Reliability and Accuracy of Clinical Assessment and Digital Image Analysis for Steatosis Evaluation in Discarded Human Livers

Artificial intelligence-aided steatosis assessment in donor livers according to the Banff consensus recommendations

OBJECTIVES

Severe macrovesicular steatosis in donor livers is associated with primary graft dysfunction. The Banff Working Group on Liver Allograft Pathology has proposed recommendations for steatosis assessment of donor liver biopsy specimens with a consensus for defining "large droplet fat" (LDF) and a 3-step algorithmic approach.

METHODS

We retrieved slides and initial pathology reports from potential liver donor biopsy specimens from 2010 to 2021. Following the Banff approach, we reevaluated LDF steatosis and employed a computer-assisted manual quantification protocol and artificial intelligence (AI) model for analysis.

RESULTS

In a total of 113 slides from 88 donors, no to mild (<33%) macrovesicular steatosis was reported in 88.5% (100/113) of slides; 8.8% (10/113) was reported as at least moderate steatosis (≥33%) initially. Subsequent pathology evaluation, following the Banff recommendation, revealed that all slides had LDF below 33%, a finding confirmed through computer-assisted manual quantification and an AI model. Correlation coefficients between pathologist and computer-assisted manual quantification, between computer-assisted manual quantification and the AI model, and between the AI model and pathologist were 0.94, 0.88, and 0.81, respectively (P < .0001 for all).

CONCLUSIONS

The 3-step approach proposed by the Banff Working Group on Liver Allograft Pathology may be followed when evaluating steatosis in donor livers. The AI model can provide a rapid and objective assessment of liver steatosis.

Development of a portable device to quantify hepatic steatosis in potential donor livers

An accurate estimation of liver fat content is necessary to predict how a donated liver will function after transplantation. Currently, a pathologist needs to be available at all hours of the day, even at remote hospitals, when an organ donor is procured. Even among expert pathologists, the estimation of liver fat content is operator-dependent. Here we describe the development of a low-cost, end-to-end artificial intelligence platform to evaluate liver fat content on a donor liver biopsy slide in real-time. The hardware includes a high-resolution camera, display, and GPU to acquire and process donor liver biopsy slides. A deep learning model was trained to label and quantify fat globules in liver tissue. The algorithm was deployed on the device to enable real-time quantification and characterization of fat content for transplant decision-making. This information is displayed on the device and can also be sent to a cloud platform for further analysis.

Digital imaging software versus the "eyeball" method in quantifying steatosis in a liver biopsy

Steatotic livers represent a potentially underutilized resource to increase the donor graft pool; however, 1 barrier to the increased utilization of such grafts is the heterogeneity in the definition and the measurement of macrovesicular steatosis (MaS). Digital imaging software (DIS) may better standardize definitions to study posttransplant outcomes. Using HALO, a DIS, we analyzed 63 liver biopsies, from 3 transplant centers, transplanted between 2016 and 2018, and compared macrovesicular steatosis percentage (%MaS) as estimated by transplant center, donor hospital, and DIS. We also quantified the relationship between DIS characteristics and posttransplant outcomes using log-linear regression for peak aspartate aminotransferase, peak alanine aminotransferase, and total bilirubin on postoperative day 7, as well as logistic regression for early allograft dysfunction. Transplant centers and donor hospitals overestimated %MaS compared with DIS, with better agreement at lower %MaS and less agreement for higher %MaS. No DIS analyzed liver biopsies were calculated to be >20% %MaS; however, 40% of liver biopsies read by transplant center pathologists were read to be >30%. Percent MaS read by HALO was positively associated with peak aspartate aminotransferase (regression coefficient= 1.04 1.08 1.12 , p <0.001), peak alanine aminotransferase (regression coefficient = 1.04 1.08 1.12 , p <0.001), and early allograft dysfunction (OR= 1.10 1.40 1.78 , p =0.006). There was no association between HALO %MaS and total bilirubin on postoperative day 7 (regression coefficient = 0.99 1.01 1.04 , p =0.3). DIS provides reproducible quantification of steatosis that could standardize MaS definitions and identify phenotypes associated with good clinical outcomes to increase the utilization of steatite livers.

Artificial intelligence for prediction of donor liver allograft steatosis and early post-transplantation graft failure

BACKGROUND

Donor livers undergo subjective pathologist review of steatosis before transplantation to mitigate the risk for early allograft dysfunction (EAD). We developed an objective, computer vision artificial intelligence (CVAI) platform to score donor liver steatosis and compared its capability for predicting EAD against pathologist steatosis scores.

METHODS

Two pathologists scored digitized donor liver biopsy slides from 2014 to 2019. We trained four CVAI platforms with 1:99 training:prediction split. Mean intersection-over-union (IU) characterized CVAI model accuracy. We defined EAD using liver function tests within 1 week of transplantation. We calculated separate EAD logistic regression models with CVAI and pathologist steatosis and compared the models' discrimination and internal calibration.

RESULTS

From 90 liver biopsies, 25,494 images trained CVAI models yielding peak mean IU = 0.80. CVAI steatosis scores were lower than pathologist scores (median 3% vs 20%, P < 0.001). Among 41 transplanted grafts, 46% developed EAD. The median CVAI steatosis score was higher for those with EAD (2.9% vs 1.9%, P = 0.02). CVAI steatosis was independently associated with EAD after adjusting for donor age, donor diabetes, and MELD score (aOR = 1.34, 95%CI = 1.03-1.75, P = 0.03).

CONCLUSION

The CVAI steatosis EAD model demonstrated slightly better calibration than pathologist steatosis, meriting further investigation into which modality most accurately and reliably predicts post-transplantation outcomes.

Mitochondrial respiratory chain and Krebs cycle enzyme function in human donor livers subjected to end-ischaemic hypothermic machine perfusion

INTRODUCTION

Marginal human donor livers are highly susceptible to ischaemia reperfusion injury and mitochondrial dysfunction. Oxygenation during hypothermic machine perfusion (HMP) was proposed to protect the mitochondria but the mechanism is unclear. Additionally, the distribution and uptake of perfusate oxygen during HMP are unknown. This study aimed to examine the feasibility of mitochondrial function analysis during end-ischaemic HMP, assess potential mitochondrial viability biomarkers, and record oxygenation kinetics.

METHODS

This was a randomised pilot study using human livers retrieved for transplant but not utilised. Livers (n = 38) were randomised at stage 1 into static cold storage (n = 6), hepatic artery HMP (n = 7), and non-oxygen supplemented portal vein HMP (n = 7) and at stage 2 into oxygen supplemented and non-oxygen supplemented portal vein HMP (n = 11 and 7, respectively). Mitochondrial parameters were compared between the groups and between low- and high-risk marginal livers based on donor history, organ steatosis and preservation period. The oxygen delivery efficiency was assessed in additional 6 livers using real-time measurements of perfusate and parenchymal oxygen.

RESULTS

The change in mitochondrial respiratory chain (complex I, II, III, IV) and Krebs cycle enzyme activity (aconitase, citrate synthase) before and after 4-hour preservation was not different between groups in both study stages (p > 0.05). Low-risk livers that could have been used clinically (n = 8) had lower complex II-III activities after 4-hour perfusion, compared with high-risk livers (73 nmol/mg/min vs. 113 nmol/mg/min, p = 0.01). Parenchymal pO2 was consistently lower than perfusate pO2 (p ≤ 0.001), stabilised in 28 minutes compared to 3 minutes in perfusate (p = 0.003), and decreased faster upon oxygen cessation (75 vs. 36 minutes, p = 0.003).

CONCLUSIONS

Actively oxygenated and air-equilibrated end-ischaemic HMP did not induce oxidative damage of aconitase, and respiratory chain complexes remained intact. Mitochondria likely respond to variable perfusate oxygen levels by adapting their respiratory function during end-ischaemic HMP. Complex II-III activities should be further investigated as viability biomarkers.

Contemporary strategies to assess and manage liver donor steatosis: a review

PURPOSE OF REVIEW

Due to a persistent shortage of donor livers, attention has turned toward ways of utilizing marginal grafts, particularly those with steatosis, without incurring inferior outcomes. Here we review the evaluation and utilization of steatotic liver allografts, highlight recently published data, and discuss novel methods of graft rehabilitation.

RECENT FINDINGS

Although severe liver allograft (>60%) steatosis has been associated with inferior graft and recipient outcomes, mild (<30%) steatosis has not. There is ongoing debate regarding safe utilization of grafts with moderate (30-60%) steatosis. Presently, no established protocols for evaluating steatosis in donor candidates or utilizing such grafts exist. Liver biopsy is accepted as the gold standard technique, though noninvasive methods have shown promise in accurately predicting steatosis. More recently, machine perfusion has been shown to enhance ex situ liver function and reduce steatosis, emerging as a potential means of optimizing steatotic grafts prior to transplantation.

SUMMARY

Steatotic liver allografts constitute a large proportion of deceased donor organs. Further work is necessary to define safe upper limits for the acceptable degree of steatosis, develop standardized evaluation protocols, and establish utilization guidelines that prioritize safety. Machine perfusion has shown promise in rehabilitating steatotic grafts and offers the possibility of expanding the deceased donor pool.

Precision transplant pathology

PURPOSE OF REVIEW

Transplant pathology contributes substantially to personalized treatment of organ allograft recipients. Rapidly advancing next-generation human leukocyte antigen (HLA) sequencing and pathology are enhancing the abilities to improve donor/recipient matching and allograft monitoring.

RECENT FINDINGS

The present review summarizes the workflow of a prototypical patient through a pathology practice, highlighting histocompatibility assessment and pathologic review of tissues as areas that are evolving to incorporate next-generation technologies while emphasizing critical needs of the field.

SUMMARY

Successful organ transplantation starts with the most precise pratical donor-recipient histocompatibility matching. Next-generation sequencing provides the highest resolution donor-recipient matching and enables eplet mismatch scores and more precise monitoring of donor-specific antibodies (DSAs) that may arise after transplant. Multiplex labeling combined with hand-crafted machine learning is transforming traditional histopathology. The combination of traditional blood/body fluid laboratory tests, eplet and DSA analysis, traditional and next-generation histopathology, and -omics-based platforms enables risk stratification and identification of early subclinical molecular-based changes that precede a decline in allograft function. Needs include software integration of data derived from diverse platforms that can render the most accurate assessment of allograft health and needs for immunosuppression adjustments.