Acute Kidney Injury Prediction Model After Liver Transplantation Using Grafts From Donors After Cardiac Death

doi:10.21203/rs.3.rs-1442996/v1

Download PDF

Research Article

Acute Kidney Injury Prediction Model After Liver Transplantation Using Grafts From Donors After Cardiac Death

https://doi.org/10.21203/rs.3.rs-1442996/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background and Aims: Using grafts from donors after cardiac death (DCD) influence the risk of acute kidney injury (AKI) after liver transplantation (LT). The goal of this study is to develop a novel prediction model that quantifies the impact of each risk factor on AKI after transplantation using DCD grafts.

Methods: Total of 132 patients undergoing LT using DCD grafts were evaluated retrospectively to develop a prediction model using the Cox proportional hazards regression model. The independent validation cohort included 112 patients recruited prospectively in the same institution.

Results: Overall, 103 (42.2%) of the recipients developed AKI. A prediction score model included five risk factors leading to a range of -2 to 8 score points was establishment. The predicted probability of AKI by the Model for End-Stage Liver Disease (MELD) score≥15 alone was 43.9% rising to 85.3% when combined with cold ischemia time (CIT) ≥ 7 hours. Excessive red blood cell (RBC) transfusions during operation also attribute to AKI. Surprisingly, a mild elevate AST pre-transplant was negative with the AKI, while mechanical ventilation ≥ 40 hours was associate with AKI after LT. Three risk classes were defined based on the risk of AKI: high (30%), medium (11-29%), and low (10%). Model fits were adequate in both derivation (P=0.3185) and validation cohorts (P=0.3247).

Conclusion: MELD score and CIT were the key determinants of post-LT AKI using DCD grafts. Improved organ preservation techniques and other strategies reducing the IRI may lower the risk of AKI.

acute kidney injury

liver transplantation

graft

donors after cardiac death

risk factor

ischemia-reperfusion injury

Model for End-stage Liver Disease scores

cold ischemia time

RBC transfusion

mechanical ventilation

Acute kidney injury (AKI) is the most prevalent complication following liver transplantation (LT), with a prevalence ranging from 25–70% depending on the chosen definition of AKI [1–3]. Post-transplant AKI is associated with the increase of morbidities and mortality of liver recipients [4,5]. Although various risk factors have been identified and preventive measures have been proposed for avoiding nephrotoxic medications and hyperglycemia, as well as optimizing hemodynamics during surgery, but the incidence of perioperative acute kidney injury has remained unchanged [1,6–8,9]. It maybe duo to the significant changes of liver transplantation, such as the expansion of donation criteria and the switch of transplant allocation strategy to MELD score-based system.

The use of organs from donation after cardiac death (DCD) has become an established routine but increases the risk of AKI [10]. This risk is further aggravated by ischemia reperfusion injury (IRI) during the process of transplantation [11]. Organ preservation technology has advanced that can be applied clinically in DCD liver grafts to reduce IRI [12,13]. The necessary of demanding a preservation technology to ameliorate IRI is need to be evaluated, and also need for assessment of other risk factors prior to transplantation. Therefore, our goal is to establish a model to predict the risk of AKI in patients receiving DCD donor liver transplantation, and quantify the risk of each risk factor to AKI, which provide basis for adoption preventive measures.

All adult recipients who received grafts from DCD donors at the liver transplantation center of Beijing You’an Hospital, Capital Medical University, from November 2016 to June 2018 were enrolled. The medical records including preoperative, intraoperative, and postoperative data for up to 7 days were retrospectively assessed to develop this prediction model. Exclusion criteria were renal replacement therapy (RRT) or serum creatinine (SCr) > 1.5mg/dl before the LT, combined liver-kidney transplantation, pediatric or pregnancy patients, and the patients who discharge or death within 24 hours from operation.

Data collection

Pre-operative data included age, sex, body mass index (BMI), underlying liver disease, laboratory examination results, Model for End-Stage Liver Disease (MELD) score and Childs- Pugh-Turcott scores. Comorbidities and complications were also recorded.

Intra-operative data included surgical type, operative time, cold ischemia time (CIT), warm ischemia time, intraoperative blood loss, red blood cells (RBCs) transfusion, fresh frozen plasma transfusion, total intravenous infusion, and urine output.

Post-operative data included postoperative complications, peak serum ALT, AST, and SCr, need for RRT, mechanical ventilation (MV), hospital days, and follow-up duration.

AKI Definition

Post-LT AKI was defined according to the Kidney Disease Improving Global Outcomes (KDIGO) criteria [14], that is an increase in SCr by ≥ 26.5 umol/L within 48h or an increase in SCr to ≥ 1.5 times baseline within the first 7 postoperative days. AKI was classified into 3 stages: AKI stage 1 (AKI-1) with SCr to ≥ 26.5 umol/L or increase to 1.5–1.9-fold from baseline; AKI stage 2 (AKI-2), increase to 2–2.9-fold; AKI stage 3 (AKI-3), increase > 3-fold or increase in SCr to ≥ 354 umol/L or the initiation of RRT. Baseline SCr was identified as the first SCr measurement during hospitalization.

Model validation

AKI score model was prospectively validated in a new data set of 112 patients recruited in the same institution in the other period (July 2018 to January 2020). The exclusion criteria of the validation cohort were the same as those of the derivation cohort.

Statistical analysis

Baseline characteristics were compared using student- t-test for continuous variables and chi-square test for categorical variables as appropriate. For the development of the final prediction model, continuous variables were categorized according to the inflection point on the ROC curve and categorical variables underwent to χ² tests.

Cox proportional hazard regression models were used to determine risk factors and prediction model using a score-based method [15]. All significant factors in univariable analysis (p < 0.5) were further analyzed by the multivariable stepwise regression method, and p-value < 0.05 was used as the inclusion threshold. The probabilities of AKI for all possible cumulative prediction scores were estimated according to the formula

p\(=1-{S}_{0}^{\text{exp}\left({\sum }_{i=1}^{p}{\beta }_{i}{X}_{i}-{\sum }_{i=1}^{p}{\beta }_{i}{\stackrel{-}{X}}_{i}\right)}\)

where S₀ is the baseline disease-free probability, β_i is the regression coefficient for the ith variable, and X_i is the actual level for the ith variable, and \({\stackrel{-}{X}}_{i}\) is the mean level for the ith variable.

Calibration of the prediction model was assessed by comparing the predicted probability of three AKI risk groups to the actual probability. Model fit was assessed using the Hosmer– Lemeshow goodness-of-fit test. Statistical significance was determined by 2-tailed tests (p < 0.05). The statistical analyses were performed using SAS 9.4 (Windows, SAS Institute, Inc., Cary, NC, USA).

Baseline characteristics

141 adult patients who received a deceased donor transplant were enrolled in the derivation cohort, 9 patients were excluded from the study for the following reasons: AKI prior to liver transplantation (n = 7) and retransplantation (n = 2). In the validation cohort, 119 patients were recruited, 7 cases were excluded: 5 patients with AKI prior to liver transplantation and 2 cases with retransplantation. Finally, 132 patients were included in the derivation cohort and 112 in the validation cohort. The baseline characteristics of the derivation and validation cohorts are shown in Table 1. Patients in both cohorts were mostly male and the most common etiology of liver disease was hepatitis B. More recipients in cohort study combined with hypertension (6.8% vs. 14.3%; p = 0.0017). Importantly, no differences were observed for serum creatinine levels in both cohort(median 60 umol/l vs 62 umol/l; p = 0.271)before LT. CIT was longer in the derivation cohort (median 5.5 vs. 4 h; p < 0.0001). Overall, 103 (42.2%) of the recipients developed AKI, which was 49 (37.1%) in derivation cohort, compared to 54 (48.2%) in validation cohort (p = 0.084). Additionally, 7 (5.5%) of the recipients in the derivation cohort required postoperative RRT, and 9 (8.0%) recipients in the validation cohort needed RRT (Table 2).

Table 1

Characteristics of patients in developing and validation cohorts. Data are shown as medians and interquartile ranges and numbers (%). The p-value is obtained using a t-test or chi-square test as appropriate.
Characteristics	Derivation Cohort (n = 132)	Validation Cohort (n = 112)	ཐValue
Age, years	52(45,59)	54(45.8,61)	0.1150
Male gender (%)	114(86%)	93(83%)	0.4702
BMI	24.7(22.3,26.7)	24.5(21.7,26.2)	0.3610
Etiology of liver disease (%)
Hepatitis B(%)	87(65.9%)	66(58.9%)	0.1377
Hepatitis C(%)	5(3.8%)	8(7.1%)
Alcoholic(%)	21(15.9%)	12(10.7%)
Other(%)	19(14.4%)	26(23.2%)
HCC (%)	69(52.3%)	64(57.1%)	0.4465
sCR, µmol/l	60(49,71)	62(52,70)	0.8813
MELD score	9(6,15)	10(6,18)	0.4088
AST u/l	42.2(28,71.2)	49(31,93)	0.559
Medical history
Hypertension(%)	9(10.8%)	9(18.3%)	0.3261*
DM(%)	4(4.8%)	5(10.2%)
Hypertension + DM	2(2.4%)	2(4.1%)
CIT (hours)	7(5.5,8.6)	5(4,5.3)	< 0.0001
RBCs transfusion (ml)	1600(800,400)	1600(800,2000)	0.3478
MV (hours)	26.5(17.8,47.3)	32(15.8,60)	0.7869
BMI body mass index, HCC hepatocellular carcinoma, DM diabetes mellitus, MELD Model for End-stage Liver Disease, sCR serum creatinine, AST aspartate aminotransferase, CIT cold ischemia time, RBC red blood cell transfusion, MV mechanical ventilation

Table 2

Development of AKI after liver transplantation in both cohorts. Values are presented as numbers (%). The p-value is obtained using a chi-square test.
	Derivation cohort cohort(n = 132)	Validation cohort (n = 112)	Total (n = 244)	ཐ-value
No AKI	83(62.9%)	58(51.8%)	141(57.8%)	0.0804
Overall AKI	49(37.1%)	54(48.2%)	103(42.2%)
AKI stage 1	21(15.9%)	25(22.3%)	46(18.9%)
AKI stage 2	18(13.6%)	16(14.3%)	34(13.9%)
AKI stage 3	10(7.6%)	13(11.6%)	23(9.4%)
RRT	7(5.3%)	9(8%)	16(6.6%)
AKI acute kidney injury, RRT renal replacement treatment

model established

The risk variables associate with AKI are analysised using multivariable Cox proportional hazards models (Fig. 1). Five factors were identified as the strongest predictors included in final prediction model: preoperative AST ≥ 40u/l, CIT ≥ 7 hours, intraoperative RBC transfusion ≥ 1800ml, post-OLT MV ≥ 40 hours, pre-operative MELD score ≥ 15, leading to a range of 0–25 score points (Table 3). The cumulative incidence of AKI increased exponentially with every single point of the AKI prediction score (Fig. 2). Participants were divided into three risk groups according to the probability of developing AKI: low (< 10%), intermediate (11–29%) and high risk (≥ 30%). The predicted risks of AKI were 7.1%, 22.6%, and 61. 3% for the three risk categories matching with the observed risk 7.7%, 19.6%, and 55.6%. The calibration was good and Hosmer–Lemeshow test for the new model showed a nonsignificant difference between predicted and observed risks (P = 0.3185) (Fig. 3a).

Table 3

Multivariate analysis of risk factors for the development of AKI prediction model score
		Hazard ratio	95% CI	Coefficient β	P-Value	Points
Pre-op AST	< 40u/l				0.0143	0
Pre-op AST	≥ 40 u/l	0.386	0.18–0.827	-0.9512	0.0143	-2
CIT	< 7 hours				0.005	0
CIT	≥ 7 hours	2.841	1.37–5.888	1.0440	0.005	2
RBCs transfusion	< 1800 ml				0.0495	0
RBCs transfusion	≥ 1800 ml	2.064	1.002–4.251	0.7244	0.0495	1
MV	< 40 hours				0.0011	0
MV	≥ 40 hours	3.123	1.58–6.175	1.1389	0.0011	2
MELD score	< 15				< 0.0001	0
MELD score	≥ 15	5.937	2.544–13.854	1.7812	< 0.0001	3
Pre-op AST Preoperative aspartate aminotransferase, CIT cold ischemia time, RBCs red blood cells transfusion, MV mechanical ventilation, MELD Model for end-stage liver Disease

Validation of the new prediction model

In a new data set of 112 patients, the AKI score was prospectively validated. The prediction risk of AKI was 6.3%, 20.7%, and 63%, for the three risk categories established based on the predicted AKI. Meanwhile, the actual incidence of AKI in each risk category was 6.0%, 31.6%, and 67.9%. The Model fits were adequate in the derivation cohort (P = 0.3185) (Fig. 3b).

The overall incidence of AKI following LT using grafts from DCD in this study was 42.2%, which is similar to those used living donor liver grafts or partly used DCD donors [1,3,16]. CIT reduction may be the reason as CIT is an important risk factor for the development of AKI [7]. The most DCD donors come from our organ procurement organization for transplantation at our center, allowing us to keep the CIT short. As expected, MELD score pre-operative of recipient and CIT of graft were the most important factors following LT. The predictive risk of post-LT AKI when recipient with a MELD score ≥ 15 alone was 43.9%. The risk increases in the presence of prolonged CIT and excess RBC transfusions, especially the donor CIT ≥ 7 hours, the incidence of AKI will rise to 85.3%.

Surprisingly, Cox multivariate stepwise regression revealed that a mild increase in AST pre- LT was negatively associated with AKI. However, AST was not an independent predictor of AKI in univariate analysis, implying that AST indirectly affects AKI through ischemia preconditioning (IP), a method that may alleviate IRI injury [17]. This is consistent with the findings of another study, which found that mild injury of renal function prior to OLT may protect against AKI post-OLT via the IP [18]. IP is another method with extensive preclinical evidence that has a protective effect on the graft and other organs, but clinical evidence has been more ambiguous [19]. But from the pathophysiological mechanism, there is hope for reducing IRI through IP.

The fifth predictor was mechanical ventilation duration that more than 40 hours post-operation potentially increasing the risk of AKI. This may be controversial for patients who require prolonged ventilatory support after transplantation are usually severely ill or have a more complicated procedure than those who extubated early. However, theoretically MV may lead to the development of AKI through haemodynamic factors or ventilator-induced lung injury by triggering a pulmonary inflammatory reaction and subsequent systemic release of inflammatory mediators [20]. Clinically, prolonged mechanical ventilation time is a risk factor for AKI 48 hours after surgery [21]. Third, early extubation can improve splanchnic and liver blood flow, which may result in better liver graft recovery, which is the foundation of stabilizing other organs function in liver transplantation [22,23]. Early extubation is feasible and safe in LT patients, with potential physiologically benefits optimal patient outcomes [24]. This does not simply imply extubating at a specific time, but suggests that personalized care for patients on a liver transplant and early extubation successful is dependent on the recipient's pretransplant condition, intraoperative events, and clinicians' ability to predict posttransplant complications [25].

This is the first model to predict the risk of AKI in liver transplant recipients using DCD grafts, and evaluated the impact of MELD and CIT on AKI after LT simultaneously, which is more suitable for the current MELD based liver transplantation policy and donor selection criteria expansion situation.This prediction model has some limitations: it is a single-center study with a small sample size, false-positive outcomes are possible. Furthermore, other parameters impacting donor quality, such as age, BMI, and so on, were not considered. Third, more modifiable risk factors such as low blood pressure and intraoperative vasopressor use were not included in the model.

In conclusion, significant modifications in recipient and donor criteria have a significant impact on the outcome of liver transplantation. Despite the MELD score and CIT are the most important risk factors for renal dysfunction in DCD liver transplants, the prognosis may improve due to novel organ preservation techniques and other methods that can reduce IRI.

Acknowledgements The authors want to thank the Department of liver transplantation center staff for providing the cases.

Author contributions HL and XW conceived and designed the project with supervision from GL, QM, HL, XB, and ML acquired the clinical data. MX provided the statistical analysis. HL and JW involved in drafting and approved the final manuscript.

Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Availability of data and materials the datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Conflict of interest the authors HL, XW, MX, YW, JW, ML, GL and QM have no relationships with any companies whose products or services may be related to the article's subject matter.

Ethics approval and consent to participate All procedures were performed in accordance with the ethical standards of the institution and the 1964 Helsinki Declaration. The Ethics Committee on Clinical Trials of Beijing You’an Hospital, Capital Medical University approved this study.

Consent to participate informed consent was not needed since the nature of the study was observed

Kalisvaart M, Schlegel A, Umbro I, et al. The AKI Prediction Score: a new prediction model for acute kidney injury after liver transplantation. HPB (Oxford). 2019; 21:1707-1717
Wyssusek KH, Keys AL, Yung J, Moloney ET, Sivalingam P, Paul SK. Evaluation of perioperative predictors of acute kidney injury post orthotopic liver transplantation. Anaesth Intensive Care. 2015; 43:757-763
Hilmi IA, Damian D, Al-Khafaji A, et al. Acute kidney injury following orthotopic liver transplantation: incidence, risk factors, and effects on patient and graft outcomes. Br J Anaesth. 2015; 114:919-926
Trinh E, Alam A, Tchervenkov J, Cantarovich M. Impact of acute kidney injury following liver transplantation on long-term outcomes. Clin Transplant. 2017;31. https://doi.org/10.1111/ctr.12863
Barri YM, Sanchez EQ, Jennings LW, Melton LB, Hays S, Levy MF, Klintmalm GB. Acute kidney injury following liver transplantation: definition and outcome. Liver Transpl. 2009; 15:475-483
Kim JM, Jo YY, Na SW, Kim SI, Choi YS, Kim NO, Park JE, Koh SO. The predictors for continuous renal replacement therapy in liver transplant recipients. Transplant Proc. 2014; 46:184-191
Zongyi Y, Baifeng L, Funian Z, Hao L, Xin W. Risk factors of acute kidney injury after orthotopic liver transplantation in China. Sci Rep. 2017; 7:41555. https://doi.org/10.1038/srep41555
Park MH, Shim HS, Kim WH, Kim HJ, Kim DJ, Lee SH, et al. Clinical Risk Scoring Models for Prediction of Acute Kidney Injury after Living Donor Liver Transplantation: A Retrospective Observational Study. PLoS One. 2015;10: e0136230. https://doi.org/10.1371/journal.pone.0136230
Meersch M, Schmidt C, Hoffmeier A, Van Aken H, Wempe C, Gerss J, et al. Prevention of cardiac surgery-associated AKI by implementing the KDIGO guidelines in high risk patients identified by biomarkers: the PrevAKI randomized controlled trial. Intensive Care Med 2017; 43:1551–1561
Leithead JA, Rajoriya N, Gunson BK, Muiesan P, Ferguson JW. The evolving use of higher risk grafts is associated with an increased incidence of acute kidney injury after liver transplantation. J Hepatol. 2014; 60:1180-1186
Collard CD, Gelman S. Pathophysiology, clinical manifestations, and prevention of ischemia-reperfusion injury. Anesthesiology. 2001; 9:1133-1138
Nasralla D, Coussios CC, Mergental H, Akhtar MZ, Butler AJ, Ceresa CDL, et al. Consortium for Organ Preservation in Europe. A randomized trial of normothermic preservation in liver transplantation. Nature. 2018; 557:50-56
Guo Z, Zhao Q, Huang S, Huang C, Wang D, Yang L, et al. Ischaemia-free liver transplantation in humans: a first-in-human trial. Lancet Reg Health West Pac.2021; 16:100260. https://doi.org/ 10.1016/j.lanwpc.2021.100260
Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120:c179-184
Liu J, Tseng TC, Yang HI, Lee MH, Batrla-Utermann R, Jen CL, et al. Predicting Hepatitis B Virus (HBV) Surface Antigen Seroclearance in HBV e Antigen-Negative Patients With Chronic Hepatitis B: External Validation of a Scoring System. J Infect Dis. 2015; 211:1566-1573
Mogawer MS, El-Rahman El-Shazly MA, Ali AY, Abd El-Ghany AM, Elhamid SA. Novel biomarkers of acute kidney injury following living donor liver transplantation. Saudi J Kidney Dis Transpl. 2020; 31:360-367
Rampes S, Ma D. Hepatic ischemia-reperfusion injury in liver transplant setting: mechanisms and protective strategies. J Biomed Res. 2019; 33:221-234
Iglesias JI, DePalma JA, Levine JS. Risk factors for acute kidney injury following orthotopic liver transplantation: the impact of changes in renal function while patients await transplantation. BMC Nephrol. 2010; 11:30. https://doi.org/10.1186/1471-2369-11-30
Li DY, Shi XJ, Li W, Sun XD, Wang GY. Ischemic preconditioning and remote ischemic preconditioning provide combined protective effect against ischemia/reperfusion injury. Life Sci. 2016; 150:76-80
Kuiper JW, Groeneveld AB, Slutsky AS, Plotz FB: Mechanical ventilation and acute renal failure. Crit Care Med 2005, 33:1408-1415
Li S, Wang S, Priyanka P, Kellum JA. Acute Kidney Injury in Critically Ill Patients After Noncardiac Major Surgery: Early Versus Late Onset. Crit Care Med. 2019;47: e437-e44
Yuan H, Tuttle-Newhall JE, Chawa V, et al. Prognostic impact of mechanical ventilation after liver transplantation: a national database study. Am J Surg. 2014; 208:582-590
O’Meara ME, Whiteley SM, Sellors JM, Luntley JM, Davison S, McClean P, et al. Immediate extubation of children following liver transplantation is safe and may be beneficial. Transplantation 2005; 80:959-963
Taner CB, Willingham DL, Bulatao IG, Shine TS, Peiris P, Torp KD, et al. Is a mandatory intensive care unit stay needed after liver transplantation? Feasibility of fast-tracking to the surgical ward after liver transplantation. Liver Transpl. 2012; 18:361-369
Biancofiore G, Tomescu DR, Mandell MS. Rapid recovery of liver transplantation recipients by implementation of fast-track care steps: What is holding us back? Semin Cardiothorac Vasc Anesth 2018; 22:191–196

Download PDF

Version 1

posted

You are reading this latest preprint version

Acute Kidney Injury Prediction Model After Liver Transplantation Using Grafts From Donors After Cardiac Death

Status:

Version 1

Abstract

Figures

Introduction

Methods

Data collection

AKI Definition

Model validation

Statistical analysis

Results

Baseline characteristics

model established

Validation of the new prediction model

Discussion

Declarations

References

Status:

Version 1