Liver transplantation is a critical and evolving field in modern medicine, offering life-saving treatment for patients with end-stage liver disease and other hepatic conditions. Despite its transformative potential, transplantation faces persistent challenges, including a global organ shortage, increasing liver disease prevalence, and significant waitlist mortality rates. Current donor evaluation practices often discard potentially viable livers, underscoring the need for refined graft assessment tools. This review explores advancements in graft evaluation and utilization aimed at expanding the donor pool and optimizing outcomes. Emerging technologies, such as imaging techniques, dynamic functional tests, and biomarkers, are increasingly critical for donor assessment, especially for marginal grafts. Machine learning and artificial intelligence, exemplified by tools like LiverColor, promise to revolutionize donor-recipient matching and liver viability predictions, while bioengineered liver grafts offer a future solution to the organ shortage. Advances in perfusion techniques are improving graft preservation and function, particularly for donation after circulatory death (DCD) grafts. While challenges remain-such as graft rejection, ischemia-reperfusion injury, and recurrence of liver disease-technological and procedural advancements are driving significant improvements in graft allocation, preservation, and post-transplant outcomes. This review highlights the transformative potential of integrating modern technologies and multidisciplinary approaches to expand the donor pool and improve equity and survival rates in liver transplantation.

Liver transplantation (LT) is a life-saving intervention for patients with end-stage liver disease. However, the equitable allocation of scarce donor organs remains a formidable challenge. Prognostic tools are pivotal in identifying the most suitable transplant candidates. Traditionally, scoring systems like the model for end-stage liver disease have been instrumental in this process. Nevertheless, the landscape of prognostication is undergoing a transformation with the integration of machine learning (ML) and artificial intelligence models.

To assess the utility of ML models in prognostication for LT, comparing their per formance and reliability to established traditional scoring systems.

Following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines, we conducted a thorough and standardized literature search using the PubMed/MEDLINE database. Our search imposed no restrictions on publication year, age, or gender. Exclusion criteria encompassed non-English stu dies, review articles, case reports, conference papers, studies with missing data, or those exhibiting evident methodological flaws.

Our search yielded a total of 64 articles, with 23 meeting the inclusion criteria. Among the selected studies, 60.8% originated from the United States and China combined. Only one pediatric study met the criteria. Notably, 91% of the studies were published within the past five years. ML models consistently demonstrated satisfactory to excellent area under the receiver operating characteristic curve values (ranging from 0.6 to 1) across all studies, surpassing the performance of traditional scoring systems. Random forest exhibited superior predictive capa bilities for 90-d mortality following LT, sepsis, and acute kidney injury (AKI). In contrast, gradient boosting excelled in predicting the risk of graft-versus-host disease, pneumonia, and AKI.

This study underscores the potential of ML models in guiding decisions related to allograft allocation and LT, marking a significant evolution in the field of prognostication.

Early allograft dysfunction (EAD) is a criterion to evaluate initial graft dysfunction associated with inferior graft survival and postoperative complications after liver transplantation (LT). This study defined the lactate-to-platelet ratio (LPR) as lactate level immediately post-LT/platelet count on postoperative day 1 and evaluated its association with EAD and short-term graft failure.

This study reviewed 434 deceased-donor LTs from individuals with confirmed brain death between January 2008 and December 2014. The area under the curve (AUC) was used to compare the predictive capacity for 3-month graft survival between EAD and the LPR. Along with LPR, the risk factors for 3-month graft failure were analyzed by multivariate analysis.

EAD was reported in 127 patients (31%). The LPR in patients with EAD was significantly higher than that in patients without EAD (9.8 vs 5.9, P < .001). In the multivariate analysis, both the LPR (per 1.0 increase) and EAD were independent risk factors for 3-month graft failure (hazard ratio [HR] =1.03, P = .03; and HR = 9.14, P = .001). The comparison of the AUCs between the LPR and EAD showed no significant difference (0.79 vs 0.78, P = .84), whereas the combination of EAD and LPR had a better predictive capacity than EAD alone (0.86 vs 0.78, P < .001). The LPR showed an inverse relationship for predicting 3-month graft survival.

The LPR is a continuous parameter that enables prediction of initial graft function and estimation of the 3-month graft failure rate with the advantages of early availability and simple calculations.

For liver transplant candidates on the waiting list, deciding to accept a donor organ with known or potential risk factors can be stressful and can lead to declined offers. Current education for patients and family often takes place during transplant evaluations and can be overwhelming and result in low retention and poor understanding of donor quality.

In the first phase, we sought to understand provider experiences when counseling patients about donor risks and donor offers. We conducted interviews and focus groups with liver transplant providers at 1 local center and at a national clinician conference. Twenty providers participated: 15 hepatologists and 5 surgeons. The provider feedback was used to create an initial outline of content that is consistent with decision support frameworks. In a second phase, graphic design collaborators created mockups of a patient-friendly tool. We reviewed mockups with 4 transplant coordinators and 9 liver transplant candidates for feedback on clarity and utility to prepare for an organ offer. Patient responses allowed a comparison of perceived readiness to receive an offer call before and after viewing mockups.

We identified themes relating to the offer process, repetition and timing of education, and standardization and tailoring of content. The results indicated a gap in available education after the evaluation session, and information specific to offer decisions is needed. Patient feedback emphasized the need to review the offer process before a real offer.

Patients and providers responded favorably to a patient tool addressing existing gaps in education while waiting for a donor offer. Additional patient, family, and provider feedback will guide the development of an interactive tool to prepare patients and families for an offer decision.

In an attempt to further improve liver allograft utilization and outcome in orthotopic liver transplantation (OLT), a variety of clinical scoring systems have been developed. Here we aimed to comparatively investigate the association of the Balance-of-Risk (BAR), Survival-Outcomes-Following-Liver-Transplant (SOFT), Preallocation-Survival-Outcomes-Following-Liver-Transplant (pSOFT), Donor-Risk-Index (DRI), and the Eurotransplant-Donor-Risk-Index (ET-DRI) scores with short- and long-term outcome following OLT.

We included 338 consecutive patients, who underwent OLT in our institution between May 2010 and November 2017. For each prognostic model, the optimal cutoff values were determined with the help of the Youden-index and their diagnostic accuracy for 90-day post OLT-mortality and major postoperative complications was measured by the area under the receiver operating characteristic curve (AUROC). Patient- and graft survival were analyzed using the Kaplan-Meier method and the log-rank test. Morbidity was assessed using the Clavien-Dindo classification and the Comprehensive-Complication-Index.

BAR, SOFT, and pSOFT performed well above the conventional AUROC-threshold of 0.70 with good prediction of early mortality. Only BAR showed AUC>0.70 for both mortality and major morbidity. With the cutoffs of 14, 31, and 22 respectively for BAR, SOFT, and pSOFT, subgroup analysis showed significant differences (p<0.001) in morbidity and mortality, length of intensive care- and hospital-stay and early allograft dysfunction rates. Five-years patient survival was inferior in the high BAR, pSOFT, and SOFT groups.

Out of all scores tested, the BAR-score had the best value in predicting both 90-day morbidity and mortality after OLT showing the highest AUCs. The pSOFT and SOFT scores demonstrated an acceptable accuracy in predicting 90-day morbidity and mortality. The used BAR, SOFT, and pSOFT cutoffs allowed the identification of patients at risk in terms of five-year patient survival. The DRI and ET-DRI scores have failed to predict recipient outcomes in the present setting.

The Share 35 policy for liver allocation prioritizes patients with Model for End-Stage Liver Disease (MELD) scores ≥ 35 for regional sharing of liver allografts. To better assess donor-recipient interactions and inform expectations, this study identified factors affecting graft survival independent of MELD score and derived a risk index for transplantation in the MELD ≥ 35 population.

The United Network for Organ Sharing (UNOS) STAR database was evaluated for deceased donor liver transplants with recipients' MELD ≥ 35, between January 2006 and June 2016. Data were randomly split into test and validate cohorts. Four individual models of graft survival spanning 90 days to 5 years were evaluated with univariate and multivariate Cox proportional hazards analyses against donor- and recipient-specific characteristics. Significant factors were compiled to generate the Liver Transplant Survival Index (LTSI-35), and survival analyses were performed.

Five risk groups (very low, low, moderate, high, and severe) were identified, with 1-year graft survival rates of 90.8% ± 0.2%, 89.3% ± 0.3%, 85.0% ± 0.3%, 79.8% ± 0.3%, and 70.3% ± 0.4% (p < 0.001 across groups), respectively. The greatest risk of graft loss was associated with donation after circulatory death (DCD) donors (1-year hazard ratio [HR] = 1.61 [95% CI 1.26 to 2.05], p = 0.001), recipients' requiring ventilator support (HR 1.32 [95% CI 1.17 to 1.51], p < 0.001), and recipient portal vein thrombosis (HR 1.21 [95% CI 1.03 to 1.42], p = 0.003). Subgroup analysis revealed increased risk of graft loss with graft macrosteatosis ≥ 30% on pre-donation biopsy at 90 days (HR 1.64 [1.33 to 1.99], p < 0.001).

The LTSI-35 identifies risk factors for graft loss in a high-MELD population which, when combined, may portend worse outcomes. The LTSI-35 may be used to influence donor selection, organ allocation, and to inform expectations for allograft survival.

In 2006, derivation of the donor risk index (DRI) highlighted the importance of donor factors for successful liver transplantation. Over the last decade, the DRI has served as a useful metric of donor quality and has enhanced our understanding of donor factors and their impact upon recipients with hepatitis C virus, those with low Model for End-Stage Liver Disease (MELD) score, and individuals undergoing retransplantation. DRI has provided the transplant community with a common language for describing donor organ characteristics and has served as the foundation for several tools for organ risk assessment. It is a useful tool in assessing the interactions of donor factors with recipient factors and their impact on posttransplant outcomes. However, limitations of statistical modeling, choice of donor factors, exclusion of unaccounted donor and geographic factors, and the changing face of the liver transplant recipient have tempered its widespread use. In addition, the DRI was derived from data before the MELD era but is currently being applied to expand the donor pool while concurrently meeting the demands of a dynamic allocation system. A decade after its introduction, DRI remains relevant but may benefit from being updated to provide guidance in the use of extended criteria donors by accounting for the impact of geography and unmeasured donor characteristics. DRI could be better adapted for recipients with nonalcoholic fatty liver disease by examining and including recipient factors unique to this population. Liver Transplantation 23 1216-1225 2017 AASLD.

The ability to predict graft failure or primary nonfunction at liver transplant decision time assists utilization of scarce resource of donor livers, while ensuring that patients who are urgently requiring a liver transplant are prioritized. An index that is derived to predict graft failure using donor and recipient factors, based on local data sets, will be more beneficial in the Australian context.

Liver transplant data from the Austin Hospital, Melbourne, Australia, from 2010 to 2013 has been included in the study. The top 15 donor, recipient, and transplant factors influencing the outcome of graft failure within 30 days were selected using a machine learning methodology. An algorithm predicting the outcome of interest was developed using those factors.

Donor Risk Index predicts the outcome with an area under the receiver operating characteristic curve (AUC-ROC) value of 0.680 (95% confidence interval [CI], 0.669-0.690). The combination of the factors used in Donor Risk Index with the model for end-stage liver disease score yields an AUC-ROC of 0.764 (95% CI, 0.756-0.771), whereas survival outcomes after liver transplantation score obtains an AUC-ROC of 0.638 (95% CI, 0.632-0.645). The top 15 donor and recipient characteristics within random forests results in an AUC-ROC of 0.818 (95% CI, 0.812-0.824).

Using donor, transplant, and recipient characteristics known at the decision time of a transplant, high accuracy in matching donors and recipients can be achieved, potentially providing assistance with clinical decision making.

With restricted numbers of available organs, futility in liver transplantation has to be avoided. The concept of dynamic changes in MELD score (DeltaMELD) has previously been shown to be a simple tool to identify patients with the greatest risk of death after transplantation. Aim was to validate this concept with the Eurotransplant (ET) database.

A retrospective registry analysis was performed on all patients listed for liver transplantation within ET between 2006 and 2011. Patients <18 years of age, acute liver failure, malignancy and patients listed for retransplantation were excluded. Influence of MELD at listing (MELDon), MELD at transplantation (MELDoff), DeltaMELD, age, sex, underlying disease and time on the waiting list on overall survival after liver transplantation were evaluated.

A total of 16 821 patients were listed for liver transplantation, 8096 met the inclusion criteria. Age, MELD on and DeltaMELD showed significant influence on survival on the waiting list. Age and DeltaMELD showed influence on survival after liver transplantation, with DeltaMELD>10 showing a 1.6-fold increased risk of death.

The concept of DeltaMELD was validated in a large, prospective data set. It provides a simple tool to identify patients with increased risk of death after liver transplantation and might help improve long-term results.

Since March 26, 2012, the Kidney Donor Profile Index (KDPI) has been provided with all deceased-donor kidney offers, with the goal of improving the expanded criteria donor (ECD) indicator. Although an improved risk index may facilitate identification and transplantation of marginal yet viable kidneys, a granular percentile system may reduce provider-patient communication flexibility, paradoxically leading to more discards ("labeling effect"). We studied the discard rates of the kidneys recovered for transplantation between March 26, 2010 and March 25, 2012 ("ECD era," N = 28 636) and March 26, 2012 and March 25, 2014 ("KDPI era," N = 29 021) using Scientific Registry of Transplant Recipients (SRTR) data. There was no significant change in discard rate from ECD era (18.1%) to KDPI era (18.3%) among the entire population (adjusted odds ratio [aOR] = 0.97 1.041.10 , p = 0.3), or in any KDPI stratum. However, among kidneys in which ECD and KDPI indicators were discordant, "high risk" standard criteria donor (SCD) kidneys (with KDPI > 85) were at increased risk of discard in the KDPI era (aOR = 1.07 1.421.89 , p = 0.02). Yet, recipients of these kidneys were at much lower risk of death (adjusted Risk Ratio [aRR] = 0.56 0.770.94 at 2 years posttransplant) compared to those remaining on dialysis waiting for low-KDPI kidneys. Our findings suggest that there might be an unexpected, harmful labeling effect of reporting a high KDPI for SCD kidneys, without the expected advantage of providing a more granular risk index.

Recently the Eurotransplant donor risk index (ET-DRI) was published, a model based on data from the Eurotransplant database that can be used for risk indication of liver donors within the Eurotransplant region. Because outcome after liver transplantation (LT) depends both on donor and recipient risk factors, a combined donor-recipient model (DRM) would give a more complete picture of the overall risk involved. All liver transplants in adult recipients from January 1, 2008 to December 31, 2010 in the Eurotransplant region were included. Risk factors in donors and recipients for failure-free (retransplant free) survival were analyzed in univariate and multivariate analyses. A simplified recipient risk index (sRRI) was constructed using all available recipient factors. A total of 4466 liver transplants were analyzed. Median donor risk index and ET-DRI were 1.78 and 1.91, respectively. The ET-DRI was validated in this new cohort (P < 0.001; concordance index [c-index], 0.59). After construction of a simplified recipient risk index of significant recipient factors, Cox regression analysis showed that the combination ET-DRI and sRRI into a new DRM gave the highest predictive value (P < 0.001; c-index, 0.62). The combined model of ET-DRI and sRRI gave a significant prediction of outcome after orthotopic LT in the Eurotransplant region, better than the ET-DRI alone. This DRM has potential in comparing data in the literature and correcting for sickness/physical condition of transplant recipients. It is a first step toward benchmarking of graft survival in the Eurotransplant region.

Donor age has become the dominant donor factor used to predict graft failure (GF) after liver transplantation (LT) in hepatitis C virus (HCV) recipients. The purpose of this study was to develop and validate a model of corrected donor age (CDA) for HCV LT recipients that transforms the risk of other donor factors into the scale of donor age. We analyzed all first LT recipients with HCV in the United Network for Organ Sharing (UNOS) registry from January 1998 to December 2007 (development cohort, n = 14,538) and January 2008 to December 2011 (validation cohort, n = 7502) using Cox regression, excluding early GF (<90 days from LT). Accuracy in predicting 1 year GF (death or repeat LT) was assessed with the net reclassification index (NRI). In the development cohort, after controlling for pre-LT recipient factors and geotemporal trends (UNOS region, LT year), the following donor factors were independent predictors of GF, all P < 0.05: donor age (hazard ratio [HR], 1.02/year), donation after cardiac death (DCD; HR, 1.31), diabetes (HR, 1.23), height < 160 cm (HR, 1.13), aspartate aminotransferase (AST) ≥ 120 U/L (HR, 1.10), female (HR, 0.94), cold ischemia time (CIT; HR, 1.02/hour), and non-African American (non-AA) donor-African American (AA) recipient (HR, 1.65). Transforming these risk factors into the donor age scale yielded the following: DCD = +16 years; diabetes = +12 years; height < 160 cm = +7 years; AST ≥ 120 U/L = +5 years; female = -4 years; and CIT = +1 year/hour > 8 hours and -1 year/hour < 8 hours. There was a large effect of donor-recipient race combinations: +29 years for non-AA donor and an AA recipient but only +5 years for an AA donor and an AA recipient, and -2 years for an AA donor and a non-AA recipient. In a validation cohort, CDA better classified risk of 1-year GF versus actual age (NRI, 4.9%; P = 0.009) and versus the donor risk index (9.0%, P < 0.001). The CDA, compared to actual donor age, provides an intuitive and superior estimation of graft quality for HCV-positive LT recipients because it incorporates additional factors that impact LT GF rates.

Clinical and laboratory criteria are not reliable predictors of deceased donor liver graft quality. Intraoperative assessment of experienced surgeons is the gold standard. Standardizing and quantifying this assessment is especially needed now that regional sharing is the rule. We prospectively evaluated a novel, simple, rapid, noninvasive, quantitative measure of liver function performed before graft procurement.

Using a portable, finger-probe-based device, indocyanine green plasma disappearance rates (ICG-PDR) were measured in adult brain-dead donors in the local donor service area before organ procurement. Results were compared with graft function and outcomes. Both donor and recipient teams were blinded to ICG-PDR measurements.

Measurements were performed on 53 consecutive donors. Eleven liver grafts were declined by all centers because of quality; the other 42 grafts were transplanted. Logistic regression analysis showed ICG-PDR to be the only donor variable to be significantly associated with 7-d graft survival. Donor risk index, donor age, and transaminase levels at peak or procurement were not significantly associated with 7-d graft survival.

We report the successful use of a portable quantitative means of measuring liver function and its association with graft survival. These data warrant further exploration in a variety of settings to evaluate acceptable values for donated liver grafts.