The AKI Prediction Score: a new prediction model for acute kidney injury after liver transplantation

Complications in Post-Liver Transplant Patients

Liver transplantation (LT) is the treatment of choice for liver failure and selected cases of malignancies. Transplantation activity has increased over the years, and indications for LT have been widened, leading to organ shortage. To face this condition, a high selection of recipients with prioritizing systems and an enlargement of the donor pool were necessary. Several authors published their case series reporting the results obtained with the use of marginal donors, which seem to have progressively improved over the years. The introduction of in situ and ex situ machine perfusion, although still strongly debated, and better knowledge and treatment of the complications may have a role in achieving better results. With longer survival rates, a significant number of patients will suffer from long-term complications. An extensive review of the literature concerning short- and long-term outcomes is reported trying to highlight the most recent findings. The heterogeneity of the behaviors within the different centers is evident, leading to a difficult comparison of the results and making explicit the need to obtain more consent from experts.

Development of a nomogram for the prediction of acute kidney injury after liver transplantation: a model based on clinical parameters and postoperative cystatin C level

BACKGROUND

Acute kidney injury (AKI) is common after liver transplantation (LT). We developed a nomogram model to predict post-LT AKI.

METHODS

A total of 120 patients were eligible for inclusion in the study. Clinical information was extracted from the institutional electronic medical record system. Blood samples were collected prior to surgery and immediately after surgery. Univariable and multivariate logistic regression were used to identify independent risk factors. Finally, a nomogram was developed based on the final multivariable logistic regression model.

RESULTS

In total, 58 (48.3%) patients developed AKI. Multivariable logistic regression revealed four independent risk factors for post-LT AKI: operation duration [odds ratio (OR) = 1.728, 95% confidence interval (CI) = 1.121-2.663, p = 0.013], intraoperative hypotension (OR = 3.235, 95% CI = 1.316-7.952, p = 0.011), postoperative cystatin C level (OR = 1.002, 95% CI = 1.001-1.004, p = 0.005) and shock (OR = 4.002, 95% CI = 0.893-17.945, p = 0.070). Receiver operating characteristic curve analysis was used to evaluate model discrimination. The area under the curve value was 0.815 (95% CI = 0.737-0.894).

CONCLUSION

The model based on combinations of clinical parameters and postoperative cystatin C levels had a higher predictive performance for post-LT AKI than the model based on clinical parameters or postoperative cystatin C level alone. Additionally, we developed an easy-to-use nomogram based on the final model, which could aid in the early detection of AKI and improve the prognosis of patients after LT.

Characterization of Risk Prediction Models for Acute Kidney Injury: A Systematic Review and Meta-analysis

Importance

Despite the expansion of published prediction models for acute kidney injury (AKI), there is little evidence of uptake of these models beyond their local derivation nor data on their association with patient outcomes.

Objective

To systematically review published AKI prediction models across all clinical subsettings.

Data Sources

MEDLINE via PubMed (January 1946 to April 2021) and Embase (January 1947 to April 2021) were searched using medical subject headings and text words related to AKI and prediction models.

Study Selection

All studies that developed a prediction model for AKI, defined as a statistical model with at least 2 predictive variables to estimate future occurrence of AKI, were eligible for inclusion. There was no limitation on study populations or methodological designs.

Data Extraction and Synthesis

Two authors independently searched the literature, screened the studies, and extracted and analyzed the data following the Preferred Reporting Items for Systematic Review and Meta-analyses guideline. The data were pooled using a random-effects model, with subgroups defined by 4 clinical settings. Between-study heterogeneity was explored using multiple methods, and funnel plot analysis was used to identify publication bias.

Main Outcomes and Measures

C statistic was used to measure the discrimination of prediction models.

Results

Of the 6955 studies initially identified through literature searching, 150 studies, with 14.4 million participants, met the inclusion criteria. The study characteristics differed widely in design, population, AKI definition, and model performance assessments. The overall pooled C statistic was 0.80 (95% CI, 0.79-0.81), with pooled C statistics in different clinical subsettings ranging from 0.78 (95% CI, 0.75-0.80) to 0.82 (95% CI, 0.78-0.86). Between-study heterogeneity was high overall and in the different clinical settings (eg, contrast medium-associated AKI: I2 = 99.9%; P < .001), and multiple methods did not identify any clear sources. A high proportion of models had a high risk of bias (126 [84.4%]) according to the Prediction Model Risk Of Bias Assessment Tool.

Conclusions and Relevance

In this study, the discrimination of the published AKI prediction models was good, reflected by high C statistics; however, the wide variation in the clinical settings, populations, and predictive variables likely drives the highly heterogenous findings that limit clinical utility. Standardized procedures for development and validation of prediction models are urgently needed.

Lack of furosemide responsiveness predict severe acute kidney injury after liver transplantation

Acute kidney injury (AKI) remains to be a common but severe complication after liver transplantation (LT). However, there are still few clinically validated biomarkers. A total of 214 patients who underwent routine furosemide (1-2 mg/kg) after LT were retrospectively included. The urine output during the first 6 h was recorded to evaluate the predictive value of AKI stage 3 and renal replacement therapy (RRT). 105 (49.07%) patients developed AKI, including 21 (9.81%) progression to AKI stage 3 and 10 (4.67%) requiring RRT. The urine output decreased with the increasing severity of AKI. The urine output of AKI stage 3 did not significantly increase after the use of furosemide. The area under the receiver operator characteristic (ROC) curves for the total urine output in the first hour to predict progression to AKI stage 3 was 0.94 (p < 0.001). The ideal cutoff for predicting AKI progression during the first hour was a urine volume of less than 200 ml with a sensitivity of 90.48% and specificity of 86.53%. The area under the ROC curves for the total urine output in the six hours to predict progression to RRT was 0.944 (p < 0.001). The ideal cutoff was a urine volume of less than 500 ml with a sensitivity of 90% and specificity of 90.91%. Severe AKI after liver transplantation seriously affects the outcome of patients. Lack of furosemide responsiveness quickly and accurately predict AKI stage 3, and patients requiring RRT after the operation.

Machine learning for acute kidney injury: Changing the traditional disease prediction mode

Acute kidney injury (AKI) is a serious clinical comorbidity with clear short-term and long-term prognostic implications for inpatients. The diversity of risk factors for AKI has been recognized in previous studies, and a series of predictive models have been developed using traditional statistical methods in conjunction with its preventability, but they have failed to meet the expectations in limited clinical applications, the rapid spread of electronic health records and artificial intelligence machine learning technology has brought new hope for the construction of AKI prediction models. In this article, we systematically review the definition and classification of machine learning methods, modeling ideas and evaluation methods, and the characteristics and current status of modeling studies. According to the modeling objectives, we subdivided them into critical care medical setting models, all medical environment models, special surgery models, special disease models, and special nephrotoxin exposure models. As the first review article to comprehensively summarize and analyze machine learning prediction models for AKI, we aim to objectively describe the advantages and disadvantages of machine learning approaches to modeling, and help other researchers more quickly and intuitively understand the current status of modeling research, inspire ideas and learn from experience, so as to guide and stimulate more research and more in-depth exploration in the future, which will ultimately provide greater help to improve the overall status of AKI diagnosis and treatment.

A Novel Online Calculator Predicting Acute Kidney Injury After Liver Transplantation: A Retrospective Study

Acute kidney injury (AKI) after liver transplantation (LT) is a common complication, and its development is thought to be multifactorial. We aimed to investigate potential risk factors and build a model to identify high-risk patients. A total of 199 LT patients were enrolled and each patient data was collected from the electronic medical records. Our primary outcome was postoperative AKI as diagnosed and classified by the KDIGO criteria. A least absolute shrinkage and selection operating algorithm and multivariate logistic regression were utilized to select factors and construct the model. Discrimination and calibration were used to estimate the model performance. Decision curve analysis (DCA) was applied to assess the clinical application value. Five variables were identified as independent predictors for post-LT AKI, including whole blood serum lymphocyte count, RBC count, serum sodium, insulin dosage and anhepatic phase urine volume. The nomogram model showed excellent discrimination with an AUC of 0.817 (95% CI: 0.758-0.876) in the training set. The DCA showed that at a threshold probability between 1% and 70%, using this model clinically may add more benefit. In conclusion, we developed an easy-to-use tool to calculate the risk of post-LT AKI. This model may help clinicians identify high-risk patients.

Machine Perfusion for Extended Criteria Donor Livers: What Challenges Remain?

Based on the renaissance of dynamic preservation techniques, extended criteria donor (ECD) livers reclaimed a valuable eligibility in the transplantable organ pool. Being more vulnerable to ischemia, ECD livers carry an increased risk of early allograft dysfunction, primary non-function and biliary complications and, hence, unveiled the limitations of static cold storage (SCS). There is growing evidence that dynamic preservation techniques—dissimilar to SCS—mitigate reperfusion injury by reconditioning organs prior transplantation and therefore represent a useful platform to assess viability. Yet, a debate is ongoing about the advantages and disadvantages of different perfusion strategies and their best possible applications for specific categories of marginal livers, including organs from donors after circulatory death (DCD) and brain death (DBD) with extended criteria, split livers and steatotic grafts. This review critically discusses the current clinical spectrum of livers from ECD donors together with the various challenges and posttransplant outcomes in the context of standard cold storage preservation. Based on this, the potential role of machine perfusion techniques is highlighted next. Finally, future perspectives focusing on how to achieve higher utilization rates of the available donor pool are highlighted.

Modelling kidney outcomes based on MELD eras - impact of MELD score in renal endpoints after liver transplantation

Background

Acute kidney injury is a common complication in solid organ transplants, notably liver transplantation. The MELD is a score validated to predict mortality of cirrhotic patients, which is also used for organ allocation, however the influence of this allocation criteria on AKI incidence and mortality after liver transplantation is still uncertain.

Methods

This is a retrospective single center study of a cohort of patients submitted to liver transplant in a tertiary Brazilian hospital: Jan/2002 to Dec/2013, divided in two groups, before and after MELD implementation (pre-MELD and post MELD). We evaluate the differences in AKI based on KDIGO stages and mortality rates between the two groups.

Results

Eight hundred seventy-four patients were included, 408 in pre-MELD and 466 in the post MELD era. The proportion of patients that developed AKI was lower in the post MELD era (p 0.04), although renal replacement therapy requirement was more frequent in this group (p < 0.01). Overall mortality rate at 28, 90 and 365 days was respectively 7%, 11% and 15%. The 1-year mortality rate was lower in the post MELD era (20% vs. 11%, p < 0.01). AKI incidence was 50% lower in the post MELD era even when adjusted for clinically relevant covariates (p < 0.01).

Conclusion

Liver transplants performed in the post MELD era had a lower incidence of AKI, although there were more cases requiring dialysis. 1-year mortality was lower in the post MELD era, suggesting that patient care was improved during this period.

Liver perfusion strategies: what is best and do ischemia times still matter?

PURPOSE OF REVIEW

This review describes recent developments in the field of liver perfusion techniques.

RECENT FINDINGS

Dynamic preservation techniques are increasingly tested due to the urgent need to improve the overall poor donor utilization. With their exposure to warm ischemia, livers from donors after circulatory death (DCD) transmit additional risk for severe complications after transplantation. Although the superiority of dynamic approaches compared to static-cold-storage is widely accepted, the number of good quality studies remains limited. Most risk factors, particularly donor warm ischemia, and accepted thresholds are inconsistently reported, leading to difficulties to assess the impact of new preservation technologies. Normothermic regional perfusion (NRP) leads to good outcomes after DCD liver transplantation, with however short ischemia times. While randomized controlled trials (RCT) with NRP are lacking, results from the first RCTs with ex-situ perfusion were reported. Hypothermic oxygenated perfusion was shown to protect DCD liver recipients from ischemic cholangiopathy. In contrast, endischemic normothermic perfusion seems to not impact on the development of biliary complications, although this evidence is only available from retrospective studies.

SUMMARY

Dynamic perfusion strategies impact posttransplant outcomes and are increasingly commissioned in various countries along with more evidence from RCTs. Transparent reporting of risk and utilization with uniform definitions is required to compare the role of different preservation strategies in DCD livers with prolonged ischemia times.

Validation of 5 models predicting transfusion, bleeding, and mortality in liver transplantation: an observational cohort study

BACKGROUND

Historically, orthotopic liver transplantation (OLT) has been associated with massive blood loss, blood transfusion and morbidity. In order to predict such outcomes five nomograms have been published relating to transfusions and morbidity associated with OLTs. These nomograms, developed on the basis of three cohorts of patients consisting of 406, 750, and 800 having undergone OLTs, aimed to predict a transfusion of ≥1 red blood cell unit (RBC), a transfusion of >2 RBC units, a blood loss of >900 ml, as well as one-month and one-year survival rates. The aim of this study was to validate these five nomograms in a contemporary, independent cohort of patients.

METHODS

Five nomograms were previously developed based on 406, 750, and 800 OLTs. In this study we performed a temporal validation of these nomograms on contemporary patients that consisted of three cohorts of 800, 250, and 200 OLTs. Logistic regression coefficients from the historic development cohorts were applied to the three contemporary temporal validation cohorts.

RESULTS

The most accurate nomogram was able to predict transfusion of ≥1 RBC units with an area under the curve (AUC) was 0.91. The second-best nomogram was able to predict bleeding of >900 ml with an AUC of 0.70. T he AUC of the third nomogram (transfusion of >2 RBC units) was 0.70. However, is temporal validation was suboptimal, due to a low prevalence of OLTs transfused with >2 RBC units. The last 2 nomograms exhibited clearly suboptimal AUC values of 0.54 and 0.61.

CONCLUSION

Two of the five nomograms predict blood transfusion and blood loss with excellent accuracy. Transfusion of ≥1 RBC unit and blood loss of >900 ml can be predicted on the basis of these nomograms. However, these nomograms are not accurate to predict one-month and one-year survival rates. These results should be further cross-validated, ideally prospectively, in additional external independent cohorts.

[Kidney failure after liver transplantation]

One third of cirrhotic patients present impaired kidney function. It has multifactorial causes and has a harmful effect on patients' morbi-mortality before and after liver transplant. Kidney function does not improve in all patients after liver transplantation and liver-transplant recipients are at high risk of developing chronic kidney disease. Causes for renal dysfunction can be divided in three groups: preoperative, peroperative and postoperative factors. To date, there is no consensus for the modality of evaluation the risk for chronic kidney disease after liver transplantation, and for its prevention. In the present review, we describe the outcome of kidney function after liver transplantation, and the prognostic factors of chronic kidney disease to determine a risk stratification for each patient. Furthermore, we discuss therapeutic options to prevent kidney dysfunction in this setting, and highlight the indications of combined liver-kidney transplantation.

Risk factors and prediction of acute kidney injury after liver transplantation: Logistic regression and artificial neural network approaches

BACKGROUND

Acute kidney injury (AKI) has serious consequences on the prognosis of patients undergoing liver transplantation. Recently, artificial neural network (ANN) was reported to have better predictive ability than the classical logistic regression (LR) for this postoperative outcome.

AIM

To identify the risk factors of AKI after deceased-donor liver transplantation (DDLT) and compare the prediction performance of ANN with that of LR for this complication.

METHODS

Adult patients with no evidence of end-stage kidney dysfunction (KD) who underwent the first DDLT according to model for end-stage liver disease (MELD) score allocation system was evaluated. AKI was defined according to the International Club of Ascites criteria, and potential predictors of postoperative AKI were identified by LR. The prediction performance of both ANN and LR was tested.

RESULTS

The incidence of AKI was 60.6% (n = 88/145) and the following predictors were identified by LR: MELD score > 25 (odds ratio [OR] = 1.999), preoperative kidney dysfunction (OR = 1.279), extended criteria donors (OR = 1.191), intraoperative arterial hypotension (OR = 1.935), intraoperative massive blood transfusion (MBT) (OR = 1.830), and postoperative serum lactate (SL) (OR = 2.001). The area under the receiver-operating characteristic curve was best for ANN (0.81, 95% confidence interval [CI]: 0.75-0.83) than for LR (0.71, 95%CI: 0.67-0.76). The root-mean-square error and mean absolute error in the ANN model were 0.47 and 0.38, respectively.

CONCLUSION

The severity of liver disease, pre-existing kidney dysfunction, marginal grafts, hemodynamic instability, MBT, and SL are predictors of postoperative AKI, and ANN has better prediction performance than LR in this scenario.

Identifying Patients at Risk of Acute Kidney Injury Among Medicare Beneficiaries With Type 2 Diabetes Initiating SGLT2 Inhibitors: A Machine Learning Approach

Introduction: To predict acute kidney injury (AKI) risk in patients with type 2 diabetes (T2D) prescribed sodium-glucose cotransporter two inhibitors (SGLT2i).

Methods: Using a 5% random sample of Medicare claims data, we identified 17,694 patients who filled ≥1 prescriptions for canagliflozin, dapagliflozin and empagliflozin in 2013–2016. The cohort was split randomly and equally into training and testing sets. We measured 65 predictor candidates using claims data from the year prior to SGLT2i initiation. We then applied three machine learning models, including random forests (RF), elastic net and least absolute shrinkage and selection operator (LASSO) for risk prediction.

Results: The incidence rate of AKI was 1.1% over a median 1.5 year follow up. Among three machine learning methods, RF produced the best prediction (C-statistic = 0.72), followed by LASSO and elastic net (both C-statistics = 0.69). Among individuals classified in the top 10% of the RF risk score (i.e., high risk group), the actual incidence rate of AKI was as high as 3.7%. In the logistic regression model including 14 important risk factors selected by LASSO, use of loop diuretics [adjusted odds ratio (95% confidence interval): 3.72 (2.44–5.76)] had the strongest association with AKI incidence.

Disscusion: Our machine learning model efficiently identified patients at risk of AKI among Medicare beneficiaries with T2D undergoing SGLT2i treatment.

Serum Lactate Level in Early Stage Is Associated With Acute Kidney Injury in Traumatic Brain Injury Patients

Background

Acute kidney injury (AKI) is a common complication in the clinical practice of managing patients with traumatic brain injury (TBI). Avoiding the development of AKI is beneficial for the prognosis of patients with TBI. We designed this study to testify whether serum lactate could be used as a predictive marker of AKI in patients with TBI.

Materials and Methods

In total, 243 patients with TBI admitted to our hospital were included in this study. Univariate and multivariate logistic regression analyses were utilized to analyze the association between lactate and AKI. The receiver operating characteristic (ROC) curves were drawn to verify the predictive value of lactate and the logistic model.

Results

Acute kidney injury group had higher age (p = 0.016), serum creatinine (p < 0.001), lactate (p < 0.001), and lower Glasgow Coma Scale (GCS; p = 0.021) than non-AKI group. Multivariate logistic regression showed that age [odds ratio (OR) = 1.026, p = 0.022], serum creatinine (OR = 1.020, p = 0.010), lactate (OR = 1.227, p = 0.031), fresh frozen plasma (FFP) transfusion (OR = 2.421, p = 0.045), and platelet transfusion (OR = 5.502, p = 0.044) were risk factors of AKI in patients with TBI. The area under the ROC curve (AUC) values of single lactate and predictive model were 0.740 and 0.807, respectively.

Conclusion

Serum lactate level in the early phase is associated with AKI in patients with TBI. Lactate is valuable for clinicians to evaluate the probability of AKI in patients with TBI.

Machine learning approach to predict acute kidney injury after liver surgery

BACKGROUND

Acute kidney injury (AKI) after surgery appears to increase the risk of death in patients with liver cancer. In recent years, machine learning algorithms have been shown to offer higher discriminative efficiency than classical statistical analysis.

AIM

To develop prediction models for AKI after liver cancer resection using machine learning techniques.

METHODS

We screened a total of 2450 patients who had undergone primary hepatocellular carcinoma resection at Changzheng Hospital, Shanghai City, China, from January 1, 2015 to August 31, 2020. The AKI definition used was consistent with the Kidney Disease: Improving Global Outcomes. We included in our analysis preoperative data such as demographic characteristics, laboratory findings, comorbidities, and medication, as well as perioperative data such as duration of surgery. Computerized algorithms used for model development included logistic regression (LR), support vector machine (SVM), random forest (RF), extreme gradient boosting (XGboost), and decision tree (DT). Feature importance was also ranked according to its contribution to model development.

RESULTS

AKI events occurred in 296 patients (12.1%) within 7 d after surgery. Among the original models based on machine learning techniques, the RF algorithm had optimal discrimination with an area under the curve value of 0.92, compared to 0.87 for XGBoost, 0.90 for DT, 0.90 for SVM, and 0.85 for LR. The RF algorithm also had the highest concordance-index (0.86) and the lowest Brier score (0.076). The variable that contributed the most in the RF algorithm was age, followed by cholesterol, and surgery time.

CONCLUSION

Machine learning algorithms are highly effective in discriminating patients at high risk of developing AKI. The successful application of machine learning models may help guide clinical decisions and help improve the long-term prognosis of patients.

Development and validation of a nomogram for predicting acute kidney injury after orthotopic liver transplantation

BACKGROUND

We aim to develop and validate a nomogram model for predicting severe acute kidney injury (AKI) after orthotopic liver transplantation (OLT).

METHODS

A total of 576 patients who received OLT in our center were enrolled. They were assigned to the development and validation cohort according to the time of inclusion. Univariable and multivariable logistic regression using the forward variable selection routine were applied to find risk factors for post-OLT severe AKI. Based on the results of multivariable analysis, a nomogram was developed and validated. Patients were followed up to assess the long-term mortality and development of chronic kidney disease (CKD).

RESULTS

Overall, 35.9% of patients were diagnosed with severe AKI. Multivariable logistic regression analysis revealed that recipients' BMI (OR 1.10, 95% CI 1.04-1.17, p = 0.012), hypertension (OR 2.32, 95% CI 1.22-4.45, p = 0.010), preoperative serum creatine (sCr) (OR 0.96, 95% CI 0.95-0.97, p < 0.001), and intraoperative fresh frozen plasm (FFP) transfusion (OR for each 1000 ml increase 1.34, 95% CI 1.03-1.75, p = 0.031) were independent risk factors for post-OLT severe AKI. They were all incorporated into the nomogram. The area under the ROC curve (AUC) was 0.73 (p < 0.05) and 0.81 (p < 0.05) in the development and validation cohort. The calibration curve demonstrated the predicted probabilities of severe AKI agreed with the observed probabilities (p > 0.05). Kaplan-Meier survival analysis showed that patients in the high-risk group stratified by the nomogram suffered significantly poorer long-term survival than the low-risk group (HR 1.92, p < 0.01). The cumulative risk of CKD was higher in the severe AKI group than no severe AKI group after competitive risk analysis (HR 1.48, p < 0.05).

CONCLUSIONS

With excellent predictive abilities, the nomogram may be a simple and reliable tool to identify patients at high risk for severe AKI and poor long-term prognosis after OLT.

Artificial neural network for prediction of acute kidney injury after liver transplantation for cirrhosis and hepatocellular carcinoma

Acute kidney injury (AKI) has serious consequences on the prognosis of patients undergoing liver transplantation (LT) for liver cancer and cirrhosis. Artificial neural network (ANN) has recently been proposed as a useful tool in many fields in the setting of solid organ transplantation and surgical oncology, where patient prognosis depends on a multidimensional and nonlinear relationship between variables pertaining to the surgical procedure, the donor (graft characteristics), and the recipient comorbidities. In the specific case of LT, ANN models have been developed mainly to predict survival in patients with cirrhosis, to assess the best donor-to-recipient match during allocation processes, and to foresee postoperative complications and outcomes. This is a specific opinion review on the role of ANN in the prediction of AKI after LT for liver cancer and cirrhosis, highlighting potential strengths of the method to forecast this serious postoperative complication.

Recommendations for Donor and Recipient Selection and Risk Prediction: Working Group Report From the ILTS Consensus Conference in DCD Liver Transplantation

Although the utilization of donation after circulatory death donors (DCDs) for liver transplantation (LT) has increased steadily, much controversy remains, and no common acceptance criteria exist with regard to donor and recipient risk factors and prediction models. A consensus conference was organized by International Liver Transplantation Society on January 31, 2020, in Venice, Italy, to review the current clinical practice worldwide regarding DCD-LT and to develop internationally accepted guidelines. The format of the conference was based on the grade system. International experts in this field were allocated to 6 working groups and prepared evidence-based recommendations to answer-specific questions considering the currently available literature. Working group members and conference attendees served as jury to edit and confirm the final recommendations presented at the end of the conference by each working group separately. This report presents the final statements and recommendations provided by working group 2, covering the entire spectrum of donor and recipient risk factors and prediction models in DCD-LT.

Kidney Failure after Liver Transplantation

One-third of patients with cirrhosis present kidney failure (AKI and CKD). It has multifactorial causes and a harmful effect on morbidity and mortality before and after liver transplantation. Kidney function does not improve in all patients after liver transplantation, and liver transplant recipients are at a high risk of developing chronic kidney disease. The causes of renal dysfunction can be divided into three groups: pre-operative, perioperative and post-operative factors. To date, there is no consensus on the modality to evaluate the risk of chronic kidney disease after liver transplantation, or for its prevention. In this narrative review, we describe the outcome of kidney function after liver transplantation, and the prognostic factors of chronic kidney disease in order to establish a risk categorization for each patient. Furthermore, we discuss therapeutic options to prevent kidney dysfunction in this context, and highlight the indications of combined liver–kidney transplantation.

An explainable supervised machine learning predictor of acute kidney injury after adult deceased donor liver transplantation

Background

Early prediction of acute kidney injury (AKI) after liver transplantation (LT) facilitates timely recognition and intervention. We aimed to build a risk predictor of post-LT AKI via supervised machine learning and visualize the mechanism driving within to assist clinical decision-making.

Methods

Data of 894 cases that underwent liver transplantation from January 2015 to September 2019 were collected, covering demographics, donor characteristics, etiology, peri-operative laboratory results, co-morbidities and medications. The primary outcome was new-onset AKI after LT according to Kidney Disease Improving Global Outcomes guidelines. Predicting performance of five classifiers including logistic regression, support vector machine, random forest, gradient boosting machine (GBM) and adaptive boosting were respectively evaluated by the area under the receiver-operating characteristic curve (AUC), accuracy, F1-score, sensitivity and specificity. Model with the best performance was validated in an independent dataset involving 195 adult LT cases from October 2019 to March 2021. SHapley Additive exPlanations (SHAP) method was applied to evaluate feature importance and explain the predictions made by ML algorithms.

Results

430 AKI cases (55.1%) were diagnosed out of 780 included cases. The GBM model achieved the highest AUC (0.76, CI 0.70 to 0.82), F1-score (0.73, CI 0.66 to 0.79) and sensitivity (0.74, CI 0.66 to 0.8) in the internal validation set, and a comparable AUC (0.75, CI 0.67 to 0.81) in the external validation set. High preoperative indirect bilirubin, low intraoperative urine output, long anesthesia time, low preoperative platelets, and graft steatosis graded NASH CRN 1 and above were revealed by SHAP method the top 5 important variables contributing to the diagnosis of post-LT AKI made by GBM model.

Conclusions

Our GBM-based predictor of post-LT AKI provides a highly interoperable tool across institutions to assist decision-making after LT.

Graphic abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-021-02990-4.

Management of Patients With Non-alcoholic Steatohepatitis Undergoing Liver Transplantation: Considerations for the Anesthesiologist

Non-alcoholic fatty liver disease (NAFLD) currently affects more than 25% of the world population and is rising. NAFLD can progress to non-alcoholic steatohepatitis that is associated with hepatic inflammation and fibrosis and can result in cirrhosis with subsequent liver failure. Non-alcoholic steatohepatitis (NASH) has now emerged as one of the leading etiologies for a liver transplant among adults in the United States. Given the rising incidence of liver transplants in patients with NASH-related cirrhosis, it is essential for anesthesiologists to be familiar with this condition as well as with NASH-related comorbidities and perioperative complications. Not only is NASH linked to metabolic syndrome, but it also is independently associated with cardiovascular disease, renal and thyroid dysfunction, obstructive sleep apnea (OSA), and a hypercoagulable state. The association with these conditions can affect the perioperative outcome of these patients, particularly because of increased mortality from major adverse cardiovascular events and sepsis. In order to decrease the perioperative morbidity and mortality of patients with NASH undergoing a liver transplant, a multidisciplinary approach to their perioperative management is essential, along with careful preoperative evaluation and aggressive intraoperative and postoperative monitoring. The focus of this review article is to provide a comprehensive overview of challenges associated with liver transplants in patients with NASH and to provide suggestions for appropriate patient selection and perioperative management.

Application of machine learning models for predicting acute kidney injury following donation after cardiac death liver transplantation

BACKGROUND

Acute kidney injury (AKI) is a common complication after liver transplantation (LT) and is an indicator of poor prognosis. The establishment of a more accurate preoperative prediction model of AKI could help to improve the prognosis of LT. Machine learning algorithms provide a potentially effective approach.

METHODS

A total of 493 patients with donation after cardiac death LT (DCDLT) were enrolled. AKI was defined according to the clinical practice guidelines of kidney disease: improving global outcomes (KDIGO). The clinical data of patients with AKI (AKI group) and without AKI (non-AKI group) were compared. With logistic regression analysis as a conventional model, four predictive machine learning models were developed using the following algorithms: random forest, support vector machine, classical decision tree, and conditional inference tree. The predictive power of these models was then evaluated using the area under the receiver operating characteristic curve (AUC).

RESULTS

The incidence of AKI was 35.7% (176/493) during the follow-up period. Compared with the non-AKI group, the AKI group showed a remarkably lower survival rate (P < 0.001). The random forest model demonstrated the highest prediction accuracy of 0.79 with AUC of 0.850 [95% confidence interval (CI): 0.794-0.905], which was significantly higher than the AUCs of the other machine learning algorithms and logistic regression models (P < 0.001).

CONCLUSIONS

The random forest model based on machine learning algorithms for predicting AKI occurring after DCDLT demonstrated stronger predictive power than other models in our study. This suggests that machine learning methods may provide feasible tools for forecasting AKI after DCDLT.

Prediction models for acute kidney injury in critically ill patients: a protocol for systematic review and critical appraisal

INTRODUCTION

Acute kidney injury (AKI) has high morbidity and mortality in intensive care units, which can lead to chronic kidney disease, more costs and longer hospital stay. Early identification of AKI is crucial for clinical intervention. Although various risk prediction models have been developed to identify AKI, the overall predictive performance varies widely across studies. Owing to the different disease scenarios and the small number of externally validated cohorts in different prediction models, the stability and applicability of these models for AKI in critically ill patients are controversial. Moreover, there are no current risk-classification tools that are standardised for prediction of AKI in critically ill patients. The purpose of this systematic review is to map and assess prediction models for AKI in critically ill patients based on a comprehensive literature review.

METHODS AND ANALYSIS

A systematic review with meta-analysis is designed and will be conducted according to the CHecklist for critical Appraisal and data extraction for systematic Reviews of prediction Modelling Studies (CHARMS). Three databases including PubMed, Cochrane Library and EMBASE from inception through October 2020 will be searched to identify all studies describing development and/or external validation of original multivariable models for predicting AKI in critically ill patients. Random-effects meta-analyses for external validation studies will be performed to estimate the performance of each model. The restricted maximum likelihood estimation and the Hartung-Knapp-Sidik-Jonkman method under a random-effects model will be applied to estimate the summary C statistic and 95% CI. 95% prediction interval integrating the heterogeneity will also be calculated to pool C-statistics to predict a possible range of C-statistics of future validation studies. Two investigators will extract data independently using the CHARMS checklist. Study quality or risk of bias will be assessed using the Prediction Model Risk of Bias Assessment Tool.

ETHICS AND DISSEMINATION

Ethical approval and patient informed consent are not required because all information will be abstracted from published literatures. We plan to share our results with clinicians and publish them in a general or critical care medicine peer-reviewed journal. We also plan to present our results at critical care international conferences.

OSF REGISTRATION NUMBER

10.17605/OSF.IO/X25AT.

Prognostic nutritional index predicts acute kidney injury and mortality of patients in the coronary care unit

The current study aimed to investigate whether prognostic nutritional index (PNI) is an independent predictor of acute kidney injury (AKI) and mortality of patients in the coronary care unit (CCU). In the present two-stage observational study of patients in the CCU, 6,444 patients from the Medical Information Mart for Intensive Care (MIMIC) III database were first enrolled (test cohort), after which 412 patients from Zhongnan Hospital of Wuhan University were recruited in the validation cohort. AKI was defined based on the Kidney Disease Improving Global Outcomes AKI criteria. The primary endpoint was the incidence of AKI stratified by severity, while the second endpoint included in-hospital mortality and 2-year mortality. In the test cohort, 4,457 (69.2%) patients developed AKI during hospitalization. Following multivariable adjustment, the highest quartile of the PNI value was associated with a 1.8-fold increased risk of AKI compared with the lowest quartile. For the prediction of AKI, the area under the receiver operating characteristic curve outperformed the acute physiology score III score and clinical model in patients with or without preexisting chronic kidney disease, and this was further validated in the hospital cohort used in the present study. A total of 2,219 patients suffered mortality during the 2-year follow-up, and PNI was indicated to independently predict the risk of in-hospital mortality and 2-year mortality in the test cohort and in the validation cohort. Decision curve analysis indicated that the PNI values were clinically useful; Therefore, the current study demonstrated that the PNI value is an independent predictor of AKI and mortality in patients within the CCU.

Early Acute Kidney Injury Associated with Liver Transplantation: A Retrospective Case-Control Study

Background

A retrospective case-control study was carried out to assess the occurrence of acute kidney injury (AKI) in liver transplantation (LT) recipients and its related risk factors.

Material/Methods

The study enrolled 131 patients undergoing LT from December 2017 to June 2019 at Beijing Tsinghua Chang Gung Hospital, China. AKI and its classification were defined according to KDIGO guidelines. We collected patients’ demographic characteristics and perioperative parameters, and identified independent risk factors of AKI by multivariate logistic regression analysis.

Results

We included 122 patients in analysis. AKI occurred in 52 (42.6%) patients (22.1% stage I, 8.2% stage II, and 12.3% stage III). AKI was notably associated with 12 factors: sex, body mass index (BMI), hepatic etiology, MELD score, ascites, prothrombin time (PT), international normalized ratio of prothrombin time (INR), preoperative total bilirubin (TBIL), operative time, total fluid intake, fresh frozen plasma (FFP), and estimated blood loss (EBL) (P<0.05). The factors independently associated with AKI were BMI (adjusted odds ratio: 0.605, 95% confidence interval: 0.425–0.859; P=0.005) and intraoperative FFP infusion (adjusted odds ratio: 0.998, 95% confidence interval: 0.995–1.000; P=0.047). Compared with the non-AKI group, the AKI group showed higher likelihood of renal replacement therapy (RRT), and longer ICU and hospital stays, higher in-hospital mortality, and higher hospitalization costs (P<0.05).

Conclusions

There is a high risk of AKI in patients undergoing LT. BMI and intraoperative FFP infusion are factors independently correlated with AKI. AKI can result in extended hospital stays and higher hospitalization expenses.

Association of overweight with postoperative acute kidney injury among patients receiving orthotopic liver transplantation: an observational cohort study

BACKGROUND

Acute kidney injury (AKI) is a common postoperative complication of orthotopic liver transplantation (OLT). So far, little attention has been paid on the association between overweight and AKI after OLT, and animal models or clinical studies have drawn conflicting conclusions. The objective of our study was to determine whether overweight (BMI [Body Mass Index] ≥ 25 kg/m2) is associated with an increased risk of AKI after OLT.

METHODS

This retrospective cohort study included 244 patients receiving OLT in the Affiliated Hospital of Qingdao University between January 1, 2017, and August 29, 2019. Preoperative, intraoperative, and postoperative data were collected retrospectively. The primary outcome was the development of AKI as defined by Kidney Disease, Improving Global Outcome (KIDGO) staging system. Logistic regression analysis was used to determine the relationship between overweight and the occurrence of postoperative AKI. Data analysis was conducted from September to October 2019, revision in April 2020.

RESULTS

Among 244 patients receiving OLT (mean [standard deviation] age, 54.1 [9.6] years; 84.0% male) identified, 163 patients (66.8%) developed postoperative AKI. Overweight (BMI ≥ 25 kg/m2) was associated with a higher rate of postoperative severe AKI (stage 2/3) compared with normal weight (18.5 ≤ BMI < 25 kg/m2) (41 [47.7%] vs 39 [28.7%]; adjusted odds ratio [OR], 2.539; 95% confidence interval [CI], 1.389-4.642; P = 0.002). Furthermore, patients with obese were at even higher risk of postoperative severe AKI after controlling for confounding factors (adjusted OR: 3.705; 95% CI: 1.108-12.388; P = 0.033).

CONCLUSIONS

Overweight is independently associated with an increased risk of postoperative severe AKI among patients receiving OLT. The association of BMI with severe AKI after OLT is J-shaped.

Peri-transplant management of nonalcoholic fatty liver disease in liver transplant candidates

The incidence of non-alcoholic fatty liver disease (NAFLD) is rapidly growing, affecting 25% of the world population. Non-alcoholic steatohepatitis (NASH) is the most severe form of NAFLD and affects 1.5% to 6.5% of the world population. Its rising incidence will make end-stage liver disease (ESLD) due to NASH the number one indication for liver transplantation (LT) in the next 10 to 20 years, overtaking Hepatitis C. Patients with NASH also have a high prevalence of associated comorbidities such as type 2 diabetes, obesity, metabolic syndrome, cardiovascular disease, and chronic kidney disease (CKD), which must be adequately managed during the peritransplant period for optimal post-transplant outcomes. The focus of this review article is to provide a comprehensive overview of the unique challenges these patients present in the peritransplant period, which comprises the pre-transplant, intraoperative, and immediate postoperative periods.

Intraoperative Hepatic Blood Inflow Can Predict Early Acute Kidney Injury following DCD Liver Transplantation: A Retrospective Observational Study

Purpose

Acute kidney injury (AKI) is a major and severe complication following donation-after-circulatory-death (DCD) liver transplantation (LT) and is associated with increased postoperative morbidity and mortality. However, the risk factors and the prognosis factors of AKI still need to be further explored, and the relativity of intraoperative hepatic blood inflow (HBI) and AKI following LT has not been discussed yet. The purpose of this study was to investigate the correlation between HBI and AKI and to construct a prediction model of early acute kidney injury (EAKI) following DCD LT with the combination of HBI and other clinical parameters.

Methods

Clinical data of 132 patients who underwent DCD liver transplantation at the first hospital of China Medical University from April 2005 to March 2017 were analyzed. Data of 105 patients (the first ten years of patients) were used to develop the prediction model. Then we assessed the clinical usefulness of the prediction models in the validation cohort (27 patients). EAKI according to Kidney Disease Improving Global Outcomes (KDIGO) criteria based on serum creatinine increase during 7-day of postoperative follow-up.

Results

After Least Absolute Shrinkage and Selection Operator (LASSO) regression and simplification, a simplified prediction model consisting of the Child-Turcotte-Pugh (CTP) score (p=0.033), anhepatic phase (p=0.014), packed red blood cell (pRBC) transfusion (p=0.027), and the HBI indexed by height (HBI/h) (p=0.002) was established. The C-indexes of the model in the development and validation cohort were 0.823 [95% CI, 0.738-0.908] and 0.921 [95% CI, 0.816-1.000], respectively.

Conclusions

In this study, we demonstrated the utility of HBI/h as a predictor for EAKI following DCD LT, as well as the clinical usefulness of the prediction model through the combination of the CTP score, anhepatic phase, pRBC transfusion and HBI/h.