Incidence and Impact of Acute Kidney Injury after Liver Transplantation: A Meta-Analysis

Background: The study’s aim was to summarize the incidence and impacts of post-liver transplant (LTx) acute kidney injury (AKI) on outcomes after LTx. Methods: A literature search was performed using the MEDLINE, EMBASE and Cochrane Databases from inception until December 2018 to identify studies assessing the incidence of AKI (using a standard AKI definition) in adult patients undergoing LTx. Effect estimates from the individual studies were derived and consolidated utilizing random-effect, the generic inverse variance approach of DerSimonian and Laird. The protocol for this systematic review is registered with PROSPERO (no. CRD42018100664). Results: Thirty-eight cohort studies, with a total of 13,422 LTx patients, were enrolled. Overall, the pooled estimated incidence rates of post-LTx AKI and severe AKI requiring renal replacement therapy (RRT) were 40.7% (95% CI: 35.4%–46.2%) and 7.7% (95% CI: 5.1%–11.4%), respectively. Meta-regression showed that the year of study did not significantly affect the incidence of post-LTx AKI (p = 0.81). The pooled estimated in-hospital or 30-day mortality, and 1-year mortality rates of patients with post-LTx AKI were 16.5% (95% CI: 10.8%–24.3%) and 31.1% (95% CI: 22.4%–41.5%), respectively. Post-LTx AKI and severe AKI requiring RRT were associated with significantly higher mortality with pooled ORs of 2.96 (95% CI: 2.32–3.77) and 8.15 (95%CI: 4.52–14.69), respectively. Compared to those without post-LTx AKI, recipients with post-LTx AKI had significantly increased risk of liver graft failure and chronic kidney disease with pooled ORs of 3.76 (95% CI: 1.56–9.03) and 2.35 (95% CI: 1.53–3.61), respectively. Conclusion: The overall estimated incidence rates of post-LTx AKI and severe AKI requiring RRT are 40.8% and 7.0%, respectively. There are significant associations of post-LTx AKI with increased mortality and graft failure after transplantation. Furthermore, the incidence of post-LTx AKI has remained stable over the ten years of the study.


Introduction
Acute kidney injury (AKI) is associated with high mortality worldwide (1.7 million deaths per year) [1][2][3][4].Patients who survive AKI are at increased risk for significant morbidities such as hypertension and progressive chronic kidney disease (CKD) [5].The incidence of AKI has steadily increased in recent years [2].It has been suggested that AKI's global burden is 13.3 million cases a year [6].In the United States, hospitalizations for AKI have been steeply rising, and data from national inpatient sample shows that the number of hospitalizations due to AKI increased from 953,926 in 2000 to 1,823,054 in 2006 and 3,959,560 in 2014, which accounts for one hospitalization associated with AKI every 7.5 minutes [7,8].
Thus, we performed a systematic review to summarize the incidence (using standard AKI definitions of Risk, Injury, Failure, Loss of kidney function, and End-stage kidney disease (RIFLE), Acute Kidney Injury Network (AKIN), and Kidney Disease: Improving Global Outcomes (KDIGO) classifications), risk factors, and mortality and their trends for AKI in patients undergoing LTx.

Search Strategy and Literature Review
The protocol for this systematic review was registered with PROSPERO (International Prospective Register of Systematic Reviews; no.CRD42018100664).A systematic literature search of MEDLINE (1946 to December 2018), EMBASE (1988 to December 2018) and the Cochrane Database of Systematic Reviews (database inception to December 2018) was performed to evaluate the incidence of AKI in adult patients undergoing LTx.The systematic literature review was conducted independently by two investigators (C.T. and W.C.) using the search strategy that consolidated the terms "acute kidney injury" OR "renal failure" AND "liver transplantation," which is provided in online supplementary data 1.No language limitation was implemented.A manual search for conceivably related studies using references of the included articles was also performed.This study was conducted by the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement [84] and the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) [85].

Selection Criteria
Eligible studies must be clinical trials or observational studies (cohort, case-control, or cross-sectional studies) that reported the incidence of post-LTx AKI in adult patients (age >/= 18 years old).Included studies must provide data to estimate the incidence of post-LTx AKI with 95% confidence intervals (CI).Retrieved articles were individually reviewed for eligibility by the two investigators (C.T. and W.C.).Discrepancies were addressed and solved by mutual consensus.Inclusion was not limited by the size of study.

Data Abstraction
A structured data collecting form was used to obtain the following information from each study, including title, name of the first author, year of the study, publication year, country where the study was conducted, post-LTx AKI definition, incidence of AKI post-LTx, risk factors for post-LTx AKI, and impact of post-LTx AKI on patient outcomes.

Statistical Analysis
Analyses were performed utilizing the Comprehensive Meta-Analysis 3.3 software (Biostat Inc, Englewood, NJ, USA).Adjusted point estimates from each study were consolidated by the generic inverse variance approach of DerSimonian and Laird, which designated the weight of each study based on its variance [86].Given the possibility of between-study variance, we used a random-effect model rather than a fixed-effect model.Cochran's Q test and I 2 statistic were applied to determine the between-study heterogeneity.A value of I 2 of 0%-25% represents insignificant heterogeneity, 26%-50% low heterogeneity, 51%-75% moderate heterogeneity and 76-100% high heterogeneity [87].The presence of publication bias was assessed by the Egger test [88].

Incidence of Post-LTx AKI
Overall, the pooled estimated incidence rates of post-LTx AKI and severe AKI requiring RRT following LTx were 40.7% (95% CI: 35.4%-46.2%,I 2 = 97%, Figure 2) and 7.7% (95% CI: 5.1%-11.4%,I 2 = 95%, Figure 3), respectively.Meta-regression showed no significant impact of type of donor (deceased vs living donors) (p = 0.33) on the incidence of post-LTx AKI.In addition, the year of study (p = 0.81) did not significantly affect the incidence of post-LTx AKI (Figure 4).Meta-regression showed no significant impact of type of donor (deceased vs living donors) (p = 0.33) on the incidence of post-LTx AKI.In addition, the year of study (p = 0.81) did not significantly affect the incidence of post-LTx AKI (Figure 4).Meta-regression showed no significant impact of type of donor (deceased vs living donors) (p = 0.33) on the incidence of post-LTx AKI.In addition, the year of study (p = 0.81) did not significantly affect the incidence of post-LTx AKI (Figure 4).

Evaluation for Publication Bias
Funnel plot (Supplementary Figure S1) and Egger's regression asymmetry test were performed to evaluate for publication bias in the analysis evaluating incidence of post-LTx AKI and mortality risk of post-LTx AKI.There was no significant publication bias in meta-analysis assessing the incidence of post-LTx AKI, p-value = 0.12.

Discussion
In this meta-analysis, we found that AKI and severe AKI requiring RRT after LTx are common, with an incidence of 40.8% and 7.0%, respectively.In addition, our findings showed no significant correlation between the incidence of post-LTx AKI and study year for the ten years of the study.Furthermore, compared to patients without post-LTx AKI, those with post-LTx AKI carry a 2.96-fold increased risk of mortality and a 3.76-fold higher risk of liver graft failure.
Studies have shown that higher MELD scores were associated with post-LTx AKI [23,[39][40][41][42][43][44][45][46][47][48][49].In patients with high MELD scores >30, the majority required RRT post LTx [44,97].Although SCr is an important determinant of the MELD score, other components of MELD such as pre-LTx INR has also been demonstrated to be strongly associated with post-LT AKI, suggesting that the severity of the liver disease itself, as reflected by the MELD score, is associated with post-LT AKI [45].Identified perioperative factors for post-LTx AKI include cardiopulmonary failure, vasopressor requirement, hemodynamic effects of prolonged surgery, and blood loss/RBC transfusion [18,24,25,39,48,54,65].Moreover, it has been hypothesized that HIRI is an important cause of post-LTx AKI [37,38].Aspartate aminotransferase (AST), as a surrogate marker for HIRI, has been shown to be correlated with post-LTx AKI.[38,78] HIRI has a close relationship with the systemic inflammatory response, which in turn is related to AKI and multiorgan dysfunction in similar settings such as sepsis [37].Early hepatic graft dysfunction has also been shown to be associated to post-LTx AKI [98].In addition, recipients of donation after circulatory death (DCD) grafts are reported to have a higher incidence of post-LTx AKI compared to donation after brain death (DBD grafts).After DCD LTx, peak AST levels were an independent predictor of post-LTx AKI [99].Other known factors that influence HIRI such as donor age, cold and warm ischemia times and graft steatosis have also been associated with post-LTx AKI [37].
As demonstrated in our study, post-LTx AKI is associated with an increased risk of death and liver graft failure.Several pharmacological and non-pharmacological interventions have been studied, but so far these have failed to demonstrate any significant benefit in the prevention of post-LTx AKI [37,100,101].To continue efforts to mitigate post-LTx AKI, it is important to identify those who are at high-risk for post-LTx AKI in order to develop earlier protective strategies [37].There have been many attempts to develop predictive models for post-LTx AKI [37].Seven published predictive models addressing a diverse range of AKI definitions for post-LT AKI have been developed [19,23,24,33,47,54,55].However, the numbers of patients in these studies were limited [19,23,24,33,47,54,55], and future prospective external validation, ideally with multi-center studies with large number of patients, is required.
Several limitations in our meta-analysis are worth mentioning.First, there were statistical heterogeneities present in our study.Possible sources for heterogeneities were the differences in the patient characteristics in the individual studies.However, we performed a meta-regression analysis which demonstrated that the type of donor (deceased vs. living donors); the year of study did not significantly affect the incidence of post-LTx AKI.Second, there is a lack of data from included studies on novel AKI biomarkers.Novel biomarkers for AKI are emerging and could be useful for the early identification and characterization of AKI.Thus, future studies evaluating predictive models with novel biomarkers are needed.Lastly, this is a systematic review and meta-analysis of cohort studies.Thus, it can demonstrate associations of post-LTx AKI with increased risk of mortality and liver graft failure, but not a causal relationship.

Conclusions
In conclusion, there are overall high incidence rates of post-LTx AKI and severe AKI requiring RRT of 40.8% and 7.0%.Post-LTx AKI is significantly associated with increased mortality and liver graft failure.In addition, the incidence of post-LTx AKI has remained stable over time.This study provides an epidemiological perspective to support the need for future large-scale multi-center studies to identify preventive strategies for post-LTx AKI.

Figure 1 .
Figure 1. Outline of our search methodology.

Figure 1 .
Figure 1. Outline of our search methodology.

Figure 2 .
Figure 2. Forest plots of the included studies assessing incidence rates of post-LTx AKI.A diamond data marker represents the overall rate from each included study (square data marker) and 95% confidence interval.

Figure 4 .
Figure 4. Meta-regression analyses showed no significant impact of year of study on the incidence of post-LTx AKI (p = 0.81).The solid black line represents the weighted regression line based on variance-weighted least squares.The inner and outer lines show the 95% confidence interval and prediction interval around the regression line.The circles indicate log event rates in each study.

Figure 3 .
Figure 3. Forest plots of the included studies assessing incidence rates of severe AKI requiring RRT following LTx.A diamond data marker represents the overall rate from each included study (square data marker) and 95% confidence interval.

Figure 4 .
Figure 4. Meta-regression analyses showed no significant impact of year of study on the incidence of post-LTx AKI (p = 0.81).The solid black line represents the weighted regression line based on variance-weighted least squares.The inner and outer lines show the 95% confidence interval and prediction interval around the regression line.The circles indicate log event rates in each study.

Figure 4 .
Figure 4. Meta-regression analyses showed no significant impact of year of study on the incidence of post-LTx AKI (p = 0.81).The solid black line represents the weighted regression line based on variance-weighted least squares.The inner and outer lines show the 95% confidence interval and prediction interval around the regression line.The circles indicate log event rates in each study.

Figure 5 .
Figure 5. Forest plots of the included studies assessing liver graft failure among patients with post-LTx AKI.A diamond data marker represents the overall rate from each included study (square data marker) and 95% confidence interval.

Figure 6 .
Figure 6.Forest plots of the included studies assessing CKD risk among patients with post-LTx AKI.A diamond data marker represents the overall rate from each included study (square data marker) and 95% confidence interval.

Figure 5 .
Figure 5. Forest plots of the included studies assessing liver graft failure among patients with post-LTx AKI.A diamond data marker represents the overall rate from each included study (square data marker) and 95% confidence interval.

Figure 6 .
Figure 6.Forest plots of the included studies assessing CKD risk among patients with post-LTx AKI.A diamond data marker represents the overall rate from each included study (square data marker) and 95% confidence interval.
Abbreviations: AKIN, Acute Kidney Injury Network; DCD, donation after circulatory death; EDC, extended donor criteria liver allografts; KDIGO, Kidney Disease Improving Global Outcomes; RIFLE, Risk, Injury, Failure, Loss of kidney function, and End-stage kidney disease; UK, United Kingdom; USA, United States of America.

No Graft Failure Graft Failure Study name Statistics for each study Odds ratio and 95% CI
Figure 6.Forest plots of the included studies assessing CKD risk among patients with post-LTx AKI.A diamond data marker represents the overall rate from each included study (square data marker) and 95% confidence interval.

Table 2 .
Reported Potential Predictors/Associated-Risk Factors of Post-LTx AKI.