion of patients with pretransplant renal failure with high MELD scores and high SCr levels.
Our analysis of intraoperative factors indicated that MAP during surgery has a strong association with AKI development, similar to that shown previously.32 In previous studies, ischemic acute tubular necrosis and prerenal azotemia were the most probable pathogenic mechanisms of AKI in the early post-LTx period.33,34 During liver mobilization, patients may experience hypotension as a result of the prolonged transplant procedure. In addition, instability in circulation and massive blood loss might contribute to a reduction in the return of blood flow to the heart. Thus, renal vein congestion may happen and lead to lower renal perfusion and variable degrees of kidney injury in this phase.35
Our results imply that patients with higher use of vasopressor infusion, as a surrogate for serious hemodynamic instability during the surgery, have a higher risk of requiring post-LTx RRT. The existing literature has shown inconclusive findings on the association of intraoperative vasopressor infusion and the development of post-LTx AKI. Zongyi and colleagues36 demonstrated that patients with more need for vasoactive agents had the inevitable condition during operation (eg, more blood loss) and may have lower MAP, which could potentially cause renal hypoperfusion and consequently higher probability of AKI episodes. Similar results were reported in several other studies.37,38 In contrast, Wyssusek and Xu and colleagues considered intraoperative noradrenaline as a protective factor for incidence of AKI. They assumed that noradrenaline use may prevent more sustained hypotension and has a protective role in restoring blood pressure among hypotensive patients with vasodilation and could mask the intraoperative hypotension effects.31,39
Similar to our findings, previous studies provided evidence that pretransplant hypoalbuminemia is a significant independent predictor that contributes to the progression of post-LTx AKI.40,41 There are several explanations for this phenomenon. First, hypoalbuminemia could reduce starling forces, which may lead to a decrease in glomerular filtration rate. Second, it has been suggested that serum albumin may change the pharmacokinetics of nephrotoxic drugs. Third, albumin can reinforce renal perfusion, protect renal tubular cells from apoptosis, and contribute to their proliferation.42 Finally, besides its protective effect on the kidney, low serum albumin levels could indicate an overall sickness or baseline renal dysfunction.
Previous studies have mentioned some other factors, including male sex, preoperative history of diabetes mellitus, hypertension, retransplant, abnormalities in liver function tests, intraoperative blood product transfusion (including packed red blood cells and fresh frozen plasma),43 higher MELD score, longer anhepatic phase, and occurrence of intraoperative acidosis44 as probable risk factors for posttransplant AKI, which are in contrast with our findings. Furthermore, similar to Zongyi and associates, our results revealed no evidence to support the association between higher MELD score or Child-Pugh grade and post-LTx AKI.36 This may be due to our smaller sample size or the demographic differences between study populations. Although our AKI incidence rates were different from those previously shown, the risk factors of post-LTx AKI were similar.
In our study, about 80% of the AKI episodes were mild or moderate and recovered spontaneously without need for RRT. Likewise, it has been reported that most cases of mild/moderate post-LTx AKI are reversible.45 Although AKI may influence the length of ICU stay, our analysis indicated no significant impact of mild or moderate AKI on 3-month and in-hospital survival of patients (Table 2). Furthermore, patients with severe AKI, who usually have significant renal injury, had nearly 4 times higher risk of in-hospital and 3-month mortality compared with those with mild or moderate AKI. These findings are similar to those shown by Hilmi and Wyssusek and colleagues.30,31 In contrast, some studies have demonstrated that AKI was potentially associated with longer ICU stay, higher 90-day mortality,28 30-day mortality,46 higher in-hospital mortality,47 and long-term renal function and graft survival48 regardless of severity of AKI. Two other studies emphasized that even mild AKI was associated with reduced graft and patient survival rates.25,26
Our findings provided data from one of the developed LTx centers in Iran with a sufficient sample size. Because level of SCr may be influenced by reduced muscle mass and malnutrition, we also considered urine output for better detection of renal function. It is important to note that SCr level is less reliable in patients with liver disease for estimation of actual renal function. Moreover, IKHC is one the referral centers in Iran with high heterogeneity in demographic characteristics of patients from different regions of Iran; therefore, our results could be generalized to the Iranian nation.
This study has some limitations. We could not assess all of the potential risk factors associated with AKI, such as immunosuppressive regimens and fluid therapy, donor factors such as the graft quality, tumor stage (in patients with liver cancer), and socioeconomic status of the patients. Further studies should concentrate on these factors to investigate other factors associated with worse outcomes in patients with post-LTx AKI.
Our study showed an alarmingly high incidence of AKI among LTx recipients at one of the referral centers for LTx in Iran. Lower serum albumin level, pretransplant renal dysfunction, and intraoperative hemodynamic instability were associated with AKI development. Moreover, pretransplant SCr level and use of intraoperative vasopressor infusion were predictors of severe AKI and could contribute to need for RRT among recipients with AKI. Finally, we found that severe AKI with the need for RRT significantly reduced the in-hospital and short-term survival of patients. Our analysis has certain clinical implications; for preventing kidney injury after LTx, more attention should be paid to improve pretransplant and intraoperative management of recipients.
DOI : 10.6002/ect.2021.0300
From the 1Nephrology Research Center, Tehran University of Medical Sciences, Tehran, Iran; the 2Prevention of Metabolic Disorders Research Center, Research Institute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences, Tehran, Iran; the 3Department of Epidemiology and Biostatistics, School of Public Health, Tehran University of Medical Sciences, Tehran, Iran; and the 4Liver Transplantation Research Center, Tehran University of Medical Sciences, Tehran, Iran; and the 5Department of General Surgery, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest. This article is part of a thesis done by Mrs. Reyhane Hizomi Arani for MD-MPH graduate degree and was supported by Tehran University of Medical Sciences, Tehran, Iran (thesis number: 9411215200). The authors express their appreciation to the study participants and medical staff of the liver transplant ward of Imam Khomeini Hospital Complex.
Corresponding author: Mohammad Hossein Shojamoradi, Nephrology Research Center, Imam Khomeini Hospital Complex, Tehran University of Medical Sciences, Keshavarz Blvd., Tehran, Iran 1419733141
Figure 1. Mean Levels of Urine and Serum Creatinine Before and After Liver Transplant
Table 1. Baseline Characteristics of 173 Liver Transplant Recipients
Table 2. Lengths of Intensive Care Unit Stay, In-Hospital Mortality, and 3-Month Mortality Based on Incident of Acute Kidney Injury and need for Renal Replacement Therapy
Figure 2. Kaplan-Meier Curves
Table 3. Univariable and Multivariable Regression Models for Incidence of Acute Kidney Injury Among Liver Transplant Recipients
Table 4. Univariable and Multivariable Regression Models for Need for Renal Replacement Therapy Among 118 Patients With Acute Kidney Injury